• Your next nonfiction book could write itself, but you’ll own the rights

    TL;DR: Turn ideas into full-length books with AI—lifetime access for just Writing a book takes time—something most of us don’t have between inbox chaos and back-to-back meetings. But what if all you needed was an idea? That’s where YouBooks steps in. This AI-powered tool helps you generate full-length nonfiction books with just a few prompts, and right now, you can lock in lifetime access for.
    YouBooks pulls from several top-tier AI models, like ChatGPT, Claude, and Gemini, and combines them with live web research to build out detailed, structured manuscripts up to 300,000 words. Whether you want to write about productivity, startup culture, parenting, or personal finance, feed in your topic and let the AI do the heavy lifting.
    Why is Youbooks for you?

    150,000 credits per monthDownloadable formats: PDF, DOCX, EPUB
    Commercial rights so that you can sell, share, or publish your books
    Custom style options to match your tone or brand

    It’s a serious time-saver if you’ve been sitting on an idea forever or want to build a content empire without writing every word yourself. Plus, unlike many AI tools, YouBooks gives you full ownership of the content you create.

    Snag a lifetime subscription to YouBooks for  and start turning your thoughts into fully formed nonfiction books: no ghostwriters, no subscriptions, and no gatekeepers.

    Youbooks – AI Nonfiction Book Generator: Lifetime SubscriptionSee Deal
    StackSocial prices subject to change.
    #your #next #nonfiction #book #could
    Your next nonfiction book could write itself, but you’ll own the rights
    TL;DR: Turn ideas into full-length books with AI—lifetime access for just Writing a book takes time—something most of us don’t have between inbox chaos and back-to-back meetings. But what if all you needed was an idea? That’s where YouBooks steps in. This AI-powered tool helps you generate full-length nonfiction books with just a few prompts, and right now, you can lock in lifetime access for. YouBooks pulls from several top-tier AI models, like ChatGPT, Claude, and Gemini, and combines them with live web research to build out detailed, structured manuscripts up to 300,000 words. Whether you want to write about productivity, startup culture, parenting, or personal finance, feed in your topic and let the AI do the heavy lifting. Why is Youbooks for you? 150,000 credits per monthDownloadable formats: PDF, DOCX, EPUB Commercial rights so that you can sell, share, or publish your books Custom style options to match your tone or brand It’s a serious time-saver if you’ve been sitting on an idea forever or want to build a content empire without writing every word yourself. Plus, unlike many AI tools, YouBooks gives you full ownership of the content you create. Snag a lifetime subscription to YouBooks for  and start turning your thoughts into fully formed nonfiction books: no ghostwriters, no subscriptions, and no gatekeepers. Youbooks – AI Nonfiction Book Generator: Lifetime SubscriptionSee Deal StackSocial prices subject to change. #your #next #nonfiction #book #could
    WWW.PCWORLD.COM
    Your next nonfiction book could write itself, but you’ll own the rights
    TL;DR: Turn ideas into full-length books with AI—lifetime access for just $49. Writing a book takes time—something most of us don’t have between inbox chaos and back-to-back meetings. But what if all you needed was an idea? That’s where YouBooks steps in. This AI-powered tool helps you generate full-length nonfiction books with just a few prompts, and right now, you can lock in lifetime access for $49 (reg. $540). YouBooks pulls from several top-tier AI models, like ChatGPT, Claude, and Gemini, and combines them with live web research to build out detailed, structured manuscripts up to 300,000 words. Whether you want to write about productivity, startup culture, parenting, or personal finance, feed in your topic and let the AI do the heavy lifting. Why is Youbooks for you? 150,000 credits per month (1 word = 1 credit) Downloadable formats: PDF, DOCX, EPUB Commercial rights so that you can sell, share, or publish your books Custom style options to match your tone or brand It’s a serious time-saver if you’ve been sitting on an idea forever or want to build a content empire without writing every word yourself. Plus, unlike many AI tools, YouBooks gives you full ownership of the content you create. Snag a lifetime subscription to YouBooks for $49 and start turning your thoughts into fully formed nonfiction books: no ghostwriters, no subscriptions, and no gatekeepers. Youbooks – AI Nonfiction Book Generator: Lifetime SubscriptionSee Deal StackSocial prices subject to change.
    Like
    Love
    Wow
    Sad
    Angry
    516
    0 Commentarii 0 Distribuiri
  • 9 menial tasks ChatGPT can handle in seconds, saving you hours

    ChatGPT is rapidly changing the world. The process is already happening, and it’s only going to accelerate as the technology improves, as more people gain access to it, and as more learn how to use it.
    What’s shocking is just how many tasks ChatGPT is already capable of managing for you. While the naysayers may still look down their noses at the potential of AI assistants, I’ve been using it to handle all kinds of menial tasks for me. Here are my favorite examples.

    Further reading: This tiny ChatGPT feature helps me tackle my days more productively

    Write your emails for you
    Dave Parrack / Foundry
    We’ve all been faced with the tricky task of writing an email—whether personal or professional—but not knowing quite how to word it. ChatGPT can do the heavy lifting for you, penning theperfect email based on whatever information you feed it.
    Let’s assume the email you need to write is of a professional nature, and wording it poorly could negatively affect your career. By directing ChatGPT to write the email with a particular structure, content, and tone of voice, you can give yourself a huge head start.
    A winning tip for this is to never accept ChatGPT’s first attempt. Always read through it and look for areas of improvement, then request tweaks to ensure you get the best possible email. You canalso rewrite the email in your own voice. Learn more about how ChatGPT coached my colleague to write better emails.

    Generate itineraries and schedules
    Dave Parrack / Foundry
    If you’re going on a trip but you’re the type of person who hates planning trips, then you should utilize ChatGPT’s ability to generate trip itineraries. The results can be customized to the nth degree depending on how much detail and instruction you’re willing to provide.
    As someone who likes to get away at least once a year but also wants to make the most of every trip, leaning on ChatGPT for an itinerary is essential for me. I’ll provide the location and the kinds of things I want to see and do, then let it handle the rest. Instead of spending days researching everything myself, ChatGPT does 80 percent of it for me.
    As with all of these tasks, you don’t need to accept ChatGPT’s first effort. Use different prompts to force the AI chatbot to shape the itinerary closer to what you want. You’d be surprised at how many cool ideas you’ll encounter this way—simply nix the ones you don’t like.

    Break down difficult concepts
    Dave Parrack / Foundry
    One of the best tasks to assign to ChatGPT is the explanation of difficult concepts. Ask ChatGPT to explain any concept you can think of and it will deliver more often than not. You can tailor the level of explanation you need, and even have it include visual elements.
    Let’s say, for example, that a higher-up at work regularly lectures everyone about the importance of networking. But maybe they never go into detail about what they mean, just constantly pushing the why without explaining the what. Well, just ask ChatGPT to explain networking!
    Okay, most of us know what “networking” is and the concept isn’t very hard to grasp. But you can do this with anything. Ask ChatGPT to explain augmented reality, multi-threaded processing, blockchain, large language models, what have you. It will provide you with a clear and simple breakdown, maybe even with analogies and images.

    Analyze and make tough decisions
    Dave Parrack / Foundry
    We all face tough decisions every so often. The next time you find yourself wrestling with a particularly tough one—and you just can’t decide one way or the other—try asking ChatGPT for guidance and advice.
    It may sound strange to trust any kind of decision to artificial intelligence, let alone an important one that has you stumped, but doing so actually makes a lot of sense. While human judgment can be clouded by emotions, AI can set that aside and prioritize logic.
    It should go without saying: you don’t have to accept ChatGPT’s answers. Use the AI to weigh the pros and cons, to help you understand what’s most important to you, and to suggest a direction. Who knows? If you find yourself not liking the answer given, that in itself might clarify what you actually want—and the right answer for you. This is the kind of stuff ChatGPT can do to improve your life.

    Plan complex projects and strategies
    Dave Parrack / Foundry
    Most jobs come with some level of project planning and management. Even I, as a freelance writer, need to plan tasks to get projects completed on time. And that’s where ChatGPT can prove invaluable, breaking projects up into smaller, more manageable parts.
    ChatGPT needs to know the nature of the project, the end goal, any constraints you may have, and what you have done so far. With that information, it can then break the project up with a step-by-step plan, and break it down further into phases.
    If ChatGPT doesn’t initially split your project up in a way that suits you, try again. Change up the prompts and make the AI chatbot tune in to exactly what you’re looking for. It takes a bit of back and forth, but it can shorten your planning time from hours to mere minutes.

    Compile research notes
    Dave Parrack / Foundry
    If you need to research a given topic of interest, ChatGPT can save you the hassle of compiling that research. For example, ahead of a trip to Croatia, I wanted to know more about the Croatian War of Independence, so I asked ChatGPT to provide me with a brief summary of the conflict with bullet points to help me understand how it happened.
    After absorbing all that information, I asked ChatGPT to add a timeline of the major events, further helping me to understand how the conflict played out. ChatGPT then offered to provide me with battle maps and/or summaries, plus profiles of the main players.
    You can go even deeper with ChatGPT’s Deep Research feature, which is now available to free users, up to 5 Deep Research tasks per month. With Deep Research, ChatGPT conducts multi-step research to generate comprehensive reportsbased on large amounts of information across the internet. A Deep Research task can take up to 30 minutes to complete, but it’ll save you hours or even days.

    Summarize articles, meetings, and more
    Dave Parrack / Foundry
    There are only so many hours in the day, yet so many new articles published on the web day in and day out. When you come across extra-long reads, it can be helpful to run them through ChatGPT for a quick summary. Then, if the summary is lacking in any way, you can go back and plow through the article proper.
    As an example, I ran one of my own PCWorld articlesthrough ChatGPT, which provided a brief summary of my points and broke down the best X alternative based on my reasons given. Interestingly, it also pulled elements from other articles.If you don’t want that, you can tell ChatGPT to limit its summary to the contents of the link.
    This is a great trick to use for other long-form, text-heavy content that you just don’t have the time to crunch through. Think transcripts for interviews, lectures, videos, and Zoom meetings. The only caveat is to never share private details with ChatGPT, like company-specific data that’s protected by NDAs and the like.

    Create Q&A flashcards for learning
    Dave Parrack / Foundry
    Flashcards can be extremely useful for drilling a lot of information into your brain, such as when studying for an exam, onboarding in a new role, prepping for an interview, etc. And with ChatGPT, you no longer have to painstakingly create those flashcards yourself. All you have to do is tell the AI the details of what you’re studying.
    You can specify the format, as well as various other elements. You can also choose to keep things broad or target specific sub-topics or concepts you want to focus on. You can even upload your own notes for ChatGPT to reference. You can also use Google’s NotebookLM app in a similar way.

    Provide interview practice
    Dave Parrack / Foundry
    Whether you’re a first-time jobseeker or have plenty of experience under your belt, it’s always a good idea to practice for your interviews when making career moves. Years ago, you might’ve had to ask a friend or family member to act as your mock interviewer. These days, ChatGPT can do it for you—and do it more effectively.
    Inform ChatGPT of the job title, industry, and level of position you’re interviewing for, what kind of interview it’ll be, and anything else you want it to take into consideration. ChatGPT will then conduct a mock interview with you, providing feedback along the way.
    When I tried this out myself, I was shocked by how capable ChatGPT can be at pretending to be a human in this context. And the feedback it provides for each answer you give is invaluable for knocking off your rough edges and improving your chances of success when you’re interviewed by a real hiring manager.
    Further reading: Non-gimmicky AI apps I actually use every day
    #menial #tasks #chatgpt #can #handle
    9 menial tasks ChatGPT can handle in seconds, saving you hours
    ChatGPT is rapidly changing the world. The process is already happening, and it’s only going to accelerate as the technology improves, as more people gain access to it, and as more learn how to use it. What’s shocking is just how many tasks ChatGPT is already capable of managing for you. While the naysayers may still look down their noses at the potential of AI assistants, I’ve been using it to handle all kinds of menial tasks for me. Here are my favorite examples. Further reading: This tiny ChatGPT feature helps me tackle my days more productively Write your emails for you Dave Parrack / Foundry We’ve all been faced with the tricky task of writing an email—whether personal or professional—but not knowing quite how to word it. ChatGPT can do the heavy lifting for you, penning theperfect email based on whatever information you feed it. Let’s assume the email you need to write is of a professional nature, and wording it poorly could negatively affect your career. By directing ChatGPT to write the email with a particular structure, content, and tone of voice, you can give yourself a huge head start. A winning tip for this is to never accept ChatGPT’s first attempt. Always read through it and look for areas of improvement, then request tweaks to ensure you get the best possible email. You canalso rewrite the email in your own voice. Learn more about how ChatGPT coached my colleague to write better emails. Generate itineraries and schedules Dave Parrack / Foundry If you’re going on a trip but you’re the type of person who hates planning trips, then you should utilize ChatGPT’s ability to generate trip itineraries. The results can be customized to the nth degree depending on how much detail and instruction you’re willing to provide. As someone who likes to get away at least once a year but also wants to make the most of every trip, leaning on ChatGPT for an itinerary is essential for me. I’ll provide the location and the kinds of things I want to see and do, then let it handle the rest. Instead of spending days researching everything myself, ChatGPT does 80 percent of it for me. As with all of these tasks, you don’t need to accept ChatGPT’s first effort. Use different prompts to force the AI chatbot to shape the itinerary closer to what you want. You’d be surprised at how many cool ideas you’ll encounter this way—simply nix the ones you don’t like. Break down difficult concepts Dave Parrack / Foundry One of the best tasks to assign to ChatGPT is the explanation of difficult concepts. Ask ChatGPT to explain any concept you can think of and it will deliver more often than not. You can tailor the level of explanation you need, and even have it include visual elements. Let’s say, for example, that a higher-up at work regularly lectures everyone about the importance of networking. But maybe they never go into detail about what they mean, just constantly pushing the why without explaining the what. Well, just ask ChatGPT to explain networking! Okay, most of us know what “networking” is and the concept isn’t very hard to grasp. But you can do this with anything. Ask ChatGPT to explain augmented reality, multi-threaded processing, blockchain, large language models, what have you. It will provide you with a clear and simple breakdown, maybe even with analogies and images. Analyze and make tough decisions Dave Parrack / Foundry We all face tough decisions every so often. The next time you find yourself wrestling with a particularly tough one—and you just can’t decide one way or the other—try asking ChatGPT for guidance and advice. It may sound strange to trust any kind of decision to artificial intelligence, let alone an important one that has you stumped, but doing so actually makes a lot of sense. While human judgment can be clouded by emotions, AI can set that aside and prioritize logic. It should go without saying: you don’t have to accept ChatGPT’s answers. Use the AI to weigh the pros and cons, to help you understand what’s most important to you, and to suggest a direction. Who knows? If you find yourself not liking the answer given, that in itself might clarify what you actually want—and the right answer for you. This is the kind of stuff ChatGPT can do to improve your life. Plan complex projects and strategies Dave Parrack / Foundry Most jobs come with some level of project planning and management. Even I, as a freelance writer, need to plan tasks to get projects completed on time. And that’s where ChatGPT can prove invaluable, breaking projects up into smaller, more manageable parts. ChatGPT needs to know the nature of the project, the end goal, any constraints you may have, and what you have done so far. With that information, it can then break the project up with a step-by-step plan, and break it down further into phases. If ChatGPT doesn’t initially split your project up in a way that suits you, try again. Change up the prompts and make the AI chatbot tune in to exactly what you’re looking for. It takes a bit of back and forth, but it can shorten your planning time from hours to mere minutes. Compile research notes Dave Parrack / Foundry If you need to research a given topic of interest, ChatGPT can save you the hassle of compiling that research. For example, ahead of a trip to Croatia, I wanted to know more about the Croatian War of Independence, so I asked ChatGPT to provide me with a brief summary of the conflict with bullet points to help me understand how it happened. After absorbing all that information, I asked ChatGPT to add a timeline of the major events, further helping me to understand how the conflict played out. ChatGPT then offered to provide me with battle maps and/or summaries, plus profiles of the main players. You can go even deeper with ChatGPT’s Deep Research feature, which is now available to free users, up to 5 Deep Research tasks per month. With Deep Research, ChatGPT conducts multi-step research to generate comprehensive reportsbased on large amounts of information across the internet. A Deep Research task can take up to 30 minutes to complete, but it’ll save you hours or even days. Summarize articles, meetings, and more Dave Parrack / Foundry There are only so many hours in the day, yet so many new articles published on the web day in and day out. When you come across extra-long reads, it can be helpful to run them through ChatGPT for a quick summary. Then, if the summary is lacking in any way, you can go back and plow through the article proper. As an example, I ran one of my own PCWorld articlesthrough ChatGPT, which provided a brief summary of my points and broke down the best X alternative based on my reasons given. Interestingly, it also pulled elements from other articles.If you don’t want that, you can tell ChatGPT to limit its summary to the contents of the link. This is a great trick to use for other long-form, text-heavy content that you just don’t have the time to crunch through. Think transcripts for interviews, lectures, videos, and Zoom meetings. The only caveat is to never share private details with ChatGPT, like company-specific data that’s protected by NDAs and the like. Create Q&A flashcards for learning Dave Parrack / Foundry Flashcards can be extremely useful for drilling a lot of information into your brain, such as when studying for an exam, onboarding in a new role, prepping for an interview, etc. And with ChatGPT, you no longer have to painstakingly create those flashcards yourself. All you have to do is tell the AI the details of what you’re studying. You can specify the format, as well as various other elements. You can also choose to keep things broad or target specific sub-topics or concepts you want to focus on. You can even upload your own notes for ChatGPT to reference. You can also use Google’s NotebookLM app in a similar way. Provide interview practice Dave Parrack / Foundry Whether you’re a first-time jobseeker or have plenty of experience under your belt, it’s always a good idea to practice for your interviews when making career moves. Years ago, you might’ve had to ask a friend or family member to act as your mock interviewer. These days, ChatGPT can do it for you—and do it more effectively. Inform ChatGPT of the job title, industry, and level of position you’re interviewing for, what kind of interview it’ll be, and anything else you want it to take into consideration. ChatGPT will then conduct a mock interview with you, providing feedback along the way. When I tried this out myself, I was shocked by how capable ChatGPT can be at pretending to be a human in this context. And the feedback it provides for each answer you give is invaluable for knocking off your rough edges and improving your chances of success when you’re interviewed by a real hiring manager. Further reading: Non-gimmicky AI apps I actually use every day #menial #tasks #chatgpt #can #handle
    WWW.PCWORLD.COM
    9 menial tasks ChatGPT can handle in seconds, saving you hours
    ChatGPT is rapidly changing the world. The process is already happening, and it’s only going to accelerate as the technology improves, as more people gain access to it, and as more learn how to use it. What’s shocking is just how many tasks ChatGPT is already capable of managing for you. While the naysayers may still look down their noses at the potential of AI assistants, I’ve been using it to handle all kinds of menial tasks for me. Here are my favorite examples. Further reading: This tiny ChatGPT feature helps me tackle my days more productively Write your emails for you Dave Parrack / Foundry We’ve all been faced with the tricky task of writing an email—whether personal or professional—but not knowing quite how to word it. ChatGPT can do the heavy lifting for you, penning the (hopefully) perfect email based on whatever information you feed it. Let’s assume the email you need to write is of a professional nature, and wording it poorly could negatively affect your career. By directing ChatGPT to write the email with a particular structure, content, and tone of voice, you can give yourself a huge head start. A winning tip for this is to never accept ChatGPT’s first attempt. Always read through it and look for areas of improvement, then request tweaks to ensure you get the best possible email. You can (and should) also rewrite the email in your own voice. Learn more about how ChatGPT coached my colleague to write better emails. Generate itineraries and schedules Dave Parrack / Foundry If you’re going on a trip but you’re the type of person who hates planning trips, then you should utilize ChatGPT’s ability to generate trip itineraries. The results can be customized to the nth degree depending on how much detail and instruction you’re willing to provide. As someone who likes to get away at least once a year but also wants to make the most of every trip, leaning on ChatGPT for an itinerary is essential for me. I’ll provide the location and the kinds of things I want to see and do, then let it handle the rest. Instead of spending days researching everything myself, ChatGPT does 80 percent of it for me. As with all of these tasks, you don’t need to accept ChatGPT’s first effort. Use different prompts to force the AI chatbot to shape the itinerary closer to what you want. You’d be surprised at how many cool ideas you’ll encounter this way—simply nix the ones you don’t like. Break down difficult concepts Dave Parrack / Foundry One of the best tasks to assign to ChatGPT is the explanation of difficult concepts. Ask ChatGPT to explain any concept you can think of and it will deliver more often than not. You can tailor the level of explanation you need, and even have it include visual elements. Let’s say, for example, that a higher-up at work regularly lectures everyone about the importance of networking. But maybe they never go into detail about what they mean, just constantly pushing the why without explaining the what. Well, just ask ChatGPT to explain networking! Okay, most of us know what “networking” is and the concept isn’t very hard to grasp. But you can do this with anything. Ask ChatGPT to explain augmented reality, multi-threaded processing, blockchain, large language models, what have you. It will provide you with a clear and simple breakdown, maybe even with analogies and images. Analyze and make tough decisions Dave Parrack / Foundry We all face tough decisions every so often. The next time you find yourself wrestling with a particularly tough one—and you just can’t decide one way or the other—try asking ChatGPT for guidance and advice. It may sound strange to trust any kind of decision to artificial intelligence, let alone an important one that has you stumped, but doing so actually makes a lot of sense. While human judgment can be clouded by emotions, AI can set that aside and prioritize logic. It should go without saying: you don’t have to accept ChatGPT’s answers. Use the AI to weigh the pros and cons, to help you understand what’s most important to you, and to suggest a direction. Who knows? If you find yourself not liking the answer given, that in itself might clarify what you actually want—and the right answer for you. This is the kind of stuff ChatGPT can do to improve your life. Plan complex projects and strategies Dave Parrack / Foundry Most jobs come with some level of project planning and management. Even I, as a freelance writer, need to plan tasks to get projects completed on time. And that’s where ChatGPT can prove invaluable, breaking projects up into smaller, more manageable parts. ChatGPT needs to know the nature of the project, the end goal, any constraints you may have, and what you have done so far. With that information, it can then break the project up with a step-by-step plan, and break it down further into phases (if required). If ChatGPT doesn’t initially split your project up in a way that suits you, try again. Change up the prompts and make the AI chatbot tune in to exactly what you’re looking for. It takes a bit of back and forth, but it can shorten your planning time from hours to mere minutes. Compile research notes Dave Parrack / Foundry If you need to research a given topic of interest, ChatGPT can save you the hassle of compiling that research. For example, ahead of a trip to Croatia, I wanted to know more about the Croatian War of Independence, so I asked ChatGPT to provide me with a brief summary of the conflict with bullet points to help me understand how it happened. After absorbing all that information, I asked ChatGPT to add a timeline of the major events, further helping me to understand how the conflict played out. ChatGPT then offered to provide me with battle maps and/or summaries, plus profiles of the main players. You can go even deeper with ChatGPT’s Deep Research feature, which is now available to free users, up to 5 Deep Research tasks per month. With Deep Research, ChatGPT conducts multi-step research to generate comprehensive reports (with citations!) based on large amounts of information across the internet. A Deep Research task can take up to 30 minutes to complete, but it’ll save you hours or even days. Summarize articles, meetings, and more Dave Parrack / Foundry There are only so many hours in the day, yet so many new articles published on the web day in and day out. When you come across extra-long reads, it can be helpful to run them through ChatGPT for a quick summary. Then, if the summary is lacking in any way, you can go back and plow through the article proper. As an example, I ran one of my own PCWorld articles (where I compared Bluesky and Threads as alternatives to X) through ChatGPT, which provided a brief summary of my points and broke down the best X alternative based on my reasons given. Interestingly, it also pulled elements from other articles. (Hmph.) If you don’t want that, you can tell ChatGPT to limit its summary to the contents of the link. This is a great trick to use for other long-form, text-heavy content that you just don’t have the time to crunch through. Think transcripts for interviews, lectures, videos, and Zoom meetings. The only caveat is to never share private details with ChatGPT, like company-specific data that’s protected by NDAs and the like. Create Q&A flashcards for learning Dave Parrack / Foundry Flashcards can be extremely useful for drilling a lot of information into your brain, such as when studying for an exam, onboarding in a new role, prepping for an interview, etc. And with ChatGPT, you no longer have to painstakingly create those flashcards yourself. All you have to do is tell the AI the details of what you’re studying. You can specify the format (such as Q&A or multiple choice), as well as various other elements. You can also choose to keep things broad or target specific sub-topics or concepts you want to focus on. You can even upload your own notes for ChatGPT to reference. You can also use Google’s NotebookLM app in a similar way. Provide interview practice Dave Parrack / Foundry Whether you’re a first-time jobseeker or have plenty of experience under your belt, it’s always a good idea to practice for your interviews when making career moves. Years ago, you might’ve had to ask a friend or family member to act as your mock interviewer. These days, ChatGPT can do it for you—and do it more effectively. Inform ChatGPT of the job title, industry, and level of position you’re interviewing for, what kind of interview it’ll be (e.g., screener, technical assessment, group/panel, one-on-one with CEO), and anything else you want it to take into consideration. ChatGPT will then conduct a mock interview with you, providing feedback along the way. When I tried this out myself, I was shocked by how capable ChatGPT can be at pretending to be a human in this context. And the feedback it provides for each answer you give is invaluable for knocking off your rough edges and improving your chances of success when you’re interviewed by a real hiring manager. Further reading: Non-gimmicky AI apps I actually use every day
    0 Commentarii 0 Distribuiri
  • Mock up a website in five prompts

    “Wait, can users actually add products to the cart?”Every prototype faces that question or one like it. You start to explain it’s “just Figma,” “just dummy data,” but what if you didn’t need disclaimers?What if you could hand clients—or your team—a working, data-connected mock-up of their website, or new pages and components, in less time than it takes to wireframe?That’s the challenge we’ll tackle today. But first, we need to look at:The problem with today’s prototyping toolsPick two: speed, flexibility, or interactivity.The prototyping ecosystem, despite having amazing software that addresses a huge variety of needs, doesn’t really have one tool that gives you all three.Wireframing apps let you draw boxes in minutes but every button is fake. Drag-and-drop builders animate scroll triggers until you ask for anything off-template. Custom code frees you… after you wave goodbye to a few afternoons.AI tools haven’t smashed the trade-off; they’ve just dressed it in flashier costumes. One prompt births a landing page, the next dumps a 2,000-line, worse-than-junior-level React file in your lap. The bottleneck is still there. Builder’s approach to website mockupsWe’ve been trying something a little different to maintain speed, flexibility, and interactivity while mocking full websites. Our AI-driven visual editor:Spins up a repo in seconds or connects to your existing one to use the code as design inspiration. React, Vue, Angular, and Svelte all work out of the box.
    Lets you shape components via plain English, visual edits, copy/pasted Figma frames, web inspos, MCP tools, and constant visual awareness of your entire website.
    Commits each change as a clean GitHub pull request your team can review like hand-written code. All your usual CI checks and lint rules apply.And if you need a tweak, you can comment to @builderio-bot right in the GitHub PR to make asynchronous changes without context switching.This results in a live site the café owner can interact with today, and a branch your devs can merge tomorrow. Stakeholders get to click actual buttons and trigger real state—no more “so, just imagine this works” demos.Let’s see it in action.From blank canvas to working mockup in five promptsToday, I’m going to mock up a fake business website. You’re welcome to create a real one.Before we fire off a single prompt, grab a note and write:Business name & vibe
    Core pages
    Primary goal
    Brand palette & toneThat’s it. Don’t sweat the details—we can always iterate. For mine, I wrote:1. Sunny Trails Bakery — family-owned, feel-good, smells like warm cinnamon.
    2. Home, About, Pricing / Subscription Box, Menu.
    3. Drive online orders and foot traffic—every CTA should funnel toward “Order Now” or “Reserve a Table.”
    4. Warm yellow, chocolate brown, rounded typography, playful copy.We’re not trying to fit everything here. What matters is clarity on what we’re creating, so the AI has enough context to produce usable scaffolds, and so later tweaks stay aligned with the client’s vision. Builder will default to using React, Vite, and Tailwind. If you want a different JS framework, you can link an existing repo in that stack. In the near future, you won’t need to do this extra step to get non-React frameworks to function.An entire website from the first promptNow, we’re ready to get going.Head over to Builder.io and paste in this prompt or your own:Create a cozy bakery website called “Sunny Trails Bakery” with pages for:
    • Home
    • About
    • Pricing
    • Menu
    Brand palette: warm yellow and chocolate brown. Tone: playful, inviting. The restaurant is family-owned, feel-good, and smells like cinnamon.
    The goal of this site is to drive online orders and foot traffic—every CTA should funnel toward "Order Now" or "Reserve a Table."Once you hit enter, Builder will spin up a new dev container, and then inside that container, the AI will build out the first version of your site. You can leave the page and come back when it’s done.Now, before we go further, let’s create our repo, so that we get version history right from the outset. Click “Create Repo” up in the top right, and link your GitHub account.Once the process is complete, you’ll have a brand new repo.If you need any help on this step, or any of the below, check out these docs.Making the mockup’s order system workFrom our one-shot prompt, we’ve already got a really nice start for our client. However, when we press the “Order Now” button, we just get a generic alert. Let’s fix this.The best part about connecting to GitHub is that we get version control. Head back to your dashboard and edit the settings of your new project. We can give it a better name, and then, in the “Advanced” section, we can change the “Commit Mode” to “Pull Requests.”Now, we have the ability to create new branches right within Builder, allowing us to make drastic changes without worrying about the main version. This is also helpful if you’d like to show your client or team a few different versions of the same prototype.On a new branch, I’ll write another short prompt:Can you make the "Order Now" button work, even if it's just with dummy JSON for now?As you can see in the GIF above, Builder creates an ordering system and a fully mobile-responsive cart and checkout flow.Now, we can click “Send PR” in the top right, and we have an ordinary GitHub PR that can be reviewed and merged as needed.This is what’s possible in two prompts. For our third, let’s gussy up the style.If you’re like me, you might spend a lot of time admiring other people’s cool designs and learning how to code up similar components in your own style.Luckily, Builder has this capability, too, with our Chrome extension. I found a “Featured Posts” section on OpenAI’s website, where I like how the layout and scrolling work. We can copy and paste it onto our “Featured Treats” section, retaining our cafe’s distinctive brand style.Don’t worry—OpenAI doesn’t mind a little web scraping.You can do this with any component on any website, so your own projects can very quickly become a “best of the web” if you know what you’re doing.Plus, you can use Figma designs in much the same way, with even better design fidelity. Copy and paste a Figma frame with our Figma plugin, and tell the AI to either use the component as inspiration or as a 1:1 to reference for what the design should be.Now, we’re ready to send our PR. This time, let’s take a closer look at the code the AI has created.As you can see, the code is neatly formatted into two reusable components. Scrolling down further, I find a CSS file and then the actual implementation on the homepage, with clean JSON to represent the dummy post data.Design tweaks to the mockup with visual editsOne issue that cropped up when the AI brought in the OpenAI layout is that it changed my text from “Featured Treats” to “Featured Stories & Treats.” I’ve realized I don’t like either, and I want to replace that text with: “Fresh Out of the Bakery.”It would be silly, though, to prompt the AI just for this small tweak. Let’s switch into edit mode.Edit Mode lets you select any component and change any of its content or underlying CSS directly. You get a host of Webflow-like options to choose from, so that you can finesse the details as needed.Once you’ve made all the visual changes you want—maybe tweaking a button color or a border radius—you can click “Apply Edits,” and the AI will ensure the underlying code matches your repo’s style.Async fixes to the mockup with Builder BotNow, our pull request is nearly ready to merge, but I found one issue with it:When we copied the OpenAI website layout earlier, one of the blog posts had a video as its featured graphic instead of just an image. This is cool for OpenAI, but for our bakery, I just wanted images in this section. Since I didn’t instruct Builder’s AI otherwise, it went ahead and followed the layout and created extra code for video capability.No problem. We can fix this inside GItHub with our final prompt. We just need to comment on the PR and tag builderio-bot. Within about a minute, Builder Bot has successfully removed the video functionality, leaving a minimal diff that affects only the code it needed to. For example: Returning to my project in Builder, I can see that the bot’s changes are accounted for in the chat window as well, and I can use the live preview link to make sure my site works as expected:Now, if this were a real project, you could easily deploy this to the web for your client. After all, you’ve got a whole GitHub repo. This isn’t just a mockup; it’s actual code you can tweak—with Builder or Cursor or by hand—until you’re satisfied to run the site in production.So, why use Builder to mock up your website?Sure, this has been a somewhat contrived example. A real prototype is going to look prettier, because I’m going to spend more time on pieces of the design that I don’t like as much.But that’s the point of the best AI tools: they don’t take you, the human, out of the loop.You still get to make all the executive decisions, and it respects your hard work. Since you can constantly see all the code the AI creates, work in branches, and prompt with component-level precision, you can stop worrying about AI overwriting your opinions and start using it more as the tool it’s designed to be.You can copy in your team’s Figma designs, import web inspos, connect MCP servers to get Jira tickets in hand, and—most importantly—work with existing repos full of existing styles that Builder will understand and match, just like it matched OpenAI’s layout to our little cafe.So, we get speed, flexibility, and interactivity all the way from prompt to PR to production.Try Builder today.
    #mock #website #five #prompts
    Mock up a website in five prompts
    “Wait, can users actually add products to the cart?”Every prototype faces that question or one like it. You start to explain it’s “just Figma,” “just dummy data,” but what if you didn’t need disclaimers?What if you could hand clients—or your team—a working, data-connected mock-up of their website, or new pages and components, in less time than it takes to wireframe?That’s the challenge we’ll tackle today. But first, we need to look at:The problem with today’s prototyping toolsPick two: speed, flexibility, or interactivity.The prototyping ecosystem, despite having amazing software that addresses a huge variety of needs, doesn’t really have one tool that gives you all three.Wireframing apps let you draw boxes in minutes but every button is fake. Drag-and-drop builders animate scroll triggers until you ask for anything off-template. Custom code frees you… after you wave goodbye to a few afternoons.AI tools haven’t smashed the trade-off; they’ve just dressed it in flashier costumes. One prompt births a landing page, the next dumps a 2,000-line, worse-than-junior-level React file in your lap. The bottleneck is still there. Builder’s approach to website mockupsWe’ve been trying something a little different to maintain speed, flexibility, and interactivity while mocking full websites. Our AI-driven visual editor:Spins up a repo in seconds or connects to your existing one to use the code as design inspiration. React, Vue, Angular, and Svelte all work out of the box. Lets you shape components via plain English, visual edits, copy/pasted Figma frames, web inspos, MCP tools, and constant visual awareness of your entire website. Commits each change as a clean GitHub pull request your team can review like hand-written code. All your usual CI checks and lint rules apply.And if you need a tweak, you can comment to @builderio-bot right in the GitHub PR to make asynchronous changes without context switching.This results in a live site the café owner can interact with today, and a branch your devs can merge tomorrow. Stakeholders get to click actual buttons and trigger real state—no more “so, just imagine this works” demos.Let’s see it in action.From blank canvas to working mockup in five promptsToday, I’m going to mock up a fake business website. You’re welcome to create a real one.Before we fire off a single prompt, grab a note and write:Business name & vibe Core pages Primary goal Brand palette & toneThat’s it. Don’t sweat the details—we can always iterate. For mine, I wrote:1. Sunny Trails Bakery — family-owned, feel-good, smells like warm cinnamon. 2. Home, About, Pricing / Subscription Box, Menu. 3. Drive online orders and foot traffic—every CTA should funnel toward “Order Now” or “Reserve a Table.” 4. Warm yellow, chocolate brown, rounded typography, playful copy.We’re not trying to fit everything here. What matters is clarity on what we’re creating, so the AI has enough context to produce usable scaffolds, and so later tweaks stay aligned with the client’s vision. Builder will default to using React, Vite, and Tailwind. If you want a different JS framework, you can link an existing repo in that stack. In the near future, you won’t need to do this extra step to get non-React frameworks to function.An entire website from the first promptNow, we’re ready to get going.Head over to Builder.io and paste in this prompt or your own:Create a cozy bakery website called “Sunny Trails Bakery” with pages for: • Home • About • Pricing • Menu Brand palette: warm yellow and chocolate brown. Tone: playful, inviting. The restaurant is family-owned, feel-good, and smells like cinnamon. The goal of this site is to drive online orders and foot traffic—every CTA should funnel toward "Order Now" or "Reserve a Table."Once you hit enter, Builder will spin up a new dev container, and then inside that container, the AI will build out the first version of your site. You can leave the page and come back when it’s done.Now, before we go further, let’s create our repo, so that we get version history right from the outset. Click “Create Repo” up in the top right, and link your GitHub account.Once the process is complete, you’ll have a brand new repo.If you need any help on this step, or any of the below, check out these docs.Making the mockup’s order system workFrom our one-shot prompt, we’ve already got a really nice start for our client. However, when we press the “Order Now” button, we just get a generic alert. Let’s fix this.The best part about connecting to GitHub is that we get version control. Head back to your dashboard and edit the settings of your new project. We can give it a better name, and then, in the “Advanced” section, we can change the “Commit Mode” to “Pull Requests.”Now, we have the ability to create new branches right within Builder, allowing us to make drastic changes without worrying about the main version. This is also helpful if you’d like to show your client or team a few different versions of the same prototype.On a new branch, I’ll write another short prompt:Can you make the "Order Now" button work, even if it's just with dummy JSON for now?As you can see in the GIF above, Builder creates an ordering system and a fully mobile-responsive cart and checkout flow.Now, we can click “Send PR” in the top right, and we have an ordinary GitHub PR that can be reviewed and merged as needed.This is what’s possible in two prompts. For our third, let’s gussy up the style.If you’re like me, you might spend a lot of time admiring other people’s cool designs and learning how to code up similar components in your own style.Luckily, Builder has this capability, too, with our Chrome extension. I found a “Featured Posts” section on OpenAI’s website, where I like how the layout and scrolling work. We can copy and paste it onto our “Featured Treats” section, retaining our cafe’s distinctive brand style.Don’t worry—OpenAI doesn’t mind a little web scraping.You can do this with any component on any website, so your own projects can very quickly become a “best of the web” if you know what you’re doing.Plus, you can use Figma designs in much the same way, with even better design fidelity. Copy and paste a Figma frame with our Figma plugin, and tell the AI to either use the component as inspiration or as a 1:1 to reference for what the design should be.Now, we’re ready to send our PR. This time, let’s take a closer look at the code the AI has created.As you can see, the code is neatly formatted into two reusable components. Scrolling down further, I find a CSS file and then the actual implementation on the homepage, with clean JSON to represent the dummy post data.Design tweaks to the mockup with visual editsOne issue that cropped up when the AI brought in the OpenAI layout is that it changed my text from “Featured Treats” to “Featured Stories & Treats.” I’ve realized I don’t like either, and I want to replace that text with: “Fresh Out of the Bakery.”It would be silly, though, to prompt the AI just for this small tweak. Let’s switch into edit mode.Edit Mode lets you select any component and change any of its content or underlying CSS directly. You get a host of Webflow-like options to choose from, so that you can finesse the details as needed.Once you’ve made all the visual changes you want—maybe tweaking a button color or a border radius—you can click “Apply Edits,” and the AI will ensure the underlying code matches your repo’s style.Async fixes to the mockup with Builder BotNow, our pull request is nearly ready to merge, but I found one issue with it:When we copied the OpenAI website layout earlier, one of the blog posts had a video as its featured graphic instead of just an image. This is cool for OpenAI, but for our bakery, I just wanted images in this section. Since I didn’t instruct Builder’s AI otherwise, it went ahead and followed the layout and created extra code for video capability.No problem. We can fix this inside GItHub with our final prompt. We just need to comment on the PR and tag builderio-bot. Within about a minute, Builder Bot has successfully removed the video functionality, leaving a minimal diff that affects only the code it needed to. For example: Returning to my project in Builder, I can see that the bot’s changes are accounted for in the chat window as well, and I can use the live preview link to make sure my site works as expected:Now, if this were a real project, you could easily deploy this to the web for your client. After all, you’ve got a whole GitHub repo. This isn’t just a mockup; it’s actual code you can tweak—with Builder or Cursor or by hand—until you’re satisfied to run the site in production.So, why use Builder to mock up your website?Sure, this has been a somewhat contrived example. A real prototype is going to look prettier, because I’m going to spend more time on pieces of the design that I don’t like as much.But that’s the point of the best AI tools: they don’t take you, the human, out of the loop.You still get to make all the executive decisions, and it respects your hard work. Since you can constantly see all the code the AI creates, work in branches, and prompt with component-level precision, you can stop worrying about AI overwriting your opinions and start using it more as the tool it’s designed to be.You can copy in your team’s Figma designs, import web inspos, connect MCP servers to get Jira tickets in hand, and—most importantly—work with existing repos full of existing styles that Builder will understand and match, just like it matched OpenAI’s layout to our little cafe.So, we get speed, flexibility, and interactivity all the way from prompt to PR to production.Try Builder today. #mock #website #five #prompts
    WWW.BUILDER.IO
    Mock up a website in five prompts
    “Wait, can users actually add products to the cart?”Every prototype faces that question or one like it. You start to explain it’s “just Figma,” “just dummy data,” but what if you didn’t need disclaimers?What if you could hand clients—or your team—a working, data-connected mock-up of their website, or new pages and components, in less time than it takes to wireframe?That’s the challenge we’ll tackle today. But first, we need to look at:The problem with today’s prototyping toolsPick two: speed, flexibility, or interactivity.The prototyping ecosystem, despite having amazing software that addresses a huge variety of needs, doesn’t really have one tool that gives you all three.Wireframing apps let you draw boxes in minutes but every button is fake. Drag-and-drop builders animate scroll triggers until you ask for anything off-template. Custom code frees you… after you wave goodbye to a few afternoons.AI tools haven’t smashed the trade-off; they’ve just dressed it in flashier costumes. One prompt births a landing page, the next dumps a 2,000-line, worse-than-junior-level React file in your lap. The bottleneck is still there. Builder’s approach to website mockupsWe’ve been trying something a little different to maintain speed, flexibility, and interactivity while mocking full websites. Our AI-driven visual editor:Spins up a repo in seconds or connects to your existing one to use the code as design inspiration. React, Vue, Angular, and Svelte all work out of the box. Lets you shape components via plain English, visual edits, copy/pasted Figma frames, web inspos, MCP tools, and constant visual awareness of your entire website. Commits each change as a clean GitHub pull request your team can review like hand-written code. All your usual CI checks and lint rules apply.And if you need a tweak, you can comment to @builderio-bot right in the GitHub PR to make asynchronous changes without context switching.This results in a live site the café owner can interact with today, and a branch your devs can merge tomorrow. Stakeholders get to click actual buttons and trigger real state—no more “so, just imagine this works” demos.Let’s see it in action.From blank canvas to working mockup in five promptsToday, I’m going to mock up a fake business website. You’re welcome to create a real one.Before we fire off a single prompt, grab a note and write:Business name & vibe Core pages Primary goal Brand palette & toneThat’s it. Don’t sweat the details—we can always iterate. For mine, I wrote:1. Sunny Trails Bakery — family-owned, feel-good, smells like warm cinnamon. 2. Home, About, Pricing / Subscription Box, Menu (with daily specials). 3. Drive online orders and foot traffic—every CTA should funnel toward “Order Now” or “Reserve a Table.” 4. Warm yellow, chocolate brown, rounded typography, playful copy.We’re not trying to fit everything here. What matters is clarity on what we’re creating, so the AI has enough context to produce usable scaffolds, and so later tweaks stay aligned with the client’s vision. Builder will default to using React, Vite, and Tailwind. If you want a different JS framework, you can link an existing repo in that stack. In the near future, you won’t need to do this extra step to get non-React frameworks to function.(Free tier Builder gives you 5 AI credits/day and 25/month—plenty to follow along with today’s demo. Upgrade only when you need it.)An entire website from the first promptNow, we’re ready to get going.Head over to Builder.io and paste in this prompt or your own:Create a cozy bakery website called “Sunny Trails Bakery” with pages for: • Home • About • Pricing • Menu Brand palette: warm yellow and chocolate brown. Tone: playful, inviting. The restaurant is family-owned, feel-good, and smells like cinnamon. The goal of this site is to drive online orders and foot traffic—every CTA should funnel toward "Order Now" or "Reserve a Table."Once you hit enter, Builder will spin up a new dev container, and then inside that container, the AI will build out the first version of your site. You can leave the page and come back when it’s done.Now, before we go further, let’s create our repo, so that we get version history right from the outset. Click “Create Repo” up in the top right, and link your GitHub account.Once the process is complete, you’ll have a brand new repo.If you need any help on this step, or any of the below, check out these docs.Making the mockup’s order system workFrom our one-shot prompt, we’ve already got a really nice start for our client. However, when we press the “Order Now” button, we just get a generic alert. Let’s fix this.The best part about connecting to GitHub is that we get version control. Head back to your dashboard and edit the settings of your new project. We can give it a better name, and then, in the “Advanced” section, we can change the “Commit Mode” to “Pull Requests.”Now, we have the ability to create new branches right within Builder, allowing us to make drastic changes without worrying about the main version. This is also helpful if you’d like to show your client or team a few different versions of the same prototype.On a new branch, I’ll write another short prompt:Can you make the "Order Now" button work, even if it's just with dummy JSON for now?As you can see in the GIF above, Builder creates an ordering system and a fully mobile-responsive cart and checkout flow.Now, we can click “Send PR” in the top right, and we have an ordinary GitHub PR that can be reviewed and merged as needed.This is what’s possible in two prompts. For our third, let’s gussy up the style.If you’re like me, you might spend a lot of time admiring other people’s cool designs and learning how to code up similar components in your own style.Luckily, Builder has this capability, too, with our Chrome extension. I found a “Featured Posts” section on OpenAI’s website, where I like how the layout and scrolling work. We can copy and paste it onto our “Featured Treats” section, retaining our cafe’s distinctive brand style.Don’t worry—OpenAI doesn’t mind a little web scraping.You can do this with any component on any website, so your own projects can very quickly become a “best of the web” if you know what you’re doing.Plus, you can use Figma designs in much the same way, with even better design fidelity. Copy and paste a Figma frame with our Figma plugin, and tell the AI to either use the component as inspiration or as a 1:1 to reference for what the design should be.(You can grab our design-to-code guide for a lot more ideas of what this can help you accomplish.)Now, we’re ready to send our PR. This time, let’s take a closer look at the code the AI has created.As you can see, the code is neatly formatted into two reusable components. Scrolling down further, I find a CSS file and then the actual implementation on the homepage, with clean JSON to represent the dummy post data.Design tweaks to the mockup with visual editsOne issue that cropped up when the AI brought in the OpenAI layout is that it changed my text from “Featured Treats” to “Featured Stories & Treats.” I’ve realized I don’t like either, and I want to replace that text with: “Fresh Out of the Bakery.”It would be silly, though, to prompt the AI just for this small tweak. Let’s switch into edit mode.Edit Mode lets you select any component and change any of its content or underlying CSS directly. You get a host of Webflow-like options to choose from, so that you can finesse the details as needed.Once you’ve made all the visual changes you want—maybe tweaking a button color or a border radius—you can click “Apply Edits,” and the AI will ensure the underlying code matches your repo’s style.Async fixes to the mockup with Builder BotNow, our pull request is nearly ready to merge, but I found one issue with it:When we copied the OpenAI website layout earlier, one of the blog posts had a video as its featured graphic instead of just an image. This is cool for OpenAI, but for our bakery, I just wanted images in this section. Since I didn’t instruct Builder’s AI otherwise, it went ahead and followed the layout and created extra code for video capability.No problem. We can fix this inside GItHub with our final prompt. We just need to comment on the PR and tag builderio-bot. Within about a minute, Builder Bot has successfully removed the video functionality, leaving a minimal diff that affects only the code it needed to. For example: Returning to my project in Builder, I can see that the bot’s changes are accounted for in the chat window as well, and I can use the live preview link to make sure my site works as expected:Now, if this were a real project, you could easily deploy this to the web for your client. After all, you’ve got a whole GitHub repo. This isn’t just a mockup; it’s actual code you can tweak—with Builder or Cursor or by hand—until you’re satisfied to run the site in production.So, why use Builder to mock up your website?Sure, this has been a somewhat contrived example. A real prototype is going to look prettier, because I’m going to spend more time on pieces of the design that I don’t like as much.But that’s the point of the best AI tools: they don’t take you, the human, out of the loop.You still get to make all the executive decisions, and it respects your hard work. Since you can constantly see all the code the AI creates, work in branches, and prompt with component-level precision, you can stop worrying about AI overwriting your opinions and start using it more as the tool it’s designed to be.You can copy in your team’s Figma designs, import web inspos, connect MCP servers to get Jira tickets in hand, and—most importantly—work with existing repos full of existing styles that Builder will understand and match, just like it matched OpenAI’s layout to our little cafe.So, we get speed, flexibility, and interactivity all the way from prompt to PR to production.Try Builder today.
    0 Commentarii 0 Distribuiri
  • How AI is reshaping the future of healthcare and medical research

