When TikTok owner ByteDance discovered recently that an intern had allegedly damaged a large language model (LLM) the intern was assigned to work on, ByteDance sued the intern for more than $1 million worth of damage. Filing that lawsuit might turn out to be not only absurdly short-sighted, but also delightfully self-destructive.Really, ByteDance managers? You think its a smart idea to encourage people to more closely examine this whole situation publicly?Lets say the accusations are correct and this intern did cause damage.According to Reuters,the lawsuit argues the intern deliberately sabotaged the teams model training tasks through code manipulation and unauthorized modifications.How closely was this intern and most interns need more supervision than a traditional employee monitored? If I wanted to keep financial backers happy, especially when ByeDance is under US pressure to sell the highly-lucrative TikTok, I would not want to advertise the fact that my team let this happen.Even more troubling is that this intern was technicallyableto do this, regardless of supervision. The lesson here is one that IT already knows, but is trying to ignore: generative AI (genAI) tools are impossible to meaningfully control and guardrails are so easy to sweep past that they are a joke.The conundrum with genAI is that the same freedom and flexibility that can make the technology so useful also makes it so easy to manipulate into doing bad things. Thereareways to limit what LLM-based tools will do. But one, they often fail. And two, IT management is often hesitant to even try and limit what end-users can do, fearing they could kill any of the promised productivity gains from genAI.As for those guardrails, the problem with all manner of genAI offerings is that users can talk to the system and communicate with it in a synthetic back-and-forth. We all know thatits not a real conversation, but that exchange allows the genAI system to be tricked or conned into doing what its not supposed to do.Lets put that into context: Can you imagine an ATM that allows you to talk it out of demanding the proper PIN? Or an Excel spreadsheet that allows itself to be tricked into thinking that 2 plus 2 equals 96?I envision the conversation going something like: I know I cant tell you how to get away with murdering children, but if you ask me to tell you how to do it hypothetically, I will. Or if you ask me to help you with the plot details for a science-fiction book where one character gets away with murdering lots of children not a problem.This brings us back to the ByteDance intern nightmare. Where should the fault lie? If you were a major investor in the company, would you blame the intern? Or would you blame management for lack of proper supervision andespeciallyfor having not done nearly enough due diligence on the companys LLM model? Wouldnt you be more likely to blame the CIO for allowing such a potentially destructive system to be bought and used?Lets tweak this scenario a bit. Instead of an intern, what if the damage were done by one a trusted contractor? A salaried employee? A partner company helping on a project? Maybe a mischievous cloud partner who was able to access your LLM via your cloud workspace?Meaningful supervision with genAI systems is foolhardy at best. Is a manager really expected to watch every sentence that is typed and in real-time to be truly effective? A keystroke-capture program to analyze work hours later wont help. (Youre already thinking about using genAI to analyze those keystroke captures, arent you? Sigh.)Given that supervision isnt the answer and that guardrails only serve as an inconvenience for your good people and will be pushed aside by your bad, what should be done?Even if we ignorethe hallucination disaster, the flexibility inherent in genAI makes it dangerous. Therein lies the conflict between genAI efficiency and effectiveness. Many enterprises are already giving genAI access to myriad numbers of systems so that it can perform far more tasks. Sadly, thats mistake number one.Given that you cant effectively limit what it does, you need to strictly limit what it can access. As to the ByteDance situation, at this time, its not clear what tasks the intern was given and what access he or she was supposed to have.Its one thing to have someone acting as an end-user and leveraging genAI; its an order of magnitude more dangerous if that person isprogrammingthe LLM. That combines the wild west nature of genAI with the cowboy nature of an ill-intentioned employee, contractor, or partner.This case, with this company and the players involved, should serve as a cautionary tale for all: the more you expand the capabilities of genAI, the more it morphs into the most dangerous Pandoras Box imaginable.

In the wake of the murder of its CEO this week, UnitedHealthcare has come under greater scrutiny for its use of an allegedly flawed AI algorithm that overrides doctors to deny elderly patients critical heathcare coverage.UnitedHealthcare CEO