AI is rotting your brain and making you stupid
For nearly 10 years I have written about science and technology and I’ve been an early adopter of new tech for much longer. As a teenager in the mid-1990s I annoyed the hell out of my family by jamming up the phone line for hours with a dial-up modem; connecting to bulletin board communities all over the country.When I started writing professionally about technology in 2016 I was all for our seemingly inevitable transhumanist future. When the chip is ready I want it immediately stuck in my head, I remember saying proudly in our busy office. Why not improve ourselves where we can?Since then, my general view on technology has dramatically shifted. Watching a growing class of super-billionaires erode the democratizing nature of technology by maintaining corporate controls over what we use and how we use it has fundamentally changed my personal relationship with technology. Seeing deeply disturbing philosophical stances like longtermism, effective altruism, and singulartarianism envelop the minds of those rich, powerful men controlling the world has only further entrenched inequality.A recent Black Mirror episode really rammed home the perils we face by having technology so controlled by capitalist interests. A sick woman is given a brain implant connected to a cloud server to keep her alive. The system is managed through a subscription service where the user pays for monthly access to the cognitive abilities managed by the implant. As time passes, that subscription cost gets more and more expensive - and well, it’s Black Mirror, so you can imagine where things end up.
Titled 'Common People', the episode is from series 7 of Black MirrorNetflix
The enshittification of our digital world has been impossible to ignore. You’re not imagining things, Google Search is getting worse.But until the emergence of AII’ve never been truly concerned about a technological innovation, in and of itself.A recent article looked at how generative AI tech such as ChatGPT is being used by university students. The piece was authored by a tech admin at New York University and it’s filled with striking insights into how AI is shaking the foundations of educational institutions.Not unsurprisingly, students are using ChatGPT for everything from summarizing complex texts to completely writing essays from scratch. But one of the reflections quoted in the article immediately jumped out at me.When a student was asked why they relied on generative AI so much when putting work together they responded, “You’re asking me to go from point A to point B, why wouldn’t I use a car to get there?”My first response was, of course, why wouldn’t you? It made complete sense.For a second.And then I thought, hang on, what is being lost by speeding from point A to point B in a car?
What if the quickest way from point A to point B wasn't the best way to get there?Depositphotos
Let’s further the analogy. You need to go to the grocery store. It’s a 10-minute walk away but a three-minute drive. Why wouldn’t you drive?Well, the only benefit of driving is saving time. That’s inarguable. You’ll be back home and cooking up your dinner before the person on foot even gets to the grocery store.Congratulations. You saved yourself about 20 minutes. In a world where efficiency trumps everything this is the best choice. Use that extra 20 minutes in your day wisely.But what are the benefits of not driving, taking the extra time, and walking?First, you have environmental benefits. Not using a car unnecessarily; spewing emissions into the air, either directly from combustion or indirectly for those with electric cars.Secondly, you have health benefits from the little bit of exercise you get by walking. Our stationary lives are quite literally killing us so a 20-minute walk a day is likely to be incredibly positive for your health.But there are also more abstract benefits to be gained by walking this short trip from A to B.Walking connects us to our neighborhood. It slows things down. Helps us better understand the community and environment we are living in. A recent study summarized the benefits of walking around your neighborhood, suggesting the practice leads to greater social connectedness and reduced feelings of isolation.So what are we losing when we use a car to get from point A to point B? Potentially a great deal.But let’s move out of abstraction and into the real world.An article in the Columbia Journalism Review asked nearly 20 news media professionals how they were integrating AI into their personal workflow. The responses were wildly varied. Some journalists refused to use AI for anything more than superficial interview transcription, while others use it broadly, to edit text, answer research questions, summarize large bodies of science text, or search massive troves of data for salient bits of information.In general, the line almost all those media professionals shared was they would never explicitly use AI to write their articles. But for some, almost every other stage of the creative process in developing a story was fair game for AI assistance.I found this a little horrifying. Farming out certain creative development processes to AI felt not only ethically wrong but also like key cognitive stages were being lost, skipped over, considered unimportant.I’ve never considered myself to be an extraordinarily creative person. I don’t feel like I come up with new or original ideas when I work. Instead, I see myself more as a compiler. I enjoy finding connections between seemingly disparate things. Linking ideas and using those pieces as building blocks to create my own work. As a writer and journalist I see this process as the whole point.A good example of this is a story I published in late 2023 investigating the relationship between long Covid and psychedelics. The story began earlier in the year when I read an intriguing study linking long Covid with serotonin abnormalities in the gut. Being interested in the science of psychedelics, and knowing that psychedelics very much influence serotonin receptors, I wondered if there could be some kind of link between these two seemingly disparate topics.The idea sat in the back of my mind for several months, until I came across a person who told me they had been actively treating their own long Covid symptoms with a variety of psychedelic remedies. After an expansive and fascinating interview I started diving into different studies looking to understand how certain psychedelics affect the body, and whether there could be any associations with long Covid treatments.Eventually I stumbled across a few compelling associations. It took weeks of reading different scientific studies, speaking to various researchers, and thinking about how several discordant threads could be somehow linked.Could AI have assisted me in the process of developing this story?No. Because ultimately, the story comprised an assortment of novel associations that I drew between disparate ideas all encapsulated within the frame of a person’s subjective experience.And it is this idea of novelty that is key to understanding why modern AI technology is not actually intelligence but a simulation of intelligence.
LLMs are a sophisticated language imitator, delivering responses that resemble what they think a response would look likeDepositphotos
ChatGPT, and all the assorted clones that have emerged over the last couple of years, are a form of technology called LLMs. At the risk of enraging those who actually work in this mind-bendingly complex field, I’m going to dangerously over-simplify how these things work.It’s important to know that when you ask a system like ChatGPT a question it doesn’t understand what you are asking it. The response these systems generate to any prompt is simply a simulation of what it computes a response would look like based on a massive dataset.So if I were to ask the system a random question like, “What color are cats?”, the system would scrape the world’s trove of information on cats and colors to create a response that mirrors the way most pre-existing text talks about cats and colors. The system builds its response word by word, creating something that reads coherently to us, by establishing a probability for what word should follow each prior word. It’s not thinking, it’s imitating.What these generative AI systems are spitting out are word salad amalgams of what it thinks the response to your prompt should look like, based on training from millions of books and webpages that have been previously published.Setting aside for a moment the accuracy of the responses these systems deliver, I am more interestedwith the cognitive stages that this technology allows us to skip past.For thousands of years we have used technology to improve our ability to manage highly complex tasks. The idea is called cognitive offloading, and it’s as simple as writing something down on a notepad or saving a contact number on your smartphone. There are pros and cons to cognitive offloading, and scientists have been digging into the phenomenon for years.As long as we have been doing it, there have been people criticizing the practice. The legendary Greek philosopher Socrates was notorious for his skepticism around the written word. He believed knowledge emerged through a dialectical process so writing itself was reductive. He even went so far as to suggestthat writing makes us dumber.
“For this invention will produce forgetfulness in the minds of those who learn to use it, because they will not practice their memory. Their trust in writing, produced by external characters which are no part of themselves, will discourage the use of their own memory within them. You have invented an elixir not of memory, but of reminding; and you offer your pupils the appearance of wisdom, not true wisdom, for they will read many things without instruction and will therefore seem to know many things, when they are for the most part ignorant and hard to get along with, since they are not wise, but only appear wise.”
Wrote Plato, quoting Socrates
Almost every technological advancement in human history can be seen to be accompanied by someone suggesting it will be damaging. Calculators have destroyed our ability to properly do math. GPS has corrupted our spatial memory. Typewriters killed handwriting. Computer word processors killed typewriters. Video killed the radio star.And what have we lost? Well, zooming in on writing, for example, a 2020 study claimed brain activity is greater when a note is handwritten as opposed to being typed on a keyboard. And then a 2021 study suggested memory retention is better when using a pen and paper versus a stylus and tablet. So there are certainly trade-offs whenever we choose to use a technological tool to offload a cognitive task.There’s an oft-told story about gonzo journalist Hunter S. Thompson. It may be apocryphal but it certainly is meaningful. He once said he sat down and typed out the entirety of The Great Gatsby, word for word. According to Thompson, he wanted to know what it felt like to write a great novel.
Thompson was infamous for writing everything on typewriters, even when computers emerged in the 1990sPublic Domain
I don’t want to get all wishy-washy here, but these are the brass tacks we are ultimately falling on. What does it feel like to think? What does it feel like to be creative? What does it feel like to understand something?A recent interview with Satya Nadella, CEO of Microsoft, reveals how deeply AI has infiltrated his life and work. Not only does Nadella utilize nearly a dozen different custom-designed AI agents to manage every part of his workflow – from summarizing emails to managing his schedule – but he also uses AI to get through podcasts quickly on his way to work. Instead of actually listening to the podcasts he has transcripts uploaded to an AI assistant who he then chats to about the information while commuting.Why listen to the podcast when you can get the gist through a summary? Why read a book when you can listen to the audio version at X2 speed? Or better yet, watch the movie? Or just read a Wikipedia entry. Or get AI to summarize the wikipedia entry.I’m not here to judge anyone on the way they choose to use technology. Do what you want with ChatGPT. But for a moment consider what you may be skipping over by racing from point A to point B.Sure, you can give ChatGPT a set of increasingly detailed prompts; adding complexity to its summary of a scientific journal or a podcast, but at what point do the prompts get so granular that you may as well read the journal entry itself? If you get generative AI to skim and summarize something, what is it missing? If something was worth being written then surely it is worth being read?If there is a more succinct way to say something then maybe we should say it more succinctly.In a magnificent article for The New Yorker, Ted Chiang perfectly summed up the deep contradiction at the heart of modern generative AI systems. He argues language, and writing, is fundamentally about communication. If we write an email to someone we can expect the person at the other end to receive those words and consider them with some kind of thought or attention. But modern AI systemsare erasing our ability to think, consider, and write. Where does it all end? For Chiang it's pretty dystopian feedback loop of dialectical slop.
“We are entering an era where someone might use a large language model to generate a document out of a bulleted list, and send it to a person who will use a large language model to condense that document into a bulleted list. Can anyone seriously argue that this is an improvement?”
Ted Chiang
#rotting #your #brain #making #youAI is rotting your brain and making you stupid
For nearly 10 years I have written about science and technology and I’ve been an early adopter of new tech for much longer. As a teenager in the mid-1990s I annoyed the hell out of my family by jamming up the phone line for hours with a dial-up modem; connecting to bulletin board communities all over the country.When I started writing professionally about technology in 2016 I was all for our seemingly inevitable transhumanist future. When the chip is ready I want it immediately stuck in my head, I remember saying proudly in our busy office. Why not improve ourselves where we can?Since then, my general view on technology has dramatically shifted. Watching a growing class of super-billionaires erode the democratizing nature of technology by maintaining corporate controls over what we use and how we use it has fundamentally changed my personal relationship with technology. Seeing deeply disturbing philosophical stances like longtermism, effective altruism, and singulartarianism envelop the minds of those rich, powerful men controlling the world has only further entrenched inequality.A recent Black Mirror episode really rammed home the perils we face by having technology so controlled by capitalist interests. A sick woman is given a brain implant connected to a cloud server to keep her alive. The system is managed through a subscription service where the user pays for monthly access to the cognitive abilities managed by the implant. As time passes, that subscription cost gets more and more expensive - and well, it’s Black Mirror, so you can imagine where things end up.
Titled 'Common People', the episode is from series 7 of Black MirrorNetflix
The enshittification of our digital world has been impossible to ignore. You’re not imagining things, Google Search is getting worse.But until the emergence of AII’ve never been truly concerned about a technological innovation, in and of itself.A recent article looked at how generative AI tech such as ChatGPT is being used by university students. The piece was authored by a tech admin at New York University and it’s filled with striking insights into how AI is shaking the foundations of educational institutions.Not unsurprisingly, students are using ChatGPT for everything from summarizing complex texts to completely writing essays from scratch. But one of the reflections quoted in the article immediately jumped out at me.When a student was asked why they relied on generative AI so much when putting work together they responded, “You’re asking me to go from point A to point B, why wouldn’t I use a car to get there?”My first response was, of course, why wouldn’t you? It made complete sense.For a second.And then I thought, hang on, what is being lost by speeding from point A to point B in a car?
What if the quickest way from point A to point B wasn't the best way to get there?Depositphotos
Let’s further the analogy. You need to go to the grocery store. It’s a 10-minute walk away but a three-minute drive. Why wouldn’t you drive?Well, the only benefit of driving is saving time. That’s inarguable. You’ll be back home and cooking up your dinner before the person on foot even gets to the grocery store.Congratulations. You saved yourself about 20 minutes. In a world where efficiency trumps everything this is the best choice. Use that extra 20 minutes in your day wisely.But what are the benefits of not driving, taking the extra time, and walking?First, you have environmental benefits. Not using a car unnecessarily; spewing emissions into the air, either directly from combustion or indirectly for those with electric cars.Secondly, you have health benefits from the little bit of exercise you get by walking. Our stationary lives are quite literally killing us so a 20-minute walk a day is likely to be incredibly positive for your health.But there are also more abstract benefits to be gained by walking this short trip from A to B.Walking connects us to our neighborhood. It slows things down. Helps us better understand the community and environment we are living in. A recent study summarized the benefits of walking around your neighborhood, suggesting the practice leads to greater social connectedness and reduced feelings of isolation.So what are we losing when we use a car to get from point A to point B? Potentially a great deal.But let’s move out of abstraction and into the real world.An article in the Columbia Journalism Review asked nearly 20 news media professionals how they were integrating AI into their personal workflow. The responses were wildly varied. Some journalists refused to use AI for anything more than superficial interview transcription, while others use it broadly, to edit text, answer research questions, summarize large bodies of science text, or search massive troves of data for salient bits of information.In general, the line almost all those media professionals shared was they would never explicitly use AI to write their articles. But for some, almost every other stage of the creative process in developing a story was fair game for AI assistance.I found this a little horrifying. Farming out certain creative development processes to AI felt not only ethically wrong but also like key cognitive stages were being lost, skipped over, considered unimportant.I’ve never considered myself to be an extraordinarily creative person. I don’t feel like I come up with new or original ideas when I work. Instead, I see myself more as a compiler. I enjoy finding connections between seemingly disparate things. Linking ideas and using those pieces as building blocks to create my own work. As a writer and journalist I see this process as the whole point.A good example of this is a story I published in late 2023 investigating the relationship between long Covid and psychedelics. The story began earlier in the year when I read an intriguing study linking long Covid with serotonin abnormalities in the gut. Being interested in the science of psychedelics, and knowing that psychedelics very much influence serotonin receptors, I wondered if there could be some kind of link between these two seemingly disparate topics.The idea sat in the back of my mind for several months, until I came across a person who told me they had been actively treating their own long Covid symptoms with a variety of psychedelic remedies. After an expansive and fascinating interview I started diving into different studies looking to understand how certain psychedelics affect the body, and whether there could be any associations with long Covid treatments.Eventually I stumbled across a few compelling associations. It took weeks of reading different scientific studies, speaking to various researchers, and thinking about how several discordant threads could be somehow linked.Could AI have assisted me in the process of developing this story?No. Because ultimately, the story comprised an assortment of novel associations that I drew between disparate ideas all encapsulated within the frame of a person’s subjective experience.And it is this idea of novelty that is key to understanding why modern AI technology is not actually intelligence but a simulation of intelligence.
LLMs are a sophisticated language imitator, delivering responses that resemble what they think a response would look likeDepositphotos
ChatGPT, and all the assorted clones that have emerged over the last couple of years, are a form of technology called LLMs. At the risk of enraging those who actually work in this mind-bendingly complex field, I’m going to dangerously over-simplify how these things work.It’s important to know that when you ask a system like ChatGPT a question it doesn’t understand what you are asking it. The response these systems generate to any prompt is simply a simulation of what it computes a response would look like based on a massive dataset.So if I were to ask the system a random question like, “What color are cats?”, the system would scrape the world’s trove of information on cats and colors to create a response that mirrors the way most pre-existing text talks about cats and colors. The system builds its response word by word, creating something that reads coherently to us, by establishing a probability for what word should follow each prior word. It’s not thinking, it’s imitating.What these generative AI systems are spitting out are word salad amalgams of what it thinks the response to your prompt should look like, based on training from millions of books and webpages that have been previously published.Setting aside for a moment the accuracy of the responses these systems deliver, I am more interestedwith the cognitive stages that this technology allows us to skip past.For thousands of years we have used technology to improve our ability to manage highly complex tasks. The idea is called cognitive offloading, and it’s as simple as writing something down on a notepad or saving a contact number on your smartphone. There are pros and cons to cognitive offloading, and scientists have been digging into the phenomenon for years.As long as we have been doing it, there have been people criticizing the practice. The legendary Greek philosopher Socrates was notorious for his skepticism around the written word. He believed knowledge emerged through a dialectical process so writing itself was reductive. He even went so far as to suggestthat writing makes us dumber.
“For this invention will produce forgetfulness in the minds of those who learn to use it, because they will not practice their memory. Their trust in writing, produced by external characters which are no part of themselves, will discourage the use of their own memory within them. You have invented an elixir not of memory, but of reminding; and you offer your pupils the appearance of wisdom, not true wisdom, for they will read many things without instruction and will therefore seem to know many things, when they are for the most part ignorant and hard to get along with, since they are not wise, but only appear wise.”
Wrote Plato, quoting Socrates
Almost every technological advancement in human history can be seen to be accompanied by someone suggesting it will be damaging. Calculators have destroyed our ability to properly do math. GPS has corrupted our spatial memory. Typewriters killed handwriting. Computer word processors killed typewriters. Video killed the radio star.And what have we lost? Well, zooming in on writing, for example, a 2020 study claimed brain activity is greater when a note is handwritten as opposed to being typed on a keyboard. And then a 2021 study suggested memory retention is better when using a pen and paper versus a stylus and tablet. So there are certainly trade-offs whenever we choose to use a technological tool to offload a cognitive task.There’s an oft-told story about gonzo journalist Hunter S. Thompson. It may be apocryphal but it certainly is meaningful. He once said he sat down and typed out the entirety of The Great Gatsby, word for word. According to Thompson, he wanted to know what it felt like to write a great novel.
Thompson was infamous for writing everything on typewriters, even when computers emerged in the 1990sPublic Domain
I don’t want to get all wishy-washy here, but these are the brass tacks we are ultimately falling on. What does it feel like to think? What does it feel like to be creative? What does it feel like to understand something?A recent interview with Satya Nadella, CEO of Microsoft, reveals how deeply AI has infiltrated his life and work. Not only does Nadella utilize nearly a dozen different custom-designed AI agents to manage every part of his workflow – from summarizing emails to managing his schedule – but he also uses AI to get through podcasts quickly on his way to work. Instead of actually listening to the podcasts he has transcripts uploaded to an AI assistant who he then chats to about the information while commuting.Why listen to the podcast when you can get the gist through a summary? Why read a book when you can listen to the audio version at X2 speed? Or better yet, watch the movie? Or just read a Wikipedia entry. Or get AI to summarize the wikipedia entry.I’m not here to judge anyone on the way they choose to use technology. Do what you want with ChatGPT. But for a moment consider what you may be skipping over by racing from point A to point B.Sure, you can give ChatGPT a set of increasingly detailed prompts; adding complexity to its summary of a scientific journal or a podcast, but at what point do the prompts get so granular that you may as well read the journal entry itself? If you get generative AI to skim and summarize something, what is it missing? If something was worth being written then surely it is worth being read?If there is a more succinct way to say something then maybe we should say it more succinctly.In a magnificent article for The New Yorker, Ted Chiang perfectly summed up the deep contradiction at the heart of modern generative AI systems. He argues language, and writing, is fundamentally about communication. If we write an email to someone we can expect the person at the other end to receive those words and consider them with some kind of thought or attention. But modern AI systemsare erasing our ability to think, consider, and write. Where does it all end? For Chiang it's pretty dystopian feedback loop of dialectical slop.
“We are entering an era where someone might use a large language model to generate a document out of a bulleted list, and send it to a person who will use a large language model to condense that document into a bulleted list. Can anyone seriously argue that this is an improvement?”
Ted Chiang
#rotting #your #brain #making #you
English
Arabic
French
Spanish
Portuguese
Deutsch
Turkish
Dutch
Italiano
Russian
Romaian
Portuguese (Brazil)
Greek