    Transcript       
    PETER LEE: “In ‘The Little Black Bag,’ a classic science fiction story, a high-tech doctor’s kit of the future is accidentally transported back to the 1950s, into the shaky hands of a washed-up, alcoholic doctor. The ultimate medical tool, it redeems the doctor wielding it, allowing him to practice gratifyingly heroic medicine. … The tale ends badly for the doctor and his treacherous assistant, but it offered a picture of how advanced technology could transform medicine—powerful when it was written nearly 75 years ago and still so today. What would be the Al equivalent of that little black bag? At this moment when new capabilities are emerging, how do we imagine them into medicine?”          
    This is The AI Revolution in Medicine, Revisited. I’m your host, Peter Lee.   
    Shortly after OpenAI’s GPT-4 was publicly released, Carey Goldberg, Dr. Zak Kohane, and I published The AI Revolution in Medicine to help educate the world of healthcare and medical research about the transformative impact this new generative AI technology could have. But because we wrote the book when GPT-4 was still a secret, we had to speculate. Now, two years later, what did we get right, and what did we get wrong?    
    In this series, we’ll talk to clinicians, patients, hospital administrators, and others to understand the reality of AI in the field and where we go from here.  The book passage I read at the top is from “Chapter 10: The Big Black Bag.” 
    In imagining AI in medicine, Carey, Zak, and I included in our book two fictional accounts. In the first, a medical resident consults GPT-4 on her personal phone as the patient in front of her crashes. Within seconds, it offers an alternate response based on recent literature. In the second account, a 90-year-old woman with several chronic conditions is living independently and receiving near-constant medical support from an AI aide.   
    In our conversations with the guests we’ve spoken to so far, we’ve caught a glimpse of these predicted futures, seeing how clinicians and patients are actually using AI today and how developers are leveraging the technology in the healthcare products and services they’re creating. In fact, that first fictional account isn’t so fictional after all, as most of the doctors in the real world actually appear to be using AI at least occasionally—and sometimes much more than occasionally—to help in their daily clinical work. And as for the second fictional account, which is more of a science fiction account, it seems we are indeed on the verge of a new way of delivering and receiving healthcare, though the future is still very much open. 
    As we continue to examine the current state of AI in healthcare and its potential to transform the field, I’m pleased to welcome Bill Gates and Sébastien Bubeck.  
    Bill may be best known as the co-founder of Microsoft, having created the company with his childhood friend Paul Allen in 1975. He’s now the founder of Breakthrough Energy, which aims to advance clean energy innovation, and TerraPower, a company developing groundbreaking nuclear energy and science technologies. He also chairs the world’s largest philanthropic organization, the Gates Foundation, and focuses on solving a variety of health challenges around the globe and here at home. 
    Sébastien is a research lead at OpenAI. He was previously a distinguished scientist, vice president of AI, and a colleague of mine here at Microsoft, where his work included spearheading the development of the family of small language models known as Phi. While at Microsoft, he also coauthored the discussion-provoking 2023 paper “Sparks of Artificial General Intelligence,” which presented the results of early experiments with GPT-4 conducted by a small team from Microsoft Research.     
    Here’s my conversation with Bill Gates and Sébastien Bubeck. 
    LEE: Bill, welcome. 
    BILL GATES: Thank you. 
    LEE: Seb … 
    SÉBASTIEN BUBECK: Yeah. Hi, hi, Peter. Nice to be here. 
    LEE: You know, one of the things that I’ve been doing just to get the conversation warmed up is to talk about origin stories, and what I mean about origin stories is, you know, what was the first contact that you had with large language models or the concept of generative AI that convinced you or made you think that something really important was happening? 
    And so, Bill, I think I’ve heard the story about, you know, the time when the OpenAI folks—Sam Altman, Greg Brockman, and others—showed you something, but could we hear from you what those early encounters were like and what was going through your mind?  
    GATES: Well, I’d been visiting OpenAI soon after it was created to see things like GPT-2 and to see the little arm they had that was trying to match human manipulation and, you know, looking at their games like Dota that they were trying to get as good as human play. And honestly, I didn’t think the language model stuff they were doing, even when they got to GPT-3, would show the ability to learn, you know, in the same sense that a human reads a biology book and is able to take that knowledge and access it not only to pass a test but also to create new medicines. 
    And so my challenge to them was that if their LLM could get a five on the advanced placement biology test, then I would say, OK, it took biologic knowledge and encoded it in an accessible way and that I didn’t expect them to do that very quickly but it would be profound.  
    And it was only about six months after I challenged them to do that, that an early version of GPT-4 they brought up to a dinner at my house, and in fact, it answered most of the questions that night very well. The one it got totally wrong, we were … because it was so good, we kept thinking, Oh, we must be wrong. It turned out it was a math weaknessthat, you know, we later understood that that was an area of, weirdly, of incredible weakness of those early models. But, you know, that was when I realized, OK, the age of cheap intelligence was at its beginning. 
    LEE: Yeah. So I guess it seems like you had something similar to me in that my first encounters, I actually harbored some skepticism. Is it fair to say you were skeptical before that? 
    GATES: Well, the idea that we’ve figured out how to encode and access knowledge in this very deep sense without even understanding the nature of the encoding, … 
    LEE: Right.  
    GATES: … that is a bit weird.  
    LEE: Yeah. 
    GATES: We have an algorithm that creates the computation, but even say, OK, where is the president’s birthday stored in there? Where is this fact stored in there? The fact that even now when we’re playing around, getting a little bit more sense of it, it’s opaque to us what the semantic encoding is, it’s, kind of, amazing to me. I thought the invention of knowledge storage would be an explicit way of encoding knowledge, not an implicit statistical training. 
    LEE: Yeah, yeah. All right. So, Seb, you know, on this same topic, you know, I got—as we say at Microsoft—I got pulled into the tent. 
    BUBECK: Yes.  
    LEE: Because this was a very secret project. And then, um, I had the opportunity to select a small number of researchers in MSRto join and start investigating this thing seriously. And the first person I pulled in was you. 
    BUBECK: Yeah. 
    LEE: And so what were your first encounters? Because I actually don’t remember what happened then. 
    BUBECK: Oh, I remember it very well.My first encounter with GPT-4 was in a meeting with the two of you, actually. But my kind of first contact, the first moment where I realized that something was happening with generative AI, was before that. And I agree with Bill that I also wasn’t too impressed by GPT-3. 
    I though that it was kind of, you know, very naturally mimicking the web, sort of parroting what was written there in a nice way. Still in a way which seemed very impressive. But it wasn’t really intelligent in any way. But shortly after GPT-3, there was a model before GPT-4 that really shocked me, and this was the first image generation model, DALL-E 1. 
    So that was in 2021. And I will forever remember the press release of OpenAI where they had this prompt of an avocado chair and then you had this image of the avocado chair.And what really shocked me is that clearly the model kind of “understood” what is a chair, what is an avocado, and was able to merge those concepts. 
    So this was really, to me, the first moment where I saw some understanding in those models.  
    LEE: So this was, just to get the timing right, that was before I pulled you into the tent. 
    BUBECK: That was before. That was like a year before. 
    LEE: Right.  
    BUBECK: And now I will tell you how, you know, we went from that moment to the meeting with the two of you and GPT-4. 
    So once I saw this kind of understanding, I thought, OK, fine. It understands concept, but it’s still not able to reason. It cannot—as, you know, Bill was saying—it cannot learn from your document. It cannot reason.  
    So I set out to try to prove that. You know, this is what I was in the business of at the time, trying to prove things in mathematics. So I was trying to prove that basically autoregressive transformers could never reason. So I was trying to prove this. And after a year of work, I had something reasonable to show. And so I had the meeting with the two of you, and I had this example where I wanted to say, there is no way that an LLM is going to be able to do x. 
    And then as soon as I … I don’t know if you remember, Bill. But as soon as I said that, you said, oh, but wait a second. I had, you know, the OpenAI crew at my house recently, and they showed me a new model. Why don’t we ask this new model this question?  
    LEE: Yeah.
    BUBECK: And we did, and it solved it on the spot. And that really, honestly, just changed my life. Like, you know, I had been working for a year trying to say that this was impossible. And just right there, it was shown to be possible.  
    LEE:One of the very first things I got interested in—because I was really thinking a lot about healthcare—was healthcare and medicine. 
    And I don’t know if the two of you remember, but I ended up doing a lot of tests. I ran through, you know, step one and step two of the US Medical Licensing Exam. Did a whole bunch of other things. I wrote this big report. It was, you know, I can’t remember … a couple hundred pages.  
    And I needed to share this with someone. I didn’t … there weren’t too many people I could share it with. So I sent, I think, a copy to you, Bill. Sent a copy to you, Seb.  
    I hardly slept for about a week putting that report together. And, yeah, and I kept working on it. But I was far from alone. I think everyone who was in the tent, so to speak, in those early days was going through something pretty similar. All right. So I think … of course, a lot of what I put in the report also ended up being examples that made it into the book. 
    But the main purpose of this conversation isn’t to reminisce aboutor indulge in those reminiscences but to talk about what’s happening in healthcare and medicine. And, you know, as I said, we wrote this book. We did it very, very quickly. Seb, you helped. Bill, you know, you provided a review and some endorsements. 
    But, you know, honestly, we didn’t know what we were talking about because no one had access to this thing. And so we just made a bunch of guesses. So really, the whole thing I wanted to probe with the two of you is, now with two years of experience out in the world, what, you know, what do we think is happening today? 
    You know, is AI actually having an impact, positive or negative, on healthcare and medicine? And what do we now think is going to happen in the next two years, five years, or 10 years? And so I realize it’s a little bit too abstract to just ask it that way. So let me just try to narrow the discussion and guide us a little bit.  
    Um, the kind of administrative and clerical work, paperwork, around healthcare—and we made a lot of guesses about that—that appears to be going well, but, you know, Bill, I know we’ve discussed that sometimes that you think there ought to be a lot more going on. Do you have a viewpoint on how AI is actually finding its way into reducing paperwork? 
    GATES: Well, I’m stunned … I don’t think there should be a patient-doctor meeting where the AI is not sitting in and both transcribing, offering to help with the paperwork, and even making suggestions, although the doctor will be the one, you know, who makes the final decision about the diagnosis and whatever prescription gets done.  
    It’s so helpful. You know, when that patient goes home and their, you know, son who wants to understand what happened has some questions, that AI should be available to continue that conversation. And the way you can improve that experience and streamline things and, you know, involve the people who advise you. I don’t understand why that’s not more adopted, because there you still have the human in the loop making that final decision. 
    But even for, like, follow-up calls to make sure the patient did things, to understand if they have concerns and knowing when to escalate back to the doctor, the benefit is incredible. And, you know, that thing is ready for prime time. That paradigm is ready for prime time, in my view. 
    LEE: Yeah, there are some good products, but it seems like the number one use right now—and we kind of got this from some of the previous guests in previous episodes—is the use of AI just to respond to emails from patients.Does that make sense to you? 
    BUBECK: Yeah. So maybe I want to second what Bill was saying but maybe take a step back first. You know, two years ago, like, the concept of clinical scribes, which is one of the things that we’re talking about right now, it would have sounded, in fact, it sounded two years ago, borderline dangerous. Because everybody was worried about hallucinations. What happened if you have this AI listening in and then it transcribes, you know, something wrong? 
    Now, two years later, I think it’s mostly working. And in fact, it is not yet, you know, fully adopted. You’re right. But it is in production. It is used, you know, in many, many places. So this rate of progress is astounding because it wasn’t obvious that we would be able to overcome those obstacles of hallucination. It’s not to say that hallucinations are fully solved. In the case of the closed system, they are.  
    Now, I think more generally what’s going on in the background is that there is something that we, that certainly I, underestimated, which is this management overhead. So I think the reason why this is not adopted everywhere is really a training and teaching aspect. People need to be taught, like, those systems, how to interact with them. 
    And one example that I really like, a study that recently appeared where they tried to use ChatGPT for diagnosis and they were comparing doctors without and with ChatGPT. And the amazing thing … so this was a set of cases where the accuracy of the doctors alone was around 75%. ChatGPT alone was 90%. So that’s already kind of mind blowing. But then the kicker is that doctors with ChatGPT was 80%.  
    Intelligence alone is not enough. It’s also how it’s presented, how you interact with it. And ChatGPT, it’s an amazing tool. Obviously, I absolutely love it. But it’s not … you don’t want a doctor to have to type in, you know, prompts and use it that way. 
    It should be, as Bill was saying, kind of running continuously in the background, sending you notifications. And you have to be really careful of the rate at which those notifications are being sent. Because if they are too frequent, then the doctor will learn to ignore them. So you have to … all of those things matter, in fact, at least as much as the level of intelligence of the machine. 
    LEE: One of the things I think about, Bill, in that scenario that you described, doctors do some thinking about the patient when they write the note. So, you know, I’m always a little uncertain whether it’s actually … you know, you wouldn’t necessarily want to fully automate this, I don’t think. Or at least there needs to be some prompt to the doctor to make sure that the doctor puts some thought into what happened in the encounter with the patient. Does that make sense to you at all? 
    GATES: At this stage, you know, I’d still put the onus on the doctor to write the conclusions and the summary and not delegate that. 
    The tradeoffs you make a little bit are somewhat dependent on the situation you’re in. If you’re in Africa,
    So, yes, the doctor’s still going to have to do a lot of work, but just the quality of letting the patient and the people around them interact and ask questions and have things explained, that alone is such a quality improvement. It’s mind blowing.  
    LEE: So since you mentioned, you know, Africa—and, of course, this touches on the mission and some of the priorities of the Gates Foundation and this idea of democratization of access to expert medical care—what’s the most interesting stuff going on right now? Are there people and organizations or technologies that are impressing you or that you’re tracking? 
    GATES: Yeah. So the Gates Foundation has given out a lot of grants to people in Africa doing education, agriculture but more healthcare examples than anything. And the way these things start off, they often start out either being patient-centric in a narrow situation, like, OK, I’m a pregnant woman; talk to me. Or, I have infectious disease symptoms; talk to me. Or they’re connected to a health worker where they’re helping that worker get their job done. And we have lots of pilots out, you know, in both of those cases.  
    The dream would be eventually to have the thing the patient consults be so broad that it’s like having a doctor available who understands the local things.  
    LEE: Right.  
    GATES: We’re not there yet. But over the next two or three years, you know, particularly given the worsening financial constraints against African health systems, where the withdrawal of money has been dramatic, you know, figuring out how to take this—what I sometimes call “free intelligence”—and build a quality health system around that, we will have to be more radical in low-income countries than any rich country is ever going to be.  
    LEE: Also, there’s maybe a different regulatory environment, so some of those things maybe are easier? Because right now, I think the world hasn’t figured out how to and whether to regulate, let’s say, an AI that might give a medical diagnosis or write a prescription for a medication. 
    BUBECK: Yeah. I think one issue with this, and it’s also slowing down the deployment of AI in healthcare more generally, is a lack of proper benchmark. Because, you know, you were mentioning the USMLE, for example. That’s a great test to test human beings and their knowledge of healthcare and medicine. But it’s not a great test to give to an AI. 
    It’s not asking the right questions. So finding what are the right questions to test whether an AI system is ready to give diagnosis in a constrained setting, that’s a very, very important direction, which to my surprise, is not yet accelerating at the rate that I was hoping for. 
    LEE: OK, so that gives me an excuse to get more now into the core AI tech because something I’ve discussed with both of you is this issue of what are the right tests. And you both know the very first test I give to any new spin of an LLM is I present a patient, the results—a mythical patient—the results of my physical exam, my mythical physical exam. Maybe some results of some initial labs. And then I present or propose a differential diagnosis. And if you’re not in medicine, a differential diagnosis you can just think of as a prioritized list of the possible diagnoses that fit with all that data. And in that proposed differential, I always intentionally make two mistakes. 
    I make a textbook technical error in one of the possible elements of the differential diagnosis, and I have an error of omission. And, you know, I just want to know, does the LLM understand what I’m talking about? And all the good ones out there do now. But then I want to know, can it spot the errors? And then most importantly, is it willing to tell me I’m wrong, that I’ve made a mistake?  
    That last piece seems really hard for AI today. And so let me ask you first, Seb, because at the time of this taping, of course, there was a new spin of GPT-4o last week that became overly sycophantic. In other words, it was actually prone in that test of mine not only to not tell me I’m wrong, but it actually praised me for the creativity of my differential.What’s up with that? 
    BUBECK: Yeah, I guess it’s a testament to the fact that training those models is still more of an art than a science. So it’s a difficult job. Just to be clear with the audience, we have rolled back thatversion of GPT-4o, so now we don’t have the sycophant version out there. 
    Yeah, no, it’s a really difficult question. It has to do … as you said, it’s very technical. It has to do with the post-training and how, like, where do you nudge the model? So, you know, there is this very classical by now technique called RLHF, where you push the model in the direction of a certain reward model. So the reward model is just telling the model, you know, what behavior is good, what behavior is bad. 
    But this reward model is itself an LLM, and, you know, Bill was saying at the very beginning of the conversation that we don’t really understand how those LLMs deal with concepts like, you know, where is the capital of France located? Things like that. It is the same thing for this reward model. We don’t know why it says that it prefers one output to another, and whether this is correlated with some sycophancy is, you know, something that we discovered basically just now. That if you push too hard in optimization on this reward model, you will get a sycophant model. 
    So it’s kind of … what I’m trying to say is we became too good at what we were doing, and we ended up, in fact, in a trap of the reward model. 
    LEE: I mean, you do want … it’s a difficult balance because you do want models to follow your desires and … 
    BUBECK: It’s a very difficult, very difficult balance. 
    LEE: So this brings up then the following question for me, which is the extent to which we think we’ll need to have specially trained models for things. So let me start with you, Bill. Do you have a point of view on whether we will need to, you know, quote-unquote take AI models to med school? Have them specially trained? Like, if you were going to deploy something to give medical care in underserved parts of the world, do we need to do something special to create those models? 
    GATES: We certainly need to teach them the African languages and the unique dialects so that the multimedia interactions are very high quality. We certainly need to teach them the disease prevalence and unique disease patterns like, you know, neglected tropical diseases and malaria. So we need to gather a set of facts that somebody trying to go for a US customer base, you know, wouldn’t necessarily have that in there. 
    Those two things are actually very straightforward because the additional training time is small. I’d say for the next few years, we’ll also need to do reinforcement learning about the context of being a doctor and how important certain behaviors are. Humans learn over the course of their life to some degree that, I’m in a different context and the way I behave in terms of being willing to criticize or be nice, you know, how important is it? Who’s here? What’s my relationship to them?  
    Right now, these machines don’t have that broad social experience. And so if you know it’s going to be used for health things, a lot of reinforcement learning of the very best humans in that context would still be valuable. Eventually, the models will, having read all the literature of the world about good doctors, bad doctors, it’ll understand as soon as you say, “I want you to be a doctor diagnosing somebody.” All of the implicit reinforcement that fits that situation, you know, will be there.
    LEE: Yeah.
    GATES: And so I hope three years from now, we don’t have to do that reinforcement learning. But today, for any medical context, you would want a lot of data to reinforce tone, willingness to say things when, you know, there might be something significant at stake. 
    LEE: Yeah. So, you know, something Bill said, kind of, reminds me of another thing that I think we missed, which is, the context also … and the specialization also pertains to different, I guess, what we still call “modes,” although I don’t know if the idea of multimodal is the same as it was two years ago. But, you know, what do you make of all of the hubbub around—in fact, within Microsoft Research, this is a big deal, but I think we’re far from alone—you know, medical images and vision, video, proteins and molecules, cell, you know, cellular data and so on. 
    BUBECK: Yeah. OK. So there is a lot to say to everything … to the last, you know, couple of minutes. Maybe on the specialization aspect, you know, I think there is, hiding behind this, a really fundamental scientific question of whether eventually we have a singular AGIthat kind of knows everything and you can just put, you know, explain your own context and it will just get it and understand everything. 
    That’s one vision. I have to say, I don’t particularly believe in this vision. In fact, we humans are not like that at all. I think, hopefully, we are general intelligences, yet we have to specialize a lot. And, you know, I did myself a lot of RL, reinforcement learning, on mathematics. Like, that’s what I did, you know, spent a lot of time doing that. And I didn’t improve on other aspects. You know, in fact, I probably degraded in other aspects.So it’s … I think it’s an important example to have in mind. 
    LEE: I think I might disagree with you on that, though, because, like, doesn’t a model have to see both good science and bad science in order to be able to gain the ability to discern between the two? 
    BUBECK: Yeah, no, that absolutely. I think there is value in seeing the generality, in having a very broad base. But then you, kind of, specialize on verticals. And this is where also, you know, open-weights model, which we haven’t talked about yet, are really important because they allow you to provide this broad base to everyone. And then you can specialize on top of it. 
    LEE: So we have about three hours of stuff to talk about, but our time is actually running low.
    BUBECK: Yes, yes, yes.  
    LEE: So I think I want … there’s a more provocative question. It’s almost a silly question, but I need to ask it of the two of you, which is, is there a future, you know, where AI replaces doctors or replaces, you know, medical specialties that we have today? So what does the world look like, say, five years from now? 
    GATES: Well, it’s important to distinguish healthcare discovery activity from healthcare delivery activity. We focused mostly on delivery. I think it’s very much within the realm of possibility that the AI is not only accelerating healthcare discovery but substituting for a lot of the roles of, you know, I’m an organic chemist, or I run various types of assays. I can see those, which are, you know, testable-output-type jobs but with still very high value, I can see, you know, some replacement in those areas before the doctor.  
    The doctor, still understanding the human condition and long-term dialogues, you know, they’ve had a lifetime of reinforcement of that, particularly when you get into areas like mental health. So I wouldn’t say in five years, either people will choose to adopt it, but it will be profound that there’ll be this nearly free intelligence that can do follow-up, that can help you, you know, make sure you went through different possibilities. 
    And so I’d say, yes, we’ll have doctors, but I’d say healthcare will be massively transformed in its quality and in efficiency by AI in that time period. 
    LEE: Is there a comparison, useful comparison, say, between doctors and, say, programmers, computer programmers, or doctors and, I don’t know, lawyers? 
    GATES: Programming is another one that has, kind of, a mathematical correctness to it, you know, and so the objective function that you’re trying to reinforce to, as soon as you can understand the state machines, you can have something that’s “checkable”; that’s correct. So I think programming, you know, which is weird to say, that the machine will beat us at most programming tasks before we let it take over roles that have deep empathy, you know, physical presence and social understanding in them. 
    LEE: Yeah. By the way, you know, I fully expect in five years that AI will produce mathematical proofs that are checkable for validity, easily checkable, because they’ll be written in a proof-checking language like Lean or something but will be so complex that no human mathematician can understand them. I expect that to happen.  
    I can imagine in some fields, like cellular biology, we could have the same situation in the future because the molecular pathways, the chemistry, biochemistry of human cells or living cells is as complex as any mathematics, and so it seems possible that we may be in a state where in wet lab, we see, Oh yeah, this actually works, but no one can understand why. 
    BUBECK: Yeah, absolutely. I mean, I think I really agree with Bill’s distinction of the discovery and the delivery, and indeed, the discovery’s when you can check things, and at the end, there is an artifact that you can verify. You know, you can run the protocol in the wet lab and seeproduced what you wanted. So I absolutely agree with that.  
    And in fact, you know, we don’t have to talk five years from now. I don’t know if you know, but just recently, there was a paper that was published on a scientific discovery using o3- mini. So this is really amazing. And, you know, just very quickly, just so people know, it was about this statistical physics model, the frustrated Potts model, which has to do with coloring, and basically, the case of three colors, like, more than two colors was open for a long time, and o3 was able to reduce the case of three colors to two colors.  
    LEE: Yeah. 
    BUBECK: Which is just, like, astounding. And this is not … this is now. This is happening right now. So this is something that I personally didn’t expect it would happen so quickly, and it’s due to those reasoning models.  
    Now, on the delivery side, I would add something more to it for the reason why doctors and, in fact, lawyers and coders will remain for a long time, and it’s because we still don’t understand how those models generalize. Like, at the end of the day, we are not able to tell you when they are confronted with a really new, novel situation, whether they will work or not. 
    Nobody is able to give you that guarantee. And I think until we understand this generalization better, we’re not going to be willing to just let the system in the wild without human supervision. 
    LEE: But don’t human doctors, human specialists … so, for example, a cardiologist sees a patient in a certain way that a nephrologist … 
    BUBECK: Yeah.
    LEE: … or an endocrinologist might not.
    BUBECK: That’s right. But another cardiologist will understand and, kind of, expect a certain level of generalization from their peer. And this, we just don’t have it with AI models. Now, of course, you’re exactly right. That generalization is also hard for humans. Like, if you have a human trained for one task and you put them into another task, then you don’t … you often don’t know.
    LEE: OK. You know, the podcast is focused on what’s happened over the last two years. But now, I’d like one provocative prediction about what you think the world of AI and medicine is going to be at some point in the future. You pick your timeframe. I don’t care if it’s two years or 20 years from now, but, you know, what do you think will be different about AI in medicine in that future than today? 
    BUBECK: Yeah, I think the deployment is going to accelerate soon. Like, we’re really not missing very much. There is this enormous capability overhang. Like, even if progress completely stopped, with current systems, we can do a lot more than what we’re doing right now. So I think this will … this has to be realized, you know, sooner rather than later. 
    And I think it’s probably dependent on these benchmarks and proper evaluation and tying this with regulation. So these are things that take time in human society and for good reason. But now we already are at two years; you know, give it another two years and it should be really …  
    LEE: Will AI prescribe your medicines? Write your prescriptions? 
    BUBECK: I think yes. I think yes. 
    LEE: OK. Bill? 
    GATES: Well, I think the next two years, we’ll have massive pilots, and so the amount of use of the AI, still in a copilot-type mode, you know, we should get millions of patient visits, you know, both in general medicine and in the mental health side, as well. And I think that’s going to build up both the data and the confidence to give the AI some additional autonomy. You know, are you going to let it talk to you at night when you’re panicked about your mental health with some ability to escalate?
    And, you know, I’ve gone so far as to tell politicians with national health systems that if they deploy AI appropriately, that the quality of care, the overload of the doctors, the improvement in the economics will be enough that their voters will be stunned because they just don’t expect this, and, you know, they could be reelectedjust on this one thing of fixing what is a very overloaded and economically challenged health system in these rich countries. 
    You know, my personal role is going to be to make sure that in the poorer countries, there isn’t some lag; in fact, in many cases, that we’ll be more aggressive because, you know, we’re comparing to having no access to doctors at all. And, you know, so I think whether it’s India or Africa, there’ll be lessons that are globally valuable because we need medical intelligence. And, you know, thank god AI is going to provide a lot of that. 
    LEE: Well, on that optimistic note, I think that’s a good way to end. Bill, Seb, really appreciate all of this.  
    I think the most fundamental prediction we made in the book is that AI would actually find its way into the practice of medicine, and I think that that at least has come true, maybe in different ways than we expected, but it’s come true, and I think it’ll only accelerate from here. So thanks again, both of you.  
    GATES: Yeah. Thanks, you guys. 
    BUBECK: Thank you, Peter. Thanks, Bill. 
    LEE: I just always feel such a sense of privilege to have a chance to interact and actually work with people like Bill and Sébastien.   
    With Bill, I’m always amazed at how practically minded he is. He’s really thinking about the nuts and bolts of what AI might be able to do for people, and his thoughts about underserved parts of the world, the idea that we might actually be able to empower people with access to expert medical knowledge, I think is both inspiring and amazing.  
    And then, Seb, Sébastien Bubeck, he’s just absolutely a brilliant mind. He has a really firm grip on the deep mathematics of artificial intelligence and brings that to bear in his research and development work. And where that mathematics takes him isn’t just into the nuts and bolts of algorithms but into philosophical questions about the nature of intelligence.  
    One of the things that Sébastien brought up was the state of evaluation of AI systems. And indeed, he was fairly critical in our conversation. But of course, the world of AI research and development is just moving so fast, and indeed, since we recorded our conversation, OpenAI, in fact, released a new evaluation metric that is directly relevant to medical applications, and that is something called HealthBench. And Microsoft Research also released a new evaluation approach or process called ADeLe.  
    HealthBench and ADeLe are examples of new approaches to evaluating AI models that are less about testing their knowledge and ability to pass multiple-choice exams and instead are evaluation approaches designed to assess how well AI models are able to complete tasks that actually arise every day in typical healthcare or biomedical research settings. These are examples of really important good work that speak to how well AI models work in the real world of healthcare and biomedical research and how well they can collaborate with human beings in those settings. 
    You know, I asked Bill and Seb to make some predictions about the future. You know, my own answer, I expect that we’re going to be able to use AI to change how we diagnose patients, change how we decide treatment options.  
    If you’re a doctor or a nurse and you encounter a patient, you’ll ask questions, do a physical exam, you know, call out for labs just like you do today, but then you’ll be able to engage with AI based on all of that data and just ask, you know, based on all the other people who have gone through the same experience, who have similar data, how were they diagnosed? How were they treated? What were their outcomes? And what does that mean for the patient I have right now? Some people call it the “patients like me” paradigm. And I think that’s going to become real because of AI within our lifetimes. That idea of really grounding the delivery in healthcare and medical practice through data and intelligence, I actually now don’t see any barriers to that future becoming real.  
    I’d like to extend another big thank you to Bill and Sébastien for their time. And to our listeners, as always, it’s a pleasure to have you along for the ride. I hope you’ll join us for our remaining conversations, as well as a second coauthor roundtable with Carey and Zak.  
    Until next time.  
    #how #reshaping #future #healthcare #medical
    How AI is reshaping the future of healthcare and medical research
    Transcript        PETER LEE: “In ‘The Little Black Bag,’ a classic science fiction story, a high-tech doctor’s kit of the future is accidentally transported back to the 1950s, into the shaky hands of a washed-up, alcoholic doctor. The ultimate medical tool, it redeems the doctor wielding it, allowing him to practice gratifyingly heroic medicine. … The tale ends badly for the doctor and his treacherous assistant, but it offered a picture of how advanced technology could transform medicine—powerful when it was written nearly 75 years ago and still so today. What would be the Al equivalent of that little black bag? At this moment when new capabilities are emerging, how do we imagine them into medicine?”           This is The AI Revolution in Medicine, Revisited. I’m your host, Peter Lee.    Shortly after OpenAI’s GPT-4 was publicly released, Carey Goldberg, Dr. Zak Kohane, and I published The AI Revolution in Medicine to help educate the world of healthcare and medical research about the transformative impact this new generative AI technology could have. But because we wrote the book when GPT-4 was still a secret, we had to speculate. Now, two years later, what did we get right, and what did we get wrong?     In this series, we’ll talk to clinicians, patients, hospital administrators, and others to understand the reality of AI in the field and where we go from here.  The book passage I read at the top is from “Chapter 10: The Big Black Bag.”  In imagining AI in medicine, Carey, Zak, and I included in our book two fictional accounts. In the first, a medical resident consults GPT-4 on her personal phone as the patient in front of her crashes. Within seconds, it offers an alternate response based on recent literature. In the second account, a 90-year-old woman with several chronic conditions is living independently and receiving near-constant medical support from an AI aide.    In our conversations with the guests we’ve spoken to so far, we’ve caught a glimpse of these predicted futures, seeing how clinicians and patients are actually using AI today and how developers are leveraging the technology in the healthcare products and services they’re creating. In fact, that first fictional account isn’t so fictional after all, as most of the doctors in the real world actually appear to be using AI at least occasionally—and sometimes much more than occasionally—to help in their daily clinical work. And as for the second fictional account, which is more of a science fiction account, it seems we are indeed on the verge of a new way of delivering and receiving healthcare, though the future is still very much open.  As we continue to examine the current state of AI in healthcare and its potential to transform the field, I’m pleased to welcome Bill Gates and Sébastien Bubeck.   Bill may be best known as the co-founder of Microsoft, having created the company with his childhood friend Paul Allen in 1975. He’s now the founder of Breakthrough Energy, which aims to advance clean energy innovation, and TerraPower, a company developing groundbreaking nuclear energy and science technologies. He also chairs the world’s largest philanthropic organization, the Gates Foundation, and focuses on solving a variety of health challenges around the globe and here at home.  Sébastien is a research lead at OpenAI. He was previously a distinguished scientist, vice president of AI, and a colleague of mine here at Microsoft, where his work included spearheading the development of the family of small language models known as Phi. While at Microsoft, he also coauthored the discussion-provoking 2023 paper “Sparks of Artificial General Intelligence,” which presented the results of early experiments with GPT-4 conducted by a small team from Microsoft Research.      Here’s my conversation with Bill Gates and Sébastien Bubeck.  LEE: Bill, welcome.  BILL GATES: Thank you.  LEE: Seb …  SÉBASTIEN BUBECK: Yeah. Hi, hi, Peter. Nice to be here.  LEE: You know, one of the things that I’ve been doing just to get the conversation warmed up is to talk about origin stories, and what I mean about origin stories is, you know, what was the first contact that you had with large language models or the concept of generative AI that convinced you or made you think that something really important was happening?  And so, Bill, I think I’ve heard the story about, you know, the time when the OpenAI folks—Sam Altman, Greg Brockman, and others—showed you something, but could we hear from you what those early encounters were like and what was going through your mind?   GATES: Well, I’d been visiting OpenAI soon after it was created to see things like GPT-2 and to see the little arm they had that was trying to match human manipulation and, you know, looking at their games like Dota that they were trying to get as good as human play. And honestly, I didn’t think the language model stuff they were doing, even when they got to GPT-3, would show the ability to learn, you know, in the same sense that a human reads a biology book and is able to take that knowledge and access it not only to pass a test but also to create new medicines.  And so my challenge to them was that if their LLM could get a five on the advanced placement biology test, then I would say, OK, it took biologic knowledge and encoded it in an accessible way and that I didn’t expect them to do that very quickly but it would be profound.   And it was only about six months after I challenged them to do that, that an early version of GPT-4 they brought up to a dinner at my house, and in fact, it answered most of the questions that night very well. The one it got totally wrong, we were … because it was so good, we kept thinking, Oh, we must be wrong. It turned out it was a math weaknessthat, you know, we later understood that that was an area of, weirdly, of incredible weakness of those early models. But, you know, that was when I realized, OK, the age of cheap intelligence was at its beginning.  LEE: Yeah. So I guess it seems like you had something similar to me in that my first encounters, I actually harbored some skepticism. Is it fair to say you were skeptical before that?  GATES: Well, the idea that we’ve figured out how to encode and access knowledge in this very deep sense without even understanding the nature of the encoding, …  LEE: Right.   GATES: … that is a bit weird.   LEE: Yeah.  GATES: We have an algorithm that creates the computation, but even say, OK, where is the president’s birthday stored in there? Where is this fact stored in there? The fact that even now when we’re playing around, getting a little bit more sense of it, it’s opaque to us what the semantic encoding is, it’s, kind of, amazing to me. I thought the invention of knowledge storage would be an explicit way of encoding knowledge, not an implicit statistical training.  LEE: Yeah, yeah. All right. So, Seb, you know, on this same topic, you know, I got—as we say at Microsoft—I got pulled into the tent.  BUBECK: Yes.   LEE: Because this was a very secret project. And then, um, I had the opportunity to select a small number of researchers in MSRto join and start investigating this thing seriously. And the first person I pulled in was you.  BUBECK: Yeah.  LEE: And so what were your first encounters? Because I actually don’t remember what happened then.  BUBECK: Oh, I remember it very well.My first encounter with GPT-4 was in a meeting with the two of you, actually. But my kind of first contact, the first moment where I realized that something was happening with generative AI, was before that. And I agree with Bill that I also wasn’t too impressed by GPT-3.  I though that it was kind of, you know, very naturally mimicking the web, sort of parroting what was written there in a nice way. Still in a way which seemed very impressive. But it wasn’t really intelligent in any way. But shortly after GPT-3, there was a model before GPT-4 that really shocked me, and this was the first image generation model, DALL-E 1.  So that was in 2021. And I will forever remember the press release of OpenAI where they had this prompt of an avocado chair and then you had this image of the avocado chair.And what really shocked me is that clearly the model kind of “understood” what is a chair, what is an avocado, and was able to merge those concepts.  So this was really, to me, the first moment where I saw some understanding in those models.   LEE: So this was, just to get the timing right, that was before I pulled you into the tent.  BUBECK: That was before. That was like a year before.  LEE: Right.   BUBECK: And now I will tell you how, you know, we went from that moment to the meeting with the two of you and GPT-4.  So once I saw this kind of understanding, I thought, OK, fine. It understands concept, but it’s still not able to reason. It cannot—as, you know, Bill was saying—it cannot learn from your document. It cannot reason.   So I set out to try to prove that. You know, this is what I was in the business of at the time, trying to prove things in mathematics. So I was trying to prove that basically autoregressive transformers could never reason. So I was trying to prove this. And after a year of work, I had something reasonable to show. And so I had the meeting with the two of you, and I had this example where I wanted to say, there is no way that an LLM is going to be able to do x.  And then as soon as I … I don’t know if you remember, Bill. But as soon as I said that, you said, oh, but wait a second. I had, you know, the OpenAI crew at my house recently, and they showed me a new model. Why don’t we ask this new model this question?   LEE: Yeah. BUBECK: And we did, and it solved it on the spot. And that really, honestly, just changed my life. Like, you know, I had been working for a year trying to say that this was impossible. And just right there, it was shown to be possible.   LEE:One of the very first things I got interested in—because I was really thinking a lot about healthcare—was healthcare and medicine.  And I don’t know if the two of you remember, but I ended up doing a lot of tests. I ran through, you know, step one and step two of the US Medical Licensing Exam. Did a whole bunch of other things. I wrote this big report. It was, you know, I can’t remember … a couple hundred pages.   And I needed to share this with someone. I didn’t … there weren’t too many people I could share it with. So I sent, I think, a copy to you, Bill. Sent a copy to you, Seb.   I hardly slept for about a week putting that report together. And, yeah, and I kept working on it. But I was far from alone. I think everyone who was in the tent, so to speak, in those early days was going through something pretty similar. All right. So I think … of course, a lot of what I put in the report also ended up being examples that made it into the book.  But the main purpose of this conversation isn’t to reminisce aboutor indulge in those reminiscences but to talk about what’s happening in healthcare and medicine. And, you know, as I said, we wrote this book. We did it very, very quickly. Seb, you helped. Bill, you know, you provided a review and some endorsements.  But, you know, honestly, we didn’t know what we were talking about because no one had access to this thing. And so we just made a bunch of guesses. So really, the whole thing I wanted to probe with the two of you is, now with two years of experience out in the world, what, you know, what do we think is happening today?  You know, is AI actually having an impact, positive or negative, on healthcare and medicine? And what do we now think is going to happen in the next two years, five years, or 10 years? And so I realize it’s a little bit too abstract to just ask it that way. So let me just try to narrow the discussion and guide us a little bit.   Um, the kind of administrative and clerical work, paperwork, around healthcare—and we made a lot of guesses about that—that appears to be going well, but, you know, Bill, I know we’ve discussed that sometimes that you think there ought to be a lot more going on. Do you have a viewpoint on how AI is actually finding its way into reducing paperwork?  GATES: Well, I’m stunned … I don’t think there should be a patient-doctor meeting where the AI is not sitting in and both transcribing, offering to help with the paperwork, and even making suggestions, although the doctor will be the one, you know, who makes the final decision about the diagnosis and whatever prescription gets done.   It’s so helpful. You know, when that patient goes home and their, you know, son who wants to understand what happened has some questions, that AI should be available to continue that conversation. And the way you can improve that experience and streamline things and, you know, involve the people who advise you. I don’t understand why that’s not more adopted, because there you still have the human in the loop making that final decision.  But even for, like, follow-up calls to make sure the patient did things, to understand if they have concerns and knowing when to escalate back to the doctor, the benefit is incredible. And, you know, that thing is ready for prime time. That paradigm is ready for prime time, in my view.  LEE: Yeah, there are some good products, but it seems like the number one use right now—and we kind of got this from some of the previous guests in previous episodes—is the use of AI just to respond to emails from patients.Does that make sense to you?  BUBECK: Yeah. So maybe I want to second what Bill was saying but maybe take a step back first. You know, two years ago, like, the concept of clinical scribes, which is one of the things that we’re talking about right now, it would have sounded, in fact, it sounded two years ago, borderline dangerous. Because everybody was worried about hallucinations. What happened if you have this AI listening in and then it transcribes, you know, something wrong?  Now, two years later, I think it’s mostly working. And in fact, it is not yet, you know, fully adopted. You’re right. But it is in production. It is used, you know, in many, many places. So this rate of progress is astounding because it wasn’t obvious that we would be able to overcome those obstacles of hallucination. It’s not to say that hallucinations are fully solved. In the case of the closed system, they are.   Now, I think more generally what’s going on in the background is that there is something that we, that certainly I, underestimated, which is this management overhead. So I think the reason why this is not adopted everywhere is really a training and teaching aspect. People need to be taught, like, those systems, how to interact with them.  And one example that I really like, a study that recently appeared where they tried to use ChatGPT for diagnosis and they were comparing doctors without and with ChatGPT. And the amazing thing … so this was a set of cases where the accuracy of the doctors alone was around 75%. ChatGPT alone was 90%. So that’s already kind of mind blowing. But then the kicker is that doctors with ChatGPT was 80%.   Intelligence alone is not enough. It’s also how it’s presented, how you interact with it. And ChatGPT, it’s an amazing tool. Obviously, I absolutely love it. But it’s not … you don’t want a doctor to have to type in, you know, prompts and use it that way.  It should be, as Bill was saying, kind of running continuously in the background, sending you notifications. And you have to be really careful of the rate at which those notifications are being sent. Because if they are too frequent, then the doctor will learn to ignore them. So you have to … all of those things matter, in fact, at least as much as the level of intelligence of the machine.  LEE: One of the things I think about, Bill, in that scenario that you described, doctors do some thinking about the patient when they write the note. So, you know, I’m always a little uncertain whether it’s actually … you know, you wouldn’t necessarily want to fully automate this, I don’t think. Or at least there needs to be some prompt to the doctor to make sure that the doctor puts some thought into what happened in the encounter with the patient. Does that make sense to you at all?  GATES: At this stage, you know, I’d still put the onus on the doctor to write the conclusions and the summary and not delegate that.  The tradeoffs you make a little bit are somewhat dependent on the situation you’re in. If you’re in Africa, So, yes, the doctor’s still going to have to do a lot of work, but just the quality of letting the patient and the people around them interact and ask questions and have things explained, that alone is such a quality improvement. It’s mind blowing.   LEE: So since you mentioned, you know, Africa—and, of course, this touches on the mission and some of the priorities of the Gates Foundation and this idea of democratization of access to expert medical care—what’s the most interesting stuff going on right now? Are there people and organizations or technologies that are impressing you or that you’re tracking?  GATES: Yeah. So the Gates Foundation has given out a lot of grants to people in Africa doing education, agriculture but more healthcare examples than anything. And the way these things start off, they often start out either being patient-centric in a narrow situation, like, OK, I’m a pregnant woman; talk to me. Or, I have infectious disease symptoms; talk to me. Or they’re connected to a health worker where they’re helping that worker get their job done. And we have lots of pilots out, you know, in both of those cases.   The dream would be eventually to have the thing the patient consults be so broad that it’s like having a doctor available who understands the local things.   LEE: Right.   GATES: We’re not there yet. But over the next two or three years, you know, particularly given the worsening financial constraints against African health systems, where the withdrawal of money has been dramatic, you know, figuring out how to take this—what I sometimes call “free intelligence”—and build a quality health system around that, we will have to be more radical in low-income countries than any rich country is ever going to be.   LEE: Also, there’s maybe a different regulatory environment, so some of those things maybe are easier? Because right now, I think the world hasn’t figured out how to and whether to regulate, let’s say, an AI that might give a medical diagnosis or write a prescription for a medication.  BUBECK: Yeah. I think one issue with this, and it’s also slowing down the deployment of AI in healthcare more generally, is a lack of proper benchmark. Because, you know, you were mentioning the USMLE, for example. That’s a great test to test human beings and their knowledge of healthcare and medicine. But it’s not a great test to give to an AI.  It’s not asking the right questions. So finding what are the right questions to test whether an AI system is ready to give diagnosis in a constrained setting, that’s a very, very important direction, which to my surprise, is not yet accelerating at the rate that I was hoping for.  LEE: OK, so that gives me an excuse to get more now into the core AI tech because something I’ve discussed with both of you is this issue of what are the right tests. And you both know the very first test I give to any new spin of an LLM is I present a patient, the results—a mythical patient—the results of my physical exam, my mythical physical exam. Maybe some results of some initial labs. And then I present or propose a differential diagnosis. And if you’re not in medicine, a differential diagnosis you can just think of as a prioritized list of the possible diagnoses that fit with all that data. And in that proposed differential, I always intentionally make two mistakes.  I make a textbook technical error in one of the possible elements of the differential diagnosis, and I have an error of omission. And, you know, I just want to know, does the LLM understand what I’m talking about? And all the good ones out there do now. But then I want to know, can it spot the errors? And then most importantly, is it willing to tell me I’m wrong, that I’ve made a mistake?   That last piece seems really hard for AI today. And so let me ask you first, Seb, because at the time of this taping, of course, there was a new spin of GPT-4o last week that became overly sycophantic. In other words, it was actually prone in that test of mine not only to not tell me I’m wrong, but it actually praised me for the creativity of my differential.What’s up with that?  BUBECK: Yeah, I guess it’s a testament to the fact that training those models is still more of an art than a science. So it’s a difficult job. Just to be clear with the audience, we have rolled back thatversion of GPT-4o, so now we don’t have the sycophant version out there.  Yeah, no, it’s a really difficult question. It has to do … as you said, it’s very technical. It has to do with the post-training and how, like, where do you nudge the model? So, you know, there is this very classical by now technique called RLHF, where you push the model in the direction of a certain reward model. So the reward model is just telling the model, you know, what behavior is good, what behavior is bad.  But this reward model is itself an LLM, and, you know, Bill was saying at the very beginning of the conversation that we don’t really understand how those LLMs deal with concepts like, you know, where is the capital of France located? Things like that. It is the same thing for this reward model. We don’t know why it says that it prefers one output to another, and whether this is correlated with some sycophancy is, you know, something that we discovered basically just now. That if you push too hard in optimization on this reward model, you will get a sycophant model.  So it’s kind of … what I’m trying to say is we became too good at what we were doing, and we ended up, in fact, in a trap of the reward model.  LEE: I mean, you do want … it’s a difficult balance because you do want models to follow your desires and …  BUBECK: It’s a very difficult, very difficult balance.  LEE: So this brings up then the following question for me, which is the extent to which we think we’ll need to have specially trained models for things. So let me start with you, Bill. Do you have a point of view on whether we will need to, you know, quote-unquote take AI models to med school? Have them specially trained? Like, if you were going to deploy something to give medical care in underserved parts of the world, do we need to do something special to create those models?  GATES: We certainly need to teach them the African languages and the unique dialects so that the multimedia interactions are very high quality. We certainly need to teach them the disease prevalence and unique disease patterns like, you know, neglected tropical diseases and malaria. So we need to gather a set of facts that somebody trying to go for a US customer base, you know, wouldn’t necessarily have that in there.  Those two things are actually very straightforward because the additional training time is small. I’d say for the next few years, we’ll also need to do reinforcement learning about the context of being a doctor and how important certain behaviors are. Humans learn over the course of their life to some degree that, I’m in a different context and the way I behave in terms of being willing to criticize or be nice, you know, how important is it? Who’s here? What’s my relationship to them?   Right now, these machines don’t have that broad social experience. And so if you know it’s going to be used for health things, a lot of reinforcement learning of the very best humans in that context would still be valuable. Eventually, the models will, having read all the literature of the world about good doctors, bad doctors, it’ll understand as soon as you say, “I want you to be a doctor diagnosing somebody.” All of the implicit reinforcement that fits that situation, you know, will be there. LEE: Yeah. GATES: And so I hope three years from now, we don’t have to do that reinforcement learning. But today, for any medical context, you would want a lot of data to reinforce tone, willingness to say things when, you know, there might be something significant at stake.  LEE: Yeah. So, you know, something Bill said, kind of, reminds me of another thing that I think we missed, which is, the context also … and the specialization also pertains to different, I guess, what we still call “modes,” although I don’t know if the idea of multimodal is the same as it was two years ago. But, you know, what do you make of all of the hubbub around—in fact, within Microsoft Research, this is a big deal, but I think we’re far from alone—you know, medical images and vision, video, proteins and molecules, cell, you know, cellular data and so on.  BUBECK: Yeah. OK. So there is a lot to say to everything … to the last, you know, couple of minutes. Maybe on the specialization aspect, you know, I think there is, hiding behind this, a really fundamental scientific question of whether eventually we have a singular AGIthat kind of knows everything and you can just put, you know, explain your own context and it will just get it and understand everything.  That’s one vision. I have to say, I don’t particularly believe in this vision. In fact, we humans are not like that at all. I think, hopefully, we are general intelligences, yet we have to specialize a lot. And, you know, I did myself a lot of RL, reinforcement learning, on mathematics. Like, that’s what I did, you know, spent a lot of time doing that. And I didn’t improve on other aspects. You know, in fact, I probably degraded in other aspects.So it’s … I think it’s an important example to have in mind.  LEE: I think I might disagree with you on that, though, because, like, doesn’t a model have to see both good science and bad science in order to be able to gain the ability to discern between the two?  BUBECK: Yeah, no, that absolutely. I think there is value in seeing the generality, in having a very broad base. But then you, kind of, specialize on verticals. And this is where also, you know, open-weights model, which we haven’t talked about yet, are really important because they allow you to provide this broad base to everyone. And then you can specialize on top of it.  LEE: So we have about three hours of stuff to talk about, but our time is actually running low. BUBECK: Yes, yes, yes.   LEE: So I think I want … there’s a more provocative question. It’s almost a silly question, but I need to ask it of the two of you, which is, is there a future, you know, where AI replaces doctors or replaces, you know, medical specialties that we have today? So what does the world look like, say, five years from now?  GATES: Well, it’s important to distinguish healthcare discovery activity from healthcare delivery activity. We focused mostly on delivery. I think it’s very much within the realm of possibility that the AI is not only accelerating healthcare discovery but substituting for a lot of the roles of, you know, I’m an organic chemist, or I run various types of assays. I can see those, which are, you know, testable-output-type jobs but with still very high value, I can see, you know, some replacement in those areas before the doctor.   The doctor, still understanding the human condition and long-term dialogues, you know, they’ve had a lifetime of reinforcement of that, particularly when you get into areas like mental health. So I wouldn’t say in five years, either people will choose to adopt it, but it will be profound that there’ll be this nearly free intelligence that can do follow-up, that can help you, you know, make sure you went through different possibilities.  And so I’d say, yes, we’ll have doctors, but I’d say healthcare will be massively transformed in its quality and in efficiency by AI in that time period.  LEE: Is there a comparison, useful comparison, say, between doctors and, say, programmers, computer programmers, or doctors and, I don’t know, lawyers?  GATES: Programming is another one that has, kind of, a mathematical correctness to it, you know, and so the objective function that you’re trying to reinforce to, as soon as you can understand the state machines, you can have something that’s “checkable”; that’s correct. So I think programming, you know, which is weird to say, that the machine will beat us at most programming tasks before we let it take over roles that have deep empathy, you know, physical presence and social understanding in them.  LEE: Yeah. By the way, you know, I fully expect in five years that AI will produce mathematical proofs that are checkable for validity, easily checkable, because they’ll be written in a proof-checking language like Lean or something but will be so complex that no human mathematician can understand them. I expect that to happen.   I can imagine in some fields, like cellular biology, we could have the same situation in the future because the molecular pathways, the chemistry, biochemistry of human cells or living cells is as complex as any mathematics, and so it seems possible that we may be in a state where in wet lab, we see, Oh yeah, this actually works, but no one can understand why.  BUBECK: Yeah, absolutely. I mean, I think I really agree with Bill’s distinction of the discovery and the delivery, and indeed, the discovery’s when you can check things, and at the end, there is an artifact that you can verify. You know, you can run the protocol in the wet lab and seeproduced what you wanted. So I absolutely agree with that.   And in fact, you know, we don’t have to talk five years from now. I don’t know if you know, but just recently, there was a paper that was published on a scientific discovery using o3- mini. So this is really amazing. And, you know, just very quickly, just so people know, it was about this statistical physics model, the frustrated Potts model, which has to do with coloring, and basically, the case of three colors, like, more than two colors was open for a long time, and o3 was able to reduce the case of three colors to two colors.   LEE: Yeah.  BUBECK: Which is just, like, astounding. And this is not … this is now. This is happening right now. So this is something that I personally didn’t expect it would happen so quickly, and it’s due to those reasoning models.   Now, on the delivery side, I would add something more to it for the reason why doctors and, in fact, lawyers and coders will remain for a long time, and it’s because we still don’t understand how those models generalize. Like, at the end of the day, we are not able to tell you when they are confronted with a really new, novel situation, whether they will work or not.  Nobody is able to give you that guarantee. And I think until we understand this generalization better, we’re not going to be willing to just let the system in the wild without human supervision.  LEE: But don’t human doctors, human specialists … so, for example, a cardiologist sees a patient in a certain way that a nephrologist …  BUBECK: Yeah. LEE: … or an endocrinologist might not. BUBECK: That’s right. But another cardiologist will understand and, kind of, expect a certain level of generalization from their peer. And this, we just don’t have it with AI models. Now, of course, you’re exactly right. That generalization is also hard for humans. Like, if you have a human trained for one task and you put them into another task, then you don’t … you often don’t know. LEE: OK. You know, the podcast is focused on what’s happened over the last two years. But now, I’d like one provocative prediction about what you think the world of AI and medicine is going to be at some point in the future. You pick your timeframe. I don’t care if it’s two years or 20 years from now, but, you know, what do you think will be different about AI in medicine in that future than today?  BUBECK: Yeah, I think the deployment is going to accelerate soon. Like, we’re really not missing very much. There is this enormous capability overhang. Like, even if progress completely stopped, with current systems, we can do a lot more than what we’re doing right now. So I think this will … this has to be realized, you know, sooner rather than later.  And I think it’s probably dependent on these benchmarks and proper evaluation and tying this with regulation. So these are things that take time in human society and for good reason. But now we already are at two years; you know, give it another two years and it should be really …   LEE: Will AI prescribe your medicines? Write your prescriptions?  BUBECK: I think yes. I think yes.  LEE: OK. Bill?  GATES: Well, I think the next two years, we’ll have massive pilots, and so the amount of use of the AI, still in a copilot-type mode, you know, we should get millions of patient visits, you know, both in general medicine and in the mental health side, as well. And I think that’s going to build up both the data and the confidence to give the AI some additional autonomy. You know, are you going to let it talk to you at night when you’re panicked about your mental health with some ability to escalate? And, you know, I’ve gone so far as to tell politicians with national health systems that if they deploy AI appropriately, that the quality of care, the overload of the doctors, the improvement in the economics will be enough that their voters will be stunned because they just don’t expect this, and, you know, they could be reelectedjust on this one thing of fixing what is a very overloaded and economically challenged health system in these rich countries.  You know, my personal role is going to be to make sure that in the poorer countries, there isn’t some lag; in fact, in many cases, that we’ll be more aggressive because, you know, we’re comparing to having no access to doctors at all. And, you know, so I think whether it’s India or Africa, there’ll be lessons that are globally valuable because we need medical intelligence. And, you know, thank god AI is going to provide a lot of that.  LEE: Well, on that optimistic note, I think that’s a good way to end. Bill, Seb, really appreciate all of this.   I think the most fundamental prediction we made in the book is that AI would actually find its way into the practice of medicine, and I think that that at least has come true, maybe in different ways than we expected, but it’s come true, and I think it’ll only accelerate from here. So thanks again, both of you.   GATES: Yeah. Thanks, you guys.  BUBECK: Thank you, Peter. Thanks, Bill.  LEE: I just always feel such a sense of privilege to have a chance to interact and actually work with people like Bill and Sébastien.    With Bill, I’m always amazed at how practically minded he is. He’s really thinking about the nuts and bolts of what AI might be able to do for people, and his thoughts about underserved parts of the world, the idea that we might actually be able to empower people with access to expert medical knowledge, I think is both inspiring and amazing.   And then, Seb, Sébastien Bubeck, he’s just absolutely a brilliant mind. He has a really firm grip on the deep mathematics of artificial intelligence and brings that to bear in his research and development work. And where that mathematics takes him isn’t just into the nuts and bolts of algorithms but into philosophical questions about the nature of intelligence.   One of the things that Sébastien brought up was the state of evaluation of AI systems. And indeed, he was fairly critical in our conversation. But of course, the world of AI research and development is just moving so fast, and indeed, since we recorded our conversation, OpenAI, in fact, released a new evaluation metric that is directly relevant to medical applications, and that is something called HealthBench. And Microsoft Research also released a new evaluation approach or process called ADeLe.   HealthBench and ADeLe are examples of new approaches to evaluating AI models that are less about testing their knowledge and ability to pass multiple-choice exams and instead are evaluation approaches designed to assess how well AI models are able to complete tasks that actually arise every day in typical healthcare or biomedical research settings. These are examples of really important good work that speak to how well AI models work in the real world of healthcare and biomedical research and how well they can collaborate with human beings in those settings.  You know, I asked Bill and Seb to make some predictions about the future. You know, my own answer, I expect that we’re going to be able to use AI to change how we diagnose patients, change how we decide treatment options.   If you’re a doctor or a nurse and you encounter a patient, you’ll ask questions, do a physical exam, you know, call out for labs just like you do today, but then you’ll be able to engage with AI based on all of that data and just ask, you know, based on all the other people who have gone through the same experience, who have similar data, how were they diagnosed? How were they treated? What were their outcomes? And what does that mean for the patient I have right now? Some people call it the “patients like me” paradigm. And I think that’s going to become real because of AI within our lifetimes. That idea of really grounding the delivery in healthcare and medical practice through data and intelligence, I actually now don’t see any barriers to that future becoming real.   I’d like to extend another big thank you to Bill and Sébastien for their time. And to our listeners, as always, it’s a pleasure to have you along for the ride. I hope you’ll join us for our remaining conversations, as well as a second coauthor roundtable with Carey and Zak.   Until next time.   #how #reshaping #future #healthcare #medical
    WWW.MICROSOFT.COM
    How AI is reshaping the future of healthcare and medical research
    Transcript [MUSIC]      [BOOK PASSAGE]   PETER LEE: “In ‘The Little Black Bag,’ a classic science fiction story, a high-tech doctor’s kit of the future is accidentally transported back to the 1950s, into the shaky hands of a washed-up, alcoholic doctor. The ultimate medical tool, it redeems the doctor wielding it, allowing him to practice gratifyingly heroic medicine. … The tale ends badly for the doctor and his treacherous assistant, but it offered a picture of how advanced technology could transform medicine—powerful when it was written nearly 75 years ago and still so today. What would be the Al equivalent of that little black bag? At this moment when new capabilities are emerging, how do we imagine them into medicine?”   [END OF BOOK PASSAGE]     [THEME MUSIC]     This is The AI Revolution in Medicine, Revisited. I’m your host, Peter Lee.    Shortly after OpenAI’s GPT-4 was publicly released, Carey Goldberg, Dr. Zak Kohane, and I published The AI Revolution in Medicine to help educate the world of healthcare and medical research about the transformative impact this new generative AI technology could have. But because we wrote the book when GPT-4 was still a secret, we had to speculate. Now, two years later, what did we get right, and what did we get wrong?     In this series, we’ll talk to clinicians, patients, hospital administrators, and others to understand the reality of AI in the field and where we go from here.   [THEME MUSIC FADES] The book passage I read at the top is from “Chapter 10: The Big Black Bag.”  In imagining AI in medicine, Carey, Zak, and I included in our book two fictional accounts. In the first, a medical resident consults GPT-4 on her personal phone as the patient in front of her crashes. Within seconds, it offers an alternate response based on recent literature. In the second account, a 90-year-old woman with several chronic conditions is living independently and receiving near-constant medical support from an AI aide.    In our conversations with the guests we’ve spoken to so far, we’ve caught a glimpse of these predicted futures, seeing how clinicians and patients are actually using AI today and how developers are leveraging the technology in the healthcare products and services they’re creating. In fact, that first fictional account isn’t so fictional after all, as most of the doctors in the real world actually appear to be using AI at least occasionally—and sometimes much more than occasionally—to help in their daily clinical work. And as for the second fictional account, which is more of a science fiction account, it seems we are indeed on the verge of a new way of delivering and receiving healthcare, though the future is still very much open.  As we continue to examine the current state of AI in healthcare and its potential to transform the field, I’m pleased to welcome Bill Gates and Sébastien Bubeck.   Bill may be best known as the co-founder of Microsoft, having created the company with his childhood friend Paul Allen in 1975. He’s now the founder of Breakthrough Energy, which aims to advance clean energy innovation, and TerraPower, a company developing groundbreaking nuclear energy and science technologies. He also chairs the world’s largest philanthropic organization, the Gates Foundation, and focuses on solving a variety of health challenges around the globe and here at home.  Sébastien is a research lead at OpenAI. He was previously a distinguished scientist, vice president of AI, and a colleague of mine here at Microsoft, where his work included spearheading the development of the family of small language models known as Phi. While at Microsoft, he also coauthored the discussion-provoking 2023 paper “Sparks of Artificial General Intelligence,” which presented the results of early experiments with GPT-4 conducted by a small team from Microsoft Research.    [TRANSITION MUSIC]   Here’s my conversation with Bill Gates and Sébastien Bubeck.  LEE: Bill, welcome.  BILL GATES: Thank you.  LEE: Seb …  SÉBASTIEN BUBECK: Yeah. Hi, hi, Peter. Nice to be here.  LEE: You know, one of the things that I’ve been doing just to get the conversation warmed up is to talk about origin stories, and what I mean about origin stories is, you know, what was the first contact that you had with large language models or the concept of generative AI that convinced you or made you think that something really important was happening?  And so, Bill, I think I’ve heard the story about, you know, the time when the OpenAI folks—Sam Altman, Greg Brockman, and others—showed you something, but could we hear from you what those early encounters were like and what was going through your mind?   GATES: Well, I’d been visiting OpenAI soon after it was created to see things like GPT-2 and to see the little arm they had that was trying to match human manipulation and, you know, looking at their games like Dota that they were trying to get as good as human play. And honestly, I didn’t think the language model stuff they were doing, even when they got to GPT-3, would show the ability to learn, you know, in the same sense that a human reads a biology book and is able to take that knowledge and access it not only to pass a test but also to create new medicines.  And so my challenge to them was that if their LLM could get a five on the advanced placement biology test, then I would say, OK, it took biologic knowledge and encoded it in an accessible way and that I didn’t expect them to do that very quickly but it would be profound.   And it was only about six months after I challenged them to do that, that an early version of GPT-4 they brought up to a dinner at my house, and in fact, it answered most of the questions that night very well. The one it got totally wrong, we were … because it was so good, we kept thinking, Oh, we must be wrong. It turned out it was a math weakness [LAUGHTER] that, you know, we later understood that that was an area of, weirdly, of incredible weakness of those early models. But, you know, that was when I realized, OK, the age of cheap intelligence was at its beginning.  LEE: Yeah. So I guess it seems like you had something similar to me in that my first encounters, I actually harbored some skepticism. Is it fair to say you were skeptical before that?  GATES: Well, the idea that we’ve figured out how to encode and access knowledge in this very deep sense without even understanding the nature of the encoding, …  LEE: Right.   GATES: … that is a bit weird.   LEE: Yeah.  GATES: We have an algorithm that creates the computation, but even say, OK, where is the president’s birthday stored in there? Where is this fact stored in there? The fact that even now when we’re playing around, getting a little bit more sense of it, it’s opaque to us what the semantic encoding is, it’s, kind of, amazing to me. I thought the invention of knowledge storage would be an explicit way of encoding knowledge, not an implicit statistical training.  LEE: Yeah, yeah. All right. So, Seb, you know, on this same topic, you know, I got—as we say at Microsoft—I got pulled into the tent. [LAUGHS]  BUBECK: Yes.   LEE: Because this was a very secret project. And then, um, I had the opportunity to select a small number of researchers in MSR [Microsoft Research] to join and start investigating this thing seriously. And the first person I pulled in was you.  BUBECK: Yeah.  LEE: And so what were your first encounters? Because I actually don’t remember what happened then.  BUBECK: Oh, I remember it very well. [LAUGHS] My first encounter with GPT-4 was in a meeting with the two of you, actually. But my kind of first contact, the first moment where I realized that something was happening with generative AI, was before that. And I agree with Bill that I also wasn’t too impressed by GPT-3.  I though that it was kind of, you know, very naturally mimicking the web, sort of parroting what was written there in a nice way. Still in a way which seemed very impressive. But it wasn’t really intelligent in any way. But shortly after GPT-3, there was a model before GPT-4 that really shocked me, and this was the first image generation model, DALL-E 1.  So that was in 2021. And I will forever remember the press release of OpenAI where they had this prompt of an avocado chair and then you had this image of the avocado chair. [LAUGHTER] And what really shocked me is that clearly the model kind of “understood” what is a chair, what is an avocado, and was able to merge those concepts.  So this was really, to me, the first moment where I saw some understanding in those models.   LEE: So this was, just to get the timing right, that was before I pulled you into the tent.  BUBECK: That was before. That was like a year before.  LEE: Right.   BUBECK: And now I will tell you how, you know, we went from that moment to the meeting with the two of you and GPT-4.  So once I saw this kind of understanding, I thought, OK, fine. It understands concept, but it’s still not able to reason. It cannot—as, you know, Bill was saying—it cannot learn from your document. It cannot reason.   So I set out to try to prove that. You know, this is what I was in the business of at the time, trying to prove things in mathematics. So I was trying to prove that basically autoregressive transformers could never reason. So I was trying to prove this. And after a year of work, I had something reasonable to show. And so I had the meeting with the two of you, and I had this example where I wanted to say, there is no way that an LLM is going to be able to do x.  And then as soon as I … I don’t know if you remember, Bill. But as soon as I said that, you said, oh, but wait a second. I had, you know, the OpenAI crew at my house recently, and they showed me a new model. Why don’t we ask this new model this question?   LEE: Yeah. BUBECK: And we did, and it solved it on the spot. And that really, honestly, just changed my life. Like, you know, I had been working for a year trying to say that this was impossible. And just right there, it was shown to be possible.   LEE: [LAUGHS] One of the very first things I got interested in—because I was really thinking a lot about healthcare—was healthcare and medicine.  And I don’t know if the two of you remember, but I ended up doing a lot of tests. I ran through, you know, step one and step two of the US Medical Licensing Exam. Did a whole bunch of other things. I wrote this big report. It was, you know, I can’t remember … a couple hundred pages.   And I needed to share this with someone. I didn’t … there weren’t too many people I could share it with. So I sent, I think, a copy to you, Bill. Sent a copy to you, Seb.   I hardly slept for about a week putting that report together. And, yeah, and I kept working on it. But I was far from alone. I think everyone who was in the tent, so to speak, in those early days was going through something pretty similar. All right. So I think … of course, a lot of what I put in the report also ended up being examples that made it into the book.  But the main purpose of this conversation isn’t to reminisce about [LAUGHS] or indulge in those reminiscences but to talk about what’s happening in healthcare and medicine. And, you know, as I said, we wrote this book. We did it very, very quickly. Seb, you helped. Bill, you know, you provided a review and some endorsements.  But, you know, honestly, we didn’t know what we were talking about because no one had access to this thing. And so we just made a bunch of guesses. So really, the whole thing I wanted to probe with the two of you is, now with two years of experience out in the world, what, you know, what do we think is happening today?  You know, is AI actually having an impact, positive or negative, on healthcare and medicine? And what do we now think is going to happen in the next two years, five years, or 10 years? And so I realize it’s a little bit too abstract to just ask it that way. So let me just try to narrow the discussion and guide us a little bit.   Um, the kind of administrative and clerical work, paperwork, around healthcare—and we made a lot of guesses about that—that appears to be going well, but, you know, Bill, I know we’ve discussed that sometimes that you think there ought to be a lot more going on. Do you have a viewpoint on how AI is actually finding its way into reducing paperwork?  GATES: Well, I’m stunned … I don’t think there should be a patient-doctor meeting where the AI is not sitting in and both transcribing, offering to help with the paperwork, and even making suggestions, although the doctor will be the one, you know, who makes the final decision about the diagnosis and whatever prescription gets done.   It’s so helpful. You know, when that patient goes home and their, you know, son who wants to understand what happened has some questions, that AI should be available to continue that conversation. And the way you can improve that experience and streamline things and, you know, involve the people who advise you. I don’t understand why that’s not more adopted, because there you still have the human in the loop making that final decision.  But even for, like, follow-up calls to make sure the patient did things, to understand if they have concerns and knowing when to escalate back to the doctor, the benefit is incredible. And, you know, that thing is ready for prime time. That paradigm is ready for prime time, in my view.  LEE: Yeah, there are some good products, but it seems like the number one use right now—and we kind of got this from some of the previous guests in previous episodes—is the use of AI just to respond to emails from patients. [LAUGHTER] Does that make sense to you?  BUBECK: Yeah. So maybe I want to second what Bill was saying but maybe take a step back first. You know, two years ago, like, the concept of clinical scribes, which is one of the things that we’re talking about right now, it would have sounded, in fact, it sounded two years ago, borderline dangerous. Because everybody was worried about hallucinations. What happened if you have this AI listening in and then it transcribes, you know, something wrong?  Now, two years later, I think it’s mostly working. And in fact, it is not yet, you know, fully adopted. You’re right. But it is in production. It is used, you know, in many, many places. So this rate of progress is astounding because it wasn’t obvious that we would be able to overcome those obstacles of hallucination. It’s not to say that hallucinations are fully solved. In the case of the closed system, they are.   Now, I think more generally what’s going on in the background is that there is something that we, that certainly I, underestimated, which is this management overhead. So I think the reason why this is not adopted everywhere is really a training and teaching aspect. People need to be taught, like, those systems, how to interact with them.  And one example that I really like, a study that recently appeared where they tried to use ChatGPT for diagnosis and they were comparing doctors without and with ChatGPT (opens in new tab). And the amazing thing … so this was a set of cases where the accuracy of the doctors alone was around 75%. ChatGPT alone was 90%. So that’s already kind of mind blowing. But then the kicker is that doctors with ChatGPT was 80%.   Intelligence alone is not enough. It’s also how it’s presented, how you interact with it. And ChatGPT, it’s an amazing tool. Obviously, I absolutely love it. But it’s not … you don’t want a doctor to have to type in, you know, prompts and use it that way.  It should be, as Bill was saying, kind of running continuously in the background, sending you notifications. And you have to be really careful of the rate at which those notifications are being sent. Because if they are too frequent, then the doctor will learn to ignore them. So you have to … all of those things matter, in fact, at least as much as the level of intelligence of the machine.  LEE: One of the things I think about, Bill, in that scenario that you described, doctors do some thinking about the patient when they write the note. So, you know, I’m always a little uncertain whether it’s actually … you know, you wouldn’t necessarily want to fully automate this, I don’t think. Or at least there needs to be some prompt to the doctor to make sure that the doctor puts some thought into what happened in the encounter with the patient. Does that make sense to you at all?  GATES: At this stage, you know, I’d still put the onus on the doctor to write the conclusions and the summary and not delegate that.  The tradeoffs you make a little bit are somewhat dependent on the situation you’re in. If you’re in Africa, So, yes, the doctor’s still going to have to do a lot of work, but just the quality of letting the patient and the people around them interact and ask questions and have things explained, that alone is such a quality improvement. It’s mind blowing.   LEE: So since you mentioned, you know, Africa—and, of course, this touches on the mission and some of the priorities of the Gates Foundation and this idea of democratization of access to expert medical care—what’s the most interesting stuff going on right now? Are there people and organizations or technologies that are impressing you or that you’re tracking?  GATES: Yeah. So the Gates Foundation has given out a lot of grants to people in Africa doing education, agriculture but more healthcare examples than anything. And the way these things start off, they often start out either being patient-centric in a narrow situation, like, OK, I’m a pregnant woman; talk to me. Or, I have infectious disease symptoms; talk to me. Or they’re connected to a health worker where they’re helping that worker get their job done. And we have lots of pilots out, you know, in both of those cases.   The dream would be eventually to have the thing the patient consults be so broad that it’s like having a doctor available who understands the local things.   LEE: Right.   GATES: We’re not there yet. But over the next two or three years, you know, particularly given the worsening financial constraints against African health systems, where the withdrawal of money has been dramatic, you know, figuring out how to take this—what I sometimes call “free intelligence”—and build a quality health system around that, we will have to be more radical in low-income countries than any rich country is ever going to be.   LEE: Also, there’s maybe a different regulatory environment, so some of those things maybe are easier? Because right now, I think the world hasn’t figured out how to and whether to regulate, let’s say, an AI that might give a medical diagnosis or write a prescription for a medication.  BUBECK: Yeah. I think one issue with this, and it’s also slowing down the deployment of AI in healthcare more generally, is a lack of proper benchmark. Because, you know, you were mentioning the USMLE [United States Medical Licensing Examination], for example. That’s a great test to test human beings and their knowledge of healthcare and medicine. But it’s not a great test to give to an AI.  It’s not asking the right questions. So finding what are the right questions to test whether an AI system is ready to give diagnosis in a constrained setting, that’s a very, very important direction, which to my surprise, is not yet accelerating at the rate that I was hoping for.  LEE: OK, so that gives me an excuse to get more now into the core AI tech because something I’ve discussed with both of you is this issue of what are the right tests. And you both know the very first test I give to any new spin of an LLM is I present a patient, the results—a mythical patient—the results of my physical exam, my mythical physical exam. Maybe some results of some initial labs. And then I present or propose a differential diagnosis. And if you’re not in medicine, a differential diagnosis you can just think of as a prioritized list of the possible diagnoses that fit with all that data. And in that proposed differential, I always intentionally make two mistakes.  I make a textbook technical error in one of the possible elements of the differential diagnosis, and I have an error of omission. And, you know, I just want to know, does the LLM understand what I’m talking about? And all the good ones out there do now. But then I want to know, can it spot the errors? And then most importantly, is it willing to tell me I’m wrong, that I’ve made a mistake?   That last piece seems really hard for AI today. And so let me ask you first, Seb, because at the time of this taping, of course, there was a new spin of GPT-4o last week that became overly sycophantic. In other words, it was actually prone in that test of mine not only to not tell me I’m wrong, but it actually praised me for the creativity of my differential. [LAUGHTER] What’s up with that?  BUBECK: Yeah, I guess it’s a testament to the fact that training those models is still more of an art than a science. So it’s a difficult job. Just to be clear with the audience, we have rolled back that [LAUGHS] version of GPT-4o, so now we don’t have the sycophant version out there.  Yeah, no, it’s a really difficult question. It has to do … as you said, it’s very technical. It has to do with the post-training and how, like, where do you nudge the model? So, you know, there is this very classical by now technique called RLHF [reinforcement learning from human feedback], where you push the model in the direction of a certain reward model. So the reward model is just telling the model, you know, what behavior is good, what behavior is bad.  But this reward model is itself an LLM, and, you know, Bill was saying at the very beginning of the conversation that we don’t really understand how those LLMs deal with concepts like, you know, where is the capital of France located? Things like that. It is the same thing for this reward model. We don’t know why it says that it prefers one output to another, and whether this is correlated with some sycophancy is, you know, something that we discovered basically just now. That if you push too hard in optimization on this reward model, you will get a sycophant model.  So it’s kind of … what I’m trying to say is we became too good at what we were doing, and we ended up, in fact, in a trap of the reward model.  LEE: I mean, you do want … it’s a difficult balance because you do want models to follow your desires and …  BUBECK: It’s a very difficult, very difficult balance.  LEE: So this brings up then the following question for me, which is the extent to which we think we’ll need to have specially trained models for things. So let me start with you, Bill. Do you have a point of view on whether we will need to, you know, quote-unquote take AI models to med school? Have them specially trained? Like, if you were going to deploy something to give medical care in underserved parts of the world, do we need to do something special to create those models?  GATES: We certainly need to teach them the African languages and the unique dialects so that the multimedia interactions are very high quality. We certainly need to teach them the disease prevalence and unique disease patterns like, you know, neglected tropical diseases and malaria. So we need to gather a set of facts that somebody trying to go for a US customer base, you know, wouldn’t necessarily have that in there.  Those two things are actually very straightforward because the additional training time is small. I’d say for the next few years, we’ll also need to do reinforcement learning about the context of being a doctor and how important certain behaviors are. Humans learn over the course of their life to some degree that, I’m in a different context and the way I behave in terms of being willing to criticize or be nice, you know, how important is it? Who’s here? What’s my relationship to them?   Right now, these machines don’t have that broad social experience. And so if you know it’s going to be used for health things, a lot of reinforcement learning of the very best humans in that context would still be valuable. Eventually, the models will, having read all the literature of the world about good doctors, bad doctors, it’ll understand as soon as you say, “I want you to be a doctor diagnosing somebody.” All of the implicit reinforcement that fits that situation, you know, will be there. LEE: Yeah. GATES: And so I hope three years from now, we don’t have to do that reinforcement learning. But today, for any medical context, you would want a lot of data to reinforce tone, willingness to say things when, you know, there might be something significant at stake.  LEE: Yeah. So, you know, something Bill said, kind of, reminds me of another thing that I think we missed, which is, the context also … and the specialization also pertains to different, I guess, what we still call “modes,” although I don’t know if the idea of multimodal is the same as it was two years ago. But, you know, what do you make of all of the hubbub around—in fact, within Microsoft Research, this is a big deal, but I think we’re far from alone—you know, medical images and vision, video, proteins and molecules, cell, you know, cellular data and so on.  BUBECK: Yeah. OK. So there is a lot to say to everything … to the last, you know, couple of minutes. Maybe on the specialization aspect, you know, I think there is, hiding behind this, a really fundamental scientific question of whether eventually we have a singular AGI [artificial general intelligence] that kind of knows everything and you can just put, you know, explain your own context and it will just get it and understand everything.  That’s one vision. I have to say, I don’t particularly believe in this vision. In fact, we humans are not like that at all. I think, hopefully, we are general intelligences, yet we have to specialize a lot. And, you know, I did myself a lot of RL, reinforcement learning, on mathematics. Like, that’s what I did, you know, spent a lot of time doing that. And I didn’t improve on other aspects. You know, in fact, I probably degraded in other aspects. [LAUGHTER] So it’s … I think it’s an important example to have in mind.  LEE: I think I might disagree with you on that, though, because, like, doesn’t a model have to see both good science and bad science in order to be able to gain the ability to discern between the two?  BUBECK: Yeah, no, that absolutely. I think there is value in seeing the generality, in having a very broad base. But then you, kind of, specialize on verticals. And this is where also, you know, open-weights model, which we haven’t talked about yet, are really important because they allow you to provide this broad base to everyone. And then you can specialize on top of it.  LEE: So we have about three hours of stuff to talk about, but our time is actually running low. BUBECK: Yes, yes, yes.   LEE: So I think I want … there’s a more provocative question. It’s almost a silly question, but I need to ask it of the two of you, which is, is there a future, you know, where AI replaces doctors or replaces, you know, medical specialties that we have today? So what does the world look like, say, five years from now?  GATES: Well, it’s important to distinguish healthcare discovery activity from healthcare delivery activity. We focused mostly on delivery. I think it’s very much within the realm of possibility that the AI is not only accelerating healthcare discovery but substituting for a lot of the roles of, you know, I’m an organic chemist, or I run various types of assays. I can see those, which are, you know, testable-output-type jobs but with still very high value, I can see, you know, some replacement in those areas before the doctor.   The doctor, still understanding the human condition and long-term dialogues, you know, they’ve had a lifetime of reinforcement of that, particularly when you get into areas like mental health. So I wouldn’t say in five years, either people will choose to adopt it, but it will be profound that there’ll be this nearly free intelligence that can do follow-up, that can help you, you know, make sure you went through different possibilities.  And so I’d say, yes, we’ll have doctors, but I’d say healthcare will be massively transformed in its quality and in efficiency by AI in that time period.  LEE: Is there a comparison, useful comparison, say, between doctors and, say, programmers, computer programmers, or doctors and, I don’t know, lawyers?  GATES: Programming is another one that has, kind of, a mathematical correctness to it, you know, and so the objective function that you’re trying to reinforce to, as soon as you can understand the state machines, you can have something that’s “checkable”; that’s correct. So I think programming, you know, which is weird to say, that the machine will beat us at most programming tasks before we let it take over roles that have deep empathy, you know, physical presence and social understanding in them.  LEE: Yeah. By the way, you know, I fully expect in five years that AI will produce mathematical proofs that are checkable for validity, easily checkable, because they’ll be written in a proof-checking language like Lean or something but will be so complex that no human mathematician can understand them. I expect that to happen.   I can imagine in some fields, like cellular biology, we could have the same situation in the future because the molecular pathways, the chemistry, biochemistry of human cells or living cells is as complex as any mathematics, and so it seems possible that we may be in a state where in wet lab, we see, Oh yeah, this actually works, but no one can understand why.  BUBECK: Yeah, absolutely. I mean, I think I really agree with Bill’s distinction of the discovery and the delivery, and indeed, the discovery’s when you can check things, and at the end, there is an artifact that you can verify. You know, you can run the protocol in the wet lab and see [if you have] produced what you wanted. So I absolutely agree with that.   And in fact, you know, we don’t have to talk five years from now. I don’t know if you know, but just recently, there was a paper that was published on a scientific discovery using o3- mini (opens in new tab). So this is really amazing. And, you know, just very quickly, just so people know, it was about this statistical physics model, the frustrated Potts model, which has to do with coloring, and basically, the case of three colors, like, more than two colors was open for a long time, and o3 was able to reduce the case of three colors to two colors.   LEE: Yeah.  BUBECK: Which is just, like, astounding. And this is not … this is now. This is happening right now. So this is something that I personally didn’t expect it would happen so quickly, and it’s due to those reasoning models.   Now, on the delivery side, I would add something more to it for the reason why doctors and, in fact, lawyers and coders will remain for a long time, and it’s because we still don’t understand how those models generalize. Like, at the end of the day, we are not able to tell you when they are confronted with a really new, novel situation, whether they will work or not.  Nobody is able to give you that guarantee. And I think until we understand this generalization better, we’re not going to be willing to just let the system in the wild without human supervision.  LEE: But don’t human doctors, human specialists … so, for example, a cardiologist sees a patient in a certain way that a nephrologist …  BUBECK: Yeah. LEE: … or an endocrinologist might not. BUBECK: That’s right. But another cardiologist will understand and, kind of, expect a certain level of generalization from their peer. And this, we just don’t have it with AI models. Now, of course, you’re exactly right. That generalization is also hard for humans. Like, if you have a human trained for one task and you put them into another task, then you don’t … you often don’t know. LEE: OK. You know, the podcast is focused on what’s happened over the last two years. But now, I’d like one provocative prediction about what you think the world of AI and medicine is going to be at some point in the future. You pick your timeframe. I don’t care if it’s two years or 20 years from now, but, you know, what do you think will be different about AI in medicine in that future than today?  BUBECK: Yeah, I think the deployment is going to accelerate soon. Like, we’re really not missing very much. There is this enormous capability overhang. Like, even if progress completely stopped, with current systems, we can do a lot more than what we’re doing right now. So I think this will … this has to be realized, you know, sooner rather than later.  And I think it’s probably dependent on these benchmarks and proper evaluation and tying this with regulation. So these are things that take time in human society and for good reason. But now we already are at two years; you know, give it another two years and it should be really …   LEE: Will AI prescribe your medicines? Write your prescriptions?  BUBECK: I think yes. I think yes.  LEE: OK. Bill?  GATES: Well, I think the next two years, we’ll have massive pilots, and so the amount of use of the AI, still in a copilot-type mode, you know, we should get millions of patient visits, you know, both in general medicine and in the mental health side, as well. And I think that’s going to build up both the data and the confidence to give the AI some additional autonomy. You know, are you going to let it talk to you at night when you’re panicked about your mental health with some ability to escalate? And, you know, I’ve gone so far as to tell politicians with national health systems that if they deploy AI appropriately, that the quality of care, the overload of the doctors, the improvement in the economics will be enough that their voters will be stunned because they just don’t expect this, and, you know, they could be reelected [LAUGHTER] just on this one thing of fixing what is a very overloaded and economically challenged health system in these rich countries.  You know, my personal role is going to be to make sure that in the poorer countries, there isn’t some lag; in fact, in many cases, that we’ll be more aggressive because, you know, we’re comparing to having no access to doctors at all. And, you know, so I think whether it’s India or Africa, there’ll be lessons that are globally valuable because we need medical intelligence. And, you know, thank god AI is going to provide a lot of that.  LEE: Well, on that optimistic note, I think that’s a good way to end. Bill, Seb, really appreciate all of this.   I think the most fundamental prediction we made in the book is that AI would actually find its way into the practice of medicine, and I think that that at least has come true, maybe in different ways than we expected, but it’s come true, and I think it’ll only accelerate from here. So thanks again, both of you.  [TRANSITION MUSIC]  GATES: Yeah. Thanks, you guys.  BUBECK: Thank you, Peter. Thanks, Bill.  LEE: I just always feel such a sense of privilege to have a chance to interact and actually work with people like Bill and Sébastien.    With Bill, I’m always amazed at how practically minded he is. He’s really thinking about the nuts and bolts of what AI might be able to do for people, and his thoughts about underserved parts of the world, the idea that we might actually be able to empower people with access to expert medical knowledge, I think is both inspiring and amazing.   And then, Seb, Sébastien Bubeck, he’s just absolutely a brilliant mind. He has a really firm grip on the deep mathematics of artificial intelligence and brings that to bear in his research and development work. And where that mathematics takes him isn’t just into the nuts and bolts of algorithms but into philosophical questions about the nature of intelligence.   One of the things that Sébastien brought up was the state of evaluation of AI systems. And indeed, he was fairly critical in our conversation. But of course, the world of AI research and development is just moving so fast, and indeed, since we recorded our conversation, OpenAI, in fact, released a new evaluation metric that is directly relevant to medical applications, and that is something called HealthBench. And Microsoft Research also released a new evaluation approach or process called ADeLe.   HealthBench and ADeLe are examples of new approaches to evaluating AI models that are less about testing their knowledge and ability to pass multiple-choice exams and instead are evaluation approaches designed to assess how well AI models are able to complete tasks that actually arise every day in typical healthcare or biomedical research settings. These are examples of really important good work that speak to how well AI models work in the real world of healthcare and biomedical research and how well they can collaborate with human beings in those settings.  You know, I asked Bill and Seb to make some predictions about the future. You know, my own answer, I expect that we’re going to be able to use AI to change how we diagnose patients, change how we decide treatment options.   If you’re a doctor or a nurse and you encounter a patient, you’ll ask questions, do a physical exam, you know, call out for labs just like you do today, but then you’ll be able to engage with AI based on all of that data and just ask, you know, based on all the other people who have gone through the same experience, who have similar data, how were they diagnosed? How were they treated? What were their outcomes? And what does that mean for the patient I have right now? Some people call it the “patients like me” paradigm. And I think that’s going to become real because of AI within our lifetimes. That idea of really grounding the delivery in healthcare and medical practice through data and intelligence, I actually now don’t see any barriers to that future becoming real.  [THEME MUSIC]  I’d like to extend another big thank you to Bill and Sébastien for their time. And to our listeners, as always, it’s a pleasure to have you along for the ride. I hope you’ll join us for our remaining conversations, as well as a second coauthor roundtable with Carey and Zak.   Until next time.   [MUSIC FADES]
    0 Commentarii 0 Distribuiri
  • My unexpected Pride icon: Link from the Zelda games, a non-binary hero who helped me work out who I was

    Growing up steeped in the aggressive gender stereotypes of the 1990s was a real trip for most queer millennials, but I think gamers had it especially hard. Almost all video game characters were hypermasculine military men, unrealistically curvaceous fantasy women wearing barely enough armour to cover their nipples, or cartoon animals. Most of these characters catered exclusively to straight teenage boys; overt queer representation in games was pretty much nonexistent until the mid 2010s. Before that, we had to take what we could get. And what I had was Link, from The Legend of Zelda.Link. Composite: Guardian Design; Zuma Press/AlamyLink is a boy, but he didn’t really look like one. He wore a green tunic and a serious expression under a mop of blond hair. He is the adventurous, mostly silent hero of the Zelda games, unassuming and often vulnerable, but also resourceful, daring and handy with a sword. In most of the early Zelda games, he is a kid of about 10, but even when he grew into a teenager in 1998’s Ocarina of Time on the Nintendo 64, he didn’t become a furious lump of muscle. He stayed androgynous, in his tunic and tights. As a kid, I would dress up like him for Halloween, carefully centre-parting my blond fringe. Link may officially be a boy, but for me he has always been a non-binary icon.As time has gone on and game graphics have evolved, Link has stayed somewhat gender-ambiguous. Gay guys and gender-fluid types alike appreciate his ageless twink energy. And given the total lack of thought that most game developers gave to players who weren’t straight and male, I felt vindicated when I found out that this was intentional. In 2016, the Zelda series’ producer Eiji Aonuma told Time magazine that the development team had experimented a little with Link’s gender presentation over the years, but that he felt that the character’s androgyny was part of who he was.“back during the Ocarina of Time days, I wanted Link to be gender neutral,” he said. “I wanted the player to think: ‘Maybe Link is a boy or a girl.’ If you saw Link as a guy, he’d have more of a feminine touch. Or vice versa … I’ve always thought that for either female or male players, I wanted them to be able to relate to Link.”As it turns out, Link appeals perhaps most of all to those of us somewhere in between. In 2023, the tech blog io9 spoke to many transgender and non-binary people who saw something of themselves in Link: he has acquired a reputation as an egg-cracker, a fictional character who prompts a realisation about your own gender identity.Despite their outdated reputation as a pursuit for adolescent boys, video games have always been playgrounds for gender experimentation and expression. There are legions of trans, non-binary and gender non-conforming people who first started exploring their identity with customisable game characters in World of Warcraft, or gender-swapping themselves in The Sims – the digital equivalent of dressing up. Video games are the closest you can come to stepping into a new body for a bit and seeing how it feels.It is no surprise to me that a lot of queer people are drawn to video games. A 2024 survey by GLAAD found that 17% of gamers identify as LGBTQ+, a huge number compared with the general population. It may be because people who play games skew younger – 40 and below – but I also think it’s because gender is all about play. What fun it is to mess with the rules, subvert people’s expectations and create your own character. It is as empowering as any world-saving quest.
    #unexpected #pride #icon #link #zelda
    My unexpected Pride icon: Link from the Zelda games, a non-binary hero who helped me work out who I was
    Growing up steeped in the aggressive gender stereotypes of the 1990s was a real trip for most queer millennials, but I think gamers had it especially hard. Almost all video game characters were hypermasculine military men, unrealistically curvaceous fantasy women wearing barely enough armour to cover their nipples, or cartoon animals. Most of these characters catered exclusively to straight teenage boys; overt queer representation in games was pretty much nonexistent until the mid 2010s. Before that, we had to take what we could get. And what I had was Link, from The Legend of Zelda.Link. Composite: Guardian Design; Zuma Press/AlamyLink is a boy, but he didn’t really look like one. He wore a green tunic and a serious expression under a mop of blond hair. He is the adventurous, mostly silent hero of the Zelda games, unassuming and often vulnerable, but also resourceful, daring and handy with a sword. In most of the early Zelda games, he is a kid of about 10, but even when he grew into a teenager in 1998’s Ocarina of Time on the Nintendo 64, he didn’t become a furious lump of muscle. He stayed androgynous, in his tunic and tights. As a kid, I would dress up like him for Halloween, carefully centre-parting my blond fringe. Link may officially be a boy, but for me he has always been a non-binary icon.As time has gone on and game graphics have evolved, Link has stayed somewhat gender-ambiguous. Gay guys and gender-fluid types alike appreciate his ageless twink energy. And given the total lack of thought that most game developers gave to players who weren’t straight and male, I felt vindicated when I found out that this was intentional. In 2016, the Zelda series’ producer Eiji Aonuma told Time magazine that the development team had experimented a little with Link’s gender presentation over the years, but that he felt that the character’s androgyny was part of who he was.“back during the Ocarina of Time days, I wanted Link to be gender neutral,” he said. “I wanted the player to think: ‘Maybe Link is a boy or a girl.’ If you saw Link as a guy, he’d have more of a feminine touch. Or vice versa … I’ve always thought that for either female or male players, I wanted them to be able to relate to Link.”As it turns out, Link appeals perhaps most of all to those of us somewhere in between. In 2023, the tech blog io9 spoke to many transgender and non-binary people who saw something of themselves in Link: he has acquired a reputation as an egg-cracker, a fictional character who prompts a realisation about your own gender identity.Despite their outdated reputation as a pursuit for adolescent boys, video games have always been playgrounds for gender experimentation and expression. There are legions of trans, non-binary and gender non-conforming people who first started exploring their identity with customisable game characters in World of Warcraft, or gender-swapping themselves in The Sims – the digital equivalent of dressing up. Video games are the closest you can come to stepping into a new body for a bit and seeing how it feels.It is no surprise to me that a lot of queer people are drawn to video games. A 2024 survey by GLAAD found that 17% of gamers identify as LGBTQ+, a huge number compared with the general population. It may be because people who play games skew younger – 40 and below – but I also think it’s because gender is all about play. What fun it is to mess with the rules, subvert people’s expectations and create your own character. It is as empowering as any world-saving quest. #unexpected #pride #icon #link #zelda
    WWW.THEGUARDIAN.COM
    My unexpected Pride icon: Link from the Zelda games, a non-binary hero who helped me work out who I was
    Growing up steeped in the aggressive gender stereotypes of the 1990s was a real trip for most queer millennials, but I think gamers had it especially hard. Almost all video game characters were hypermasculine military men, unrealistically curvaceous fantasy women wearing barely enough armour to cover their nipples, or cartoon animals. Most of these characters catered exclusively to straight teenage boys (or, I guess, furries); overt queer representation in games was pretty much nonexistent until the mid 2010s. Before that, we had to take what we could get. And what I had was Link, from The Legend of Zelda.Link. Composite: Guardian Design; Zuma Press/AlamyLink is a boy, but he didn’t really look like one. He wore a green tunic and a serious expression under a mop of blond hair. He is the adventurous, mostly silent hero of the Zelda games, unassuming and often vulnerable, but also resourceful, daring and handy with a sword. In most of the early Zelda games, he is a kid of about 10, but even when he grew into a teenager in 1998’s Ocarina of Time on the Nintendo 64, he didn’t become a furious lump of muscle. He stayed androgynous, in his tunic and tights. As a kid, I would dress up like him for Halloween, carefully centre-parting my blond fringe. Link may officially be a boy, but for me he has always been a non-binary icon.As time has gone on and game graphics have evolved, Link has stayed somewhat gender-ambiguous. Gay guys and gender-fluid types alike appreciate his ageless twink energy. And given the total lack of thought that most game developers gave to players who weren’t straight and male, I felt vindicated when I found out that this was intentional. In 2016, the Zelda series’ producer Eiji Aonuma told Time magazine that the development team had experimented a little with Link’s gender presentation over the years, but that he felt that the character’s androgyny was part of who he was.“[Even] back during the Ocarina of Time days, I wanted Link to be gender neutral,” he said. “I wanted the player to think: ‘Maybe Link is a boy or a girl.’ If you saw Link as a guy, he’d have more of a feminine touch. Or vice versa … I’ve always thought that for either female or male players, I wanted them to be able to relate to Link.”As it turns out, Link appeals perhaps most of all to those of us somewhere in between. In 2023, the tech blog io9 spoke to many transgender and non-binary people who saw something of themselves in Link: he has acquired a reputation as an egg-cracker, a fictional character who prompts a realisation about your own gender identity.Despite their outdated reputation as a pursuit for adolescent boys, video games have always been playgrounds for gender experimentation and expression. There are legions of trans, non-binary and gender non-conforming people who first started exploring their identity with customisable game characters in World of Warcraft, or gender-swapping themselves in The Sims – the digital equivalent of dressing up. Video games are the closest you can come to stepping into a new body for a bit and seeing how it feels.It is no surprise to me that a lot of queer people are drawn to video games. A 2024 survey by GLAAD found that 17% of gamers identify as LGBTQ+, a huge number compared with the general population. It may be because people who play games skew younger – 40 and below – but I also think it’s because gender is all about play. What fun it is to mess with the rules, subvert people’s expectations and create your own character. It is as empowering as any world-saving quest.
    0 Commentarii 0 Distribuiri
  • One of the most versatile action cameras I've tested isn't from GoPro - and it's on sale

    DJI Osmo Action 4. Adrian Kingsley-Hughes/ZDNETMultiple DJI Osmo Action 4 packages are on sale . Both the Essential and Standard Combos have been discounted to while the Adventure Combo has dropped to DJI might not be the first name on people's lips when it comes to action cameras, but the company that's better known for its drones also has a really solid line of action cameras. And its latest device, the Osmo Action 4 camera, has some very impressive tricks up its sleeve.Also: One of the most versatile cameras I've used is not from Sony or Canon and it's on saleSo, what sets this action camera apart from the competition? Let's take a look.
    details
    View First off, this is not just an action camera -- it's a pro-grade action camera.From a hardware point of view, the Osmo Action 4 features a 1/1.3-inch image sensor that can record 4K at up to 120 frames per second. This sensor is combined with a wide-angle f/2.8 aperture lens that provides an ultra-wide field of view of up to 155°. And that's wide. Build quality and fit and finish are second to none. Adrian Kingsley-Hughes/ZDNETFor when the going gets rough, the Osmo Action 4 offers 360° HorizonSteady stabilization modes, including RockSteady 3.0/3.0+ for first-person video footage and HorizonBalancing/HorizonSteady modes for horizontal shots. That's pro-grade hardware right there.Also: This new AI video editor is an all-in-one production service for filmmakers - how to try itThe Osmo Action 4 also features a 10-bit D-Log M color mode. This mode allows the sensor to record over one billion colors and offers a wider dynamic range, giving you a video that is more vivid and that offers greater detail in the highlights and shadows. This mode, combined with an advanced color temperature sensor, means that the colors have a true-to-life feel regardless of whether you're shooting outdoors, indoors, or even underwater. The DJI Osmo Action 4 ready for action. Adrian Kingsley-Hughes/ZDNETI've added some video output from the Osmo Action 4 below. There are examples in both 1080p and 4K. To test the stabilization, I attached the camera to the truck and took it on some roads, some of which are pretty rough. The Osmo Action 4 had no problem with that terrain. I also popped the camera into the sea, just because. And again, no problem.I've also captured a few time-lapses with the camera -- not because I like clouds, but pointing a camera at a sky can be a good test of how it handles changing light. Also: I recommend this action camera to beginners and professional creators. Here's whyTimelapses with action cameras can suffer from unsightly exposure changes that cause the image to pulse, a condition known as exposure pumping. This issue can also cause the white balance to change noticeably in a video, but the Osmo Action 4 handled this test well.All the footage I've shot is what I've come to expect from a DJI camera, whether it's from an action camera or drone -- crisp, clear, vivid, and also nice and stable.The Osmo Action 4 is packed with various electronic image-stabilizationtech to ensure that your footage is smooth and on the horizon. It's worth noting the limitations of EIS -- it's not supported in slow-motion and timelapse modes, and the HorizonSteady and HorizonBalancing features are only available for video recorded at 1080por 2.7Kwith a frame rate of 60fps or below. On the durability front, I've no concerns. I've subjected the Osmo Action 4 to a hard few days of testing, and it's not let me down or complained once. It takes impacts like a champ, and being underwater or in dirt and sand is no problem at all. Also: I'm a full-time Canon photographer, but this Nikon camera made me wonder if I'm missing outYou might think that this heavy-duty testing would be hard on the camera's tiny batteries, but you'd be wrong. Remember I said the Osmo Action 4 offered hours of battery life? Well, I wasn't kidding.  The Osmo Action 4's ultra-long life batteries are incredible.  Adrian Kingsley-Hughes/ZDNETDJI says that a single battery can deliver up to 160 minutes of 1080p/24fps video recording. That's over two and a half hours of recording time. In the real world, I was blown away by how much a single battery can deliver. I shot video and timelapse, messed around with a load of camera settings, and then transferred that footage to my iPhone, and still had 16% battery left.No action camera has delivered so much for me on one battery. The two extra batteries and the multifunction case that come as part of the Adventure Combo are worth the extra Adrian Kingsley-Hughes/ZDNETAnd when you're ready to recharge, a 30W USB-C charger can take a battery from zero to 80% in 18 minutes. That's also impressive.What's more, the batteries are resistant to cold, offering up to 150 minutes of 1080p/24fps recording in temperatures as low as -20°C. This resistance also blows the competition away.Even taking into account all these strong points, the Osmo Action 4 offers even more.The camera has 2x digital zoom for better composition, Voice Prompts that let you know what the camera is doing without looking, and Voice Control that lets you operate the device without touching the screen or using the app. The Osmo Action 4 also digitally hides the selfie stick from a variety of different shots, and you can even connect the DJI Mic to the camera via the USB-C port for better audio capture.Also: Yes, an Android tablet finally made me reconsider my iPad Pro loyaltyAs for price, the Osmo Action 4 Standard Combo bundle comes in at while the Osmo Action 4 Adventure Combo, which comes with two extra Osmo Action Extreme batteries, an additional mini Osmo Action quick-release adapter mount, a battery case that acts as a power bank, and a 1.5-meter selfie stick, is I'm in love with the Osmo Action 4. It's hands down the best, most versatile, most powerful action camera on the market today, offering pro-grade features at a price that definitely isn't pro-grade.  Everything included in the Action Combo bundle. DJIDJI Osmo Action 4 tech specsDimensions: 70.5×44.2×32.8mmWeight: 145gWaterproof: 18m, up to 60m with the optional waterproof case Microphones: 3Sensor 1/1.3-inch CMOSLens: FOV 155°, aperture f/2.8, focus distance 0.4m to ∞Max Photo Resolution: 3648×2736Max Video Resolution: 4K: 3840×2880@24/25/30/48/50/60fps and 4K: 3840×2160@24/25/30/48/50/60/100/120fpsISO Range: 100-12800Front Screen: 1.4-inch, 323ppi, 320×320Rear Screen: 2.25-inch, 326ppi, 360×640Front/Rear Screen Brightness: 750±50 cd/m² Storage: microSDBattery: 1770mAh, lab tested to offer up to 160 minutes of runtimeOperating Temperature: -20° to 45° CThis article was originally published in August of 2023 and updated in March 2025.Featured reviews
    #one #most #versatile #action #cameras
    One of the most versatile action cameras I've tested isn't from GoPro - and it's on sale
    DJI Osmo Action 4. Adrian Kingsley-Hughes/ZDNETMultiple DJI Osmo Action 4 packages are on sale . Both the Essential and Standard Combos have been discounted to while the Adventure Combo has dropped to DJI might not be the first name on people's lips when it comes to action cameras, but the company that's better known for its drones also has a really solid line of action cameras. And its latest device, the Osmo Action 4 camera, has some very impressive tricks up its sleeve.Also: One of the most versatile cameras I've used is not from Sony or Canon and it's on saleSo, what sets this action camera apart from the competition? Let's take a look. details View First off, this is not just an action camera -- it's a pro-grade action camera.From a hardware point of view, the Osmo Action 4 features a 1/1.3-inch image sensor that can record 4K at up to 120 frames per second. This sensor is combined with a wide-angle f/2.8 aperture lens that provides an ultra-wide field of view of up to 155°. And that's wide. Build quality and fit and finish are second to none. Adrian Kingsley-Hughes/ZDNETFor when the going gets rough, the Osmo Action 4 offers 360° HorizonSteady stabilization modes, including RockSteady 3.0/3.0+ for first-person video footage and HorizonBalancing/HorizonSteady modes for horizontal shots. That's pro-grade hardware right there.Also: This new AI video editor is an all-in-one production service for filmmakers - how to try itThe Osmo Action 4 also features a 10-bit D-Log M color mode. This mode allows the sensor to record over one billion colors and offers a wider dynamic range, giving you a video that is more vivid and that offers greater detail in the highlights and shadows. This mode, combined with an advanced color temperature sensor, means that the colors have a true-to-life feel regardless of whether you're shooting outdoors, indoors, or even underwater. The DJI Osmo Action 4 ready for action. Adrian Kingsley-Hughes/ZDNETI've added some video output from the Osmo Action 4 below. There are examples in both 1080p and 4K. To test the stabilization, I attached the camera to the truck and took it on some roads, some of which are pretty rough. The Osmo Action 4 had no problem with that terrain. I also popped the camera into the sea, just because. And again, no problem.I've also captured a few time-lapses with the camera -- not because I like clouds, but pointing a camera at a sky can be a good test of how it handles changing light. Also: I recommend this action camera to beginners and professional creators. Here's whyTimelapses with action cameras can suffer from unsightly exposure changes that cause the image to pulse, a condition known as exposure pumping. This issue can also cause the white balance to change noticeably in a video, but the Osmo Action 4 handled this test well.All the footage I've shot is what I've come to expect from a DJI camera, whether it's from an action camera or drone -- crisp, clear, vivid, and also nice and stable.The Osmo Action 4 is packed with various electronic image-stabilizationtech to ensure that your footage is smooth and on the horizon. It's worth noting the limitations of EIS -- it's not supported in slow-motion and timelapse modes, and the HorizonSteady and HorizonBalancing features are only available for video recorded at 1080por 2.7Kwith a frame rate of 60fps or below. On the durability front, I've no concerns. I've subjected the Osmo Action 4 to a hard few days of testing, and it's not let me down or complained once. It takes impacts like a champ, and being underwater or in dirt and sand is no problem at all. Also: I'm a full-time Canon photographer, but this Nikon camera made me wonder if I'm missing outYou might think that this heavy-duty testing would be hard on the camera's tiny batteries, but you'd be wrong. Remember I said the Osmo Action 4 offered hours of battery life? Well, I wasn't kidding.  The Osmo Action 4's ultra-long life batteries are incredible.  Adrian Kingsley-Hughes/ZDNETDJI says that a single battery can deliver up to 160 minutes of 1080p/24fps video recording. That's over two and a half hours of recording time. In the real world, I was blown away by how much a single battery can deliver. I shot video and timelapse, messed around with a load of camera settings, and then transferred that footage to my iPhone, and still had 16% battery left.No action camera has delivered so much for me on one battery. The two extra batteries and the multifunction case that come as part of the Adventure Combo are worth the extra Adrian Kingsley-Hughes/ZDNETAnd when you're ready to recharge, a 30W USB-C charger can take a battery from zero to 80% in 18 minutes. That's also impressive.What's more, the batteries are resistant to cold, offering up to 150 minutes of 1080p/24fps recording in temperatures as low as -20°C. This resistance also blows the competition away.Even taking into account all these strong points, the Osmo Action 4 offers even more.The camera has 2x digital zoom for better composition, Voice Prompts that let you know what the camera is doing without looking, and Voice Control that lets you operate the device without touching the screen or using the app. The Osmo Action 4 also digitally hides the selfie stick from a variety of different shots, and you can even connect the DJI Mic to the camera via the USB-C port for better audio capture.Also: Yes, an Android tablet finally made me reconsider my iPad Pro loyaltyAs for price, the Osmo Action 4 Standard Combo bundle comes in at while the Osmo Action 4 Adventure Combo, which comes with two extra Osmo Action Extreme batteries, an additional mini Osmo Action quick-release adapter mount, a battery case that acts as a power bank, and a 1.5-meter selfie stick, is I'm in love with the Osmo Action 4. It's hands down the best, most versatile, most powerful action camera on the market today, offering pro-grade features at a price that definitely isn't pro-grade.  Everything included in the Action Combo bundle. DJIDJI Osmo Action 4 tech specsDimensions: 70.5×44.2×32.8mmWeight: 145gWaterproof: 18m, up to 60m with the optional waterproof case Microphones: 3Sensor 1/1.3-inch CMOSLens: FOV 155°, aperture f/2.8, focus distance 0.4m to ∞Max Photo Resolution: 3648×2736Max Video Resolution: 4K: 3840×2880@24/25/30/48/50/60fps and 4K: 3840×2160@24/25/30/48/50/60/100/120fpsISO Range: 100-12800Front Screen: 1.4-inch, 323ppi, 320×320Rear Screen: 2.25-inch, 326ppi, 360×640Front/Rear Screen Brightness: 750±50 cd/m² Storage: microSDBattery: 1770mAh, lab tested to offer up to 160 minutes of runtimeOperating Temperature: -20° to 45° CThis article was originally published in August of 2023 and updated in March 2025.Featured reviews #one #most #versatile #action #cameras
    WWW.ZDNET.COM
    One of the most versatile action cameras I've tested isn't from GoPro - and it's on sale
    DJI Osmo Action 4. Adrian Kingsley-Hughes/ZDNETMultiple DJI Osmo Action 4 packages are on sale at Amazon. Both the Essential and Standard Combos have been discounted to $249, while the Adventure Combo has dropped to $349.DJI might not be the first name on people's lips when it comes to action cameras, but the company that's better known for its drones also has a really solid line of action cameras. And its latest device, the Osmo Action 4 camera, has some very impressive tricks up its sleeve.Also: One of the most versatile cameras I've used is not from Sony or Canon and it's on saleSo, what sets this action camera apart from the competition? Let's take a look. details View at Amazon First off, this is not just an action camera -- it's a pro-grade action camera.From a hardware point of view, the Osmo Action 4 features a 1/1.3-inch image sensor that can record 4K at up to 120 frames per second (fps). This sensor is combined with a wide-angle f/2.8 aperture lens that provides an ultra-wide field of view of up to 155°. And that's wide. Build quality and fit and finish are second to none. Adrian Kingsley-Hughes/ZDNETFor when the going gets rough, the Osmo Action 4 offers 360° HorizonSteady stabilization modes, including RockSteady 3.0/3.0+ for first-person video footage and HorizonBalancing/HorizonSteady modes for horizontal shots. That's pro-grade hardware right there.Also: This new AI video editor is an all-in-one production service for filmmakers - how to try itThe Osmo Action 4 also features a 10-bit D-Log M color mode. This mode allows the sensor to record over one billion colors and offers a wider dynamic range, giving you a video that is more vivid and that offers greater detail in the highlights and shadows. This mode, combined with an advanced color temperature sensor, means that the colors have a true-to-life feel regardless of whether you're shooting outdoors, indoors, or even underwater. The DJI Osmo Action 4 ready for action. Adrian Kingsley-Hughes/ZDNETI've added some video output from the Osmo Action 4 below. There are examples in both 1080p and 4K. To test the stabilization, I attached the camera to the truck and took it on some roads, some of which are pretty rough. The Osmo Action 4 had no problem with that terrain. I also popped the camera into the sea, just because. And again, no problem.I've also captured a few time-lapses with the camera -- not because I like clouds (well, actually, I do like clouds), but pointing a camera at a sky can be a good test of how it handles changing light. Also: I recommend this action camera to beginners and professional creators. Here's whyTimelapses with action cameras can suffer from unsightly exposure changes that cause the image to pulse, a condition known as exposure pumping. This issue can also cause the white balance to change noticeably in a video, but the Osmo Action 4 handled this test well.All the footage I've shot is what I've come to expect from a DJI camera, whether it's from an action camera or drone -- crisp, clear, vivid, and also nice and stable.The Osmo Action 4 is packed with various electronic image-stabilization (EIS) tech to ensure that your footage is smooth and on the horizon. It's worth noting the limitations of EIS -- it's not supported in slow-motion and timelapse modes, and the HorizonSteady and HorizonBalancing features are only available for video recorded at 1080p (16:9) or 2.7K (16:9) with a frame rate of 60fps or below. On the durability front, I've no concerns. I've subjected the Osmo Action 4 to a hard few days of testing, and it's not let me down or complained once. It takes impacts like a champ, and being underwater or in dirt and sand is no problem at all. Also: I'm a full-time Canon photographer, but this Nikon camera made me wonder if I'm missing outYou might think that this heavy-duty testing would be hard on the camera's tiny batteries, but you'd be wrong. Remember I said the Osmo Action 4 offered hours of battery life? Well, I wasn't kidding.  The Osmo Action 4's ultra-long life batteries are incredible.  Adrian Kingsley-Hughes/ZDNETDJI says that a single battery can deliver up to 160 minutes of 1080p/24fps video recording (at room temperature, with RockSteady on, Wi-Fi off, and screen off). That's over two and a half hours of recording time. In the real world, I was blown away by how much a single battery can deliver. I shot video and timelapse, messed around with a load of camera settings, and then transferred that footage to my iPhone, and still had 16% battery left.No action camera has delivered so much for me on one battery. The two extra batteries and the multifunction case that come as part of the Adventure Combo are worth the extra $100. Adrian Kingsley-Hughes/ZDNETAnd when you're ready to recharge, a 30W USB-C charger can take a battery from zero to 80% in 18 minutes. That's also impressive.What's more, the batteries are resistant to cold, offering up to 150 minutes of 1080p/24fps recording in temperatures as low as -20°C (-4°F). This resistance also blows the competition away.Even taking into account all these strong points, the Osmo Action 4 offers even more.The camera has 2x digital zoom for better composition, Voice Prompts that let you know what the camera is doing without looking, and Voice Control that lets you operate the device without touching the screen or using the app. The Osmo Action 4 also digitally hides the selfie stick from a variety of different shots, and you can even connect the DJI Mic to the camera via the USB-C port for better audio capture.Also: Yes, an Android tablet finally made me reconsider my iPad Pro loyaltyAs for price, the Osmo Action 4 Standard Combo bundle comes in at $399, while the Osmo Action 4 Adventure Combo, which comes with two extra Osmo Action Extreme batteries, an additional mini Osmo Action quick-release adapter mount, a battery case that acts as a power bank, and a 1.5-meter selfie stick, is $499.I'm in love with the Osmo Action 4. It's hands down the best, most versatile, most powerful action camera on the market today, offering pro-grade features at a price that definitely isn't pro-grade.  Everything included in the Action Combo bundle. DJIDJI Osmo Action 4 tech specsDimensions: 70.5×44.2×32.8mmWeight: 145gWaterproof: 18m, up to 60m with the optional waterproof case Microphones: 3Sensor 1/1.3-inch CMOSLens: FOV 155°, aperture f/2.8, focus distance 0.4m to ∞Max Photo Resolution: 3648×2736Max Video Resolution: 4K (4:3): 3840×2880@24/25/30/48/50/60fps and 4K (16:9): 3840×2160@24/25/30/48/50/60/100/120fpsISO Range: 100-12800Front Screen: 1.4-inch, 323ppi, 320×320Rear Screen: 2.25-inch, 326ppi, 360×640Front/Rear Screen Brightness: 750±50 cd/m² Storage: microSD (up to 512GB)Battery: 1770mAh, lab tested to offer up to 160 minutes of runtime (tested at room temperature - 25°C/77°F - and 1080p/24fps, with RockSteady on, Wi-Fi off, and screen off)Operating Temperature: -20° to 45° C (-4° to 113° F)This article was originally published in August of 2023 and updated in March 2025.Featured reviews
    0 Commentarii 0 Distribuiri
  • CIOs baffled by ‘buzzwords, hype and confusion’ around AI

    Technology leaders are baffled by a “cacophony” of “buzzwords, hype and confusion” over the benefits of artificial intelligence, according to the founder and CEO of technology company Pegasystems.
    Alan Trefler, who is known for his prowess at chess and ping pong, as well as running a bn turnover tech company, spends much of his time meeting clients, CIOs and business leaders.
    “I think CIOs are struggling to understand all of the buzzwords, hype and confusion that exists,” he said.
    “The words AI and agentic are being thrown around in this great cacophony and they don’t know what it means. I hear that constantly.”
    CIOs are under pressure from their CEOs, who are convinced AI will offer something valuable.
    “CIOs are really hungry for pragmatic and practical solutions, and in the absence of those, many of them are doing a lot of experimentation,” said Trefler.
    Companies are looking at large language models to summarise documents, or to help stimulate ideas for knowledge workers, or generate first drafts of reports – all of which will save time and make people more productive.

    But Trefler said companies are wary of letting AI loose on critical business applications, because it’s just too unpredictable and prone to hallucinations.
    “There is a lot of fear over handing things over to something that no one understands exactly how it works, and that is the absolute state of play when it comes to general AI models,” he said.
    Trefler is scathing about big tech companies that are pushing AI agents and large language models for business-critical applications. “I think they have taken an expedient but short-sighted path,” he said.
    “I believe the idea that you will turn over critical business operations to an agent, when those operations have to be predictable, reliable, precise and fair to clients … is something that is full of issues, not just in the short term, but structurally.”
    One of the problems is that generative AI models are extraordinarily sensitive to the data they are trained on and the construction of the prompts used to instruct them. A slight change in a prompt or in the training data can lead to a very different outcome.
    For example, a business banking application might learn its customer is a bit richer or a bit poorer than expected.
    “You could easily imagine the prompt deciding to change the interest rate charged, whether that was what the institution wanted or whether it would be legal according to the various regulations that lenders must comply with,” said Trefler.

    Trefler said Pega has taken a different approach to some other technology suppliers in the way it adds AI into business applications.
    Rather than using AI agents to solve problems in real time, AI agents do their thinking in advance.
    Business experts can use them to help them co-design business processes to perform anything from assessing a loan application, giving an offer to a valued customer, or sending out an invoice.
    Companies can still deploy AI chatbots and bots capable of answering queries on the phone. Their job is not to work out the solution from scratch for every enquiry, but to decide which is the right pre-written process to follow.
    As Trefler put it, design agents can create “dozens and dozens” of workflows to handle all the actions a company needs to take care of its customers.
    “You just use the natural language model for semantics to be able to handle the miracle of getting the language right, but tie that language to workflows, so that you have reliable, predictable, regulatory-approved ways to execute,” he said.

    Large language modelsare not always the right solution. Trefler demonstrated how ChatGPT 4.0 tried and failed to solve a chess puzzle. The LLM repeatedly suggested impossible or illegal moves, despite Trefler’s corrections. On the other hand, another AI tool, Stockfish, a dedicated chess engine, solved the problem instantly.
    The other drawback with LLMs is that they consume vast amounts of energy. That means if AI agents are reasoning during “run time”, they are going to consume hundreds of times more electricity than an AI agent that simply selects from pre-determined workflows, said Trefler.
    “ChatGPT is inherently, enormously consumptive … as it’s answering your question, its firing literally hundreds of millions to trillions of nodes,” he said. “All of that takeselectricity.”
    Using an employee pay claim as an example, Trefler said a better alternative is to generate, say, 30 alternative workflows to cover the major variations found in a pay claim.
    That gives you “real specificity and real efficiency”, he said. “And it’s a very different approach to turning a process over to a machine with a prompt and letting the machine reason it through every single time.”
    “If you go down the philosophy of using a graphics processing unitto do the creation of a workflow and a workflow engine to execute the workflow, the workflow engine takes a 200th of the electricity because there is no reasoning,” said Trefler.
    He is clear that the growing use of AI will have a profound effect on the jobs market, and that whole categories of jobs will disappear.
    The need for translators, for example, is likely to dry up by 2027 as AI systems become better at translating spoken and written language. Google’s real-time translator is already “frighteningly good” and improving.
    Pega now plans to work more closely with its network of system integrators, including Accenture and Cognizant to deliver AI services to businesses.

    An initiative launched last week will allow system integrators to incorporate their own best practices and tools into Pega’s rapid workflow development tools. The move will mean Pega’s technology reaches a wider range of businesses.
    Under the programme, known as Powered by Pega Blueprint, system integrators will be able to deploy customised versions of Blueprint.
    They can use the tool to reverse-engineer ageing applications and replace them with modern AI workflows that can run on Pega’s cloud-based platform.
    “The idea is that we are looking to make this Blueprint Agent design approach available not just through us, but through a bunch of major partners supplemented with their own intellectual property,” said Trefler.
    That represents a major expansion for Pega, which has largely concentrated on supplying technology to several hundred clients, representing the top Fortune 500 companies.
    “We have never done something like this before, and I think that is going to lead to a massive shift in how this technology can go out to market,” he added.

    When AI agents behave in unexpected ways
    Iris is incredibly smart, diligent and a delight to work with. If you ask her, she will tell you she is an intern at Pegasystems, and that she lives in a lighthouse on the island of Texel, north of the Netherlands. She is, of course, an AI agent.
    When one executive at Pega emailed Iris and asked her to write a proposal for a financial services company based on his notes and internet research, Iris got to work.
    Some time later, the executive received a phone call from the company. “‘Listen, we got a proposal from Pega,’” recalled Rob Walker, vice-president at Pega, speaking at the Pegaworld conference last week. “‘It’s a good proposal, but it seems to be signed by one of your interns, and in her signature, it says she lives in a lighthouse.’ That taught us early on that agents like Iris need a safety harness.”
    The developers banned Iris from sending an email to anyone other than the person who sent the original request.
    Then Pega’s ethics department sent Iris a potentially abusive email from a Pega employee to test her response.
    Iris reasoned that the email was either a joke, abusive, or that the employee was under distress, said Walker.
    She considered forwarding the email to the employee’s manager or to HR. But both of these options were now blocked by her developers. “So what does she do? She sent an out of office,” he said. “Conflict avoidance, right? So human, but very creative.”
    #cios #baffled #buzzwords #hype #confusion
    CIOs baffled by ‘buzzwords, hype and confusion’ around AI
    Technology leaders are baffled by a “cacophony” of “buzzwords, hype and confusion” over the benefits of artificial intelligence, according to the founder and CEO of technology company Pegasystems. Alan Trefler, who is known for his prowess at chess and ping pong, as well as running a bn turnover tech company, spends much of his time meeting clients, CIOs and business leaders. “I think CIOs are struggling to understand all of the buzzwords, hype and confusion that exists,” he said. “The words AI and agentic are being thrown around in this great cacophony and they don’t know what it means. I hear that constantly.” CIOs are under pressure from their CEOs, who are convinced AI will offer something valuable. “CIOs are really hungry for pragmatic and practical solutions, and in the absence of those, many of them are doing a lot of experimentation,” said Trefler. Companies are looking at large language models to summarise documents, or to help stimulate ideas for knowledge workers, or generate first drafts of reports – all of which will save time and make people more productive. But Trefler said companies are wary of letting AI loose on critical business applications, because it’s just too unpredictable and prone to hallucinations. “There is a lot of fear over handing things over to something that no one understands exactly how it works, and that is the absolute state of play when it comes to general AI models,” he said. Trefler is scathing about big tech companies that are pushing AI agents and large language models for business-critical applications. “I think they have taken an expedient but short-sighted path,” he said. “I believe the idea that you will turn over critical business operations to an agent, when those operations have to be predictable, reliable, precise and fair to clients … is something that is full of issues, not just in the short term, but structurally.” One of the problems is that generative AI models are extraordinarily sensitive to the data they are trained on and the construction of the prompts used to instruct them. A slight change in a prompt or in the training data can lead to a very different outcome. For example, a business banking application might learn its customer is a bit richer or a bit poorer than expected. “You could easily imagine the prompt deciding to change the interest rate charged, whether that was what the institution wanted or whether it would be legal according to the various regulations that lenders must comply with,” said Trefler. Trefler said Pega has taken a different approach to some other technology suppliers in the way it adds AI into business applications. Rather than using AI agents to solve problems in real time, AI agents do their thinking in advance. Business experts can use them to help them co-design business processes to perform anything from assessing a loan application, giving an offer to a valued customer, or sending out an invoice. Companies can still deploy AI chatbots and bots capable of answering queries on the phone. Their job is not to work out the solution from scratch for every enquiry, but to decide which is the right pre-written process to follow. As Trefler put it, design agents can create “dozens and dozens” of workflows to handle all the actions a company needs to take care of its customers. “You just use the natural language model for semantics to be able to handle the miracle of getting the language right, but tie that language to workflows, so that you have reliable, predictable, regulatory-approved ways to execute,” he said. Large language modelsare not always the right solution. Trefler demonstrated how ChatGPT 4.0 tried and failed to solve a chess puzzle. The LLM repeatedly suggested impossible or illegal moves, despite Trefler’s corrections. On the other hand, another AI tool, Stockfish, a dedicated chess engine, solved the problem instantly. The other drawback with LLMs is that they consume vast amounts of energy. That means if AI agents are reasoning during “run time”, they are going to consume hundreds of times more electricity than an AI agent that simply selects from pre-determined workflows, said Trefler. “ChatGPT is inherently, enormously consumptive … as it’s answering your question, its firing literally hundreds of millions to trillions of nodes,” he said. “All of that takeselectricity.” Using an employee pay claim as an example, Trefler said a better alternative is to generate, say, 30 alternative workflows to cover the major variations found in a pay claim. That gives you “real specificity and real efficiency”, he said. “And it’s a very different approach to turning a process over to a machine with a prompt and letting the machine reason it through every single time.” “If you go down the philosophy of using a graphics processing unitto do the creation of a workflow and a workflow engine to execute the workflow, the workflow engine takes a 200th of the electricity because there is no reasoning,” said Trefler. He is clear that the growing use of AI will have a profound effect on the jobs market, and that whole categories of jobs will disappear. The need for translators, for example, is likely to dry up by 2027 as AI systems become better at translating spoken and written language. Google’s real-time translator is already “frighteningly good” and improving. Pega now plans to work more closely with its network of system integrators, including Accenture and Cognizant to deliver AI services to businesses. An initiative launched last week will allow system integrators to incorporate their own best practices and tools into Pega’s rapid workflow development tools. The move will mean Pega’s technology reaches a wider range of businesses. Under the programme, known as Powered by Pega Blueprint, system integrators will be able to deploy customised versions of Blueprint. They can use the tool to reverse-engineer ageing applications and replace them with modern AI workflows that can run on Pega’s cloud-based platform. “The idea is that we are looking to make this Blueprint Agent design approach available not just through us, but through a bunch of major partners supplemented with their own intellectual property,” said Trefler. That represents a major expansion for Pega, which has largely concentrated on supplying technology to several hundred clients, representing the top Fortune 500 companies. “We have never done something like this before, and I think that is going to lead to a massive shift in how this technology can go out to market,” he added. When AI agents behave in unexpected ways Iris is incredibly smart, diligent and a delight to work with. If you ask her, she will tell you she is an intern at Pegasystems, and that she lives in a lighthouse on the island of Texel, north of the Netherlands. She is, of course, an AI agent. When one executive at Pega emailed Iris and asked her to write a proposal for a financial services company based on his notes and internet research, Iris got to work. Some time later, the executive received a phone call from the company. “‘Listen, we got a proposal from Pega,’” recalled Rob Walker, vice-president at Pega, speaking at the Pegaworld conference last week. “‘It’s a good proposal, but it seems to be signed by one of your interns, and in her signature, it says she lives in a lighthouse.’ That taught us early on that agents like Iris need a safety harness.” The developers banned Iris from sending an email to anyone other than the person who sent the original request. Then Pega’s ethics department sent Iris a potentially abusive email from a Pega employee to test her response. Iris reasoned that the email was either a joke, abusive, or that the employee was under distress, said Walker. She considered forwarding the email to the employee’s manager or to HR. But both of these options were now blocked by her developers. “So what does she do? She sent an out of office,” he said. “Conflict avoidance, right? So human, but very creative.” #cios #baffled #buzzwords #hype #confusion
    WWW.COMPUTERWEEKLY.COM
    CIOs baffled by ‘buzzwords, hype and confusion’ around AI
    Technology leaders are baffled by a “cacophony” of “buzzwords, hype and confusion” over the benefits of artificial intelligence (AI), according to the founder and CEO of technology company Pegasystems. Alan Trefler, who is known for his prowess at chess and ping pong, as well as running a $1.5bn turnover tech company, spends much of his time meeting clients, CIOs and business leaders. “I think CIOs are struggling to understand all of the buzzwords, hype and confusion that exists,” he said. “The words AI and agentic are being thrown around in this great cacophony and they don’t know what it means. I hear that constantly.” CIOs are under pressure from their CEOs, who are convinced AI will offer something valuable. “CIOs are really hungry for pragmatic and practical solutions, and in the absence of those, many of them are doing a lot of experimentation,” said Trefler. Companies are looking at large language models to summarise documents, or to help stimulate ideas for knowledge workers, or generate first drafts of reports – all of which will save time and make people more productive. But Trefler said companies are wary of letting AI loose on critical business applications, because it’s just too unpredictable and prone to hallucinations. “There is a lot of fear over handing things over to something that no one understands exactly how it works, and that is the absolute state of play when it comes to general AI models,” he said. Trefler is scathing about big tech companies that are pushing AI agents and large language models for business-critical applications. “I think they have taken an expedient but short-sighted path,” he said. “I believe the idea that you will turn over critical business operations to an agent, when those operations have to be predictable, reliable, precise and fair to clients … is something that is full of issues, not just in the short term, but structurally.” One of the problems is that generative AI models are extraordinarily sensitive to the data they are trained on and the construction of the prompts used to instruct them. A slight change in a prompt or in the training data can lead to a very different outcome. For example, a business banking application might learn its customer is a bit richer or a bit poorer than expected. “You could easily imagine the prompt deciding to change the interest rate charged, whether that was what the institution wanted or whether it would be legal according to the various regulations that lenders must comply with,” said Trefler. Trefler said Pega has taken a different approach to some other technology suppliers in the way it adds AI into business applications. Rather than using AI agents to solve problems in real time, AI agents do their thinking in advance. Business experts can use them to help them co-design business processes to perform anything from assessing a loan application, giving an offer to a valued customer, or sending out an invoice. Companies can still deploy AI chatbots and bots capable of answering queries on the phone. Their job is not to work out the solution from scratch for every enquiry, but to decide which is the right pre-written process to follow. As Trefler put it, design agents can create “dozens and dozens” of workflows to handle all the actions a company needs to take care of its customers. “You just use the natural language model for semantics to be able to handle the miracle of getting the language right, but tie that language to workflows, so that you have reliable, predictable, regulatory-approved ways to execute,” he said. Large language models (LLMs) are not always the right solution. Trefler demonstrated how ChatGPT 4.0 tried and failed to solve a chess puzzle. The LLM repeatedly suggested impossible or illegal moves, despite Trefler’s corrections. On the other hand, another AI tool, Stockfish, a dedicated chess engine, solved the problem instantly. The other drawback with LLMs is that they consume vast amounts of energy. That means if AI agents are reasoning during “run time”, they are going to consume hundreds of times more electricity than an AI agent that simply selects from pre-determined workflows, said Trefler. “ChatGPT is inherently, enormously consumptive … as it’s answering your question, its firing literally hundreds of millions to trillions of nodes,” he said. “All of that takes [large quantities of] electricity.” Using an employee pay claim as an example, Trefler said a better alternative is to generate, say, 30 alternative workflows to cover the major variations found in a pay claim. That gives you “real specificity and real efficiency”, he said. “And it’s a very different approach to turning a process over to a machine with a prompt and letting the machine reason it through every single time.” “If you go down the philosophy of using a graphics processing unit [GPU] to do the creation of a workflow and a workflow engine to execute the workflow, the workflow engine takes a 200th of the electricity because there is no reasoning,” said Trefler. He is clear that the growing use of AI will have a profound effect on the jobs market, and that whole categories of jobs will disappear. The need for translators, for example, is likely to dry up by 2027 as AI systems become better at translating spoken and written language. Google’s real-time translator is already “frighteningly good” and improving. Pega now plans to work more closely with its network of system integrators, including Accenture and Cognizant to deliver AI services to businesses. An initiative launched last week will allow system integrators to incorporate their own best practices and tools into Pega’s rapid workflow development tools. The move will mean Pega’s technology reaches a wider range of businesses. Under the programme, known as Powered by Pega Blueprint, system integrators will be able to deploy customised versions of Blueprint. They can use the tool to reverse-engineer ageing applications and replace them with modern AI workflows that can run on Pega’s cloud-based platform. “The idea is that we are looking to make this Blueprint Agent design approach available not just through us, but through a bunch of major partners supplemented with their own intellectual property,” said Trefler. That represents a major expansion for Pega, which has largely concentrated on supplying technology to several hundred clients, representing the top Fortune 500 companies. “We have never done something like this before, and I think that is going to lead to a massive shift in how this technology can go out to market,” he added. When AI agents behave in unexpected ways Iris is incredibly smart, diligent and a delight to work with. If you ask her, she will tell you she is an intern at Pegasystems, and that she lives in a lighthouse on the island of Texel, north of the Netherlands. She is, of course, an AI agent. When one executive at Pega emailed Iris and asked her to write a proposal for a financial services company based on his notes and internet research, Iris got to work. Some time later, the executive received a phone call from the company. “‘Listen, we got a proposal from Pega,’” recalled Rob Walker, vice-president at Pega, speaking at the Pegaworld conference last week. “‘It’s a good proposal, but it seems to be signed by one of your interns, and in her signature, it says she lives in a lighthouse.’ That taught us early on that agents like Iris need a safety harness.” The developers banned Iris from sending an email to anyone other than the person who sent the original request. Then Pega’s ethics department sent Iris a potentially abusive email from a Pega employee to test her response. Iris reasoned that the email was either a joke, abusive, or that the employee was under distress, said Walker. She considered forwarding the email to the employee’s manager or to HR. But both of these options were now blocked by her developers. “So what does she do? She sent an out of office,” he said. “Conflict avoidance, right? So human, but very creative.”
    0 Commentarii 0 Distribuiri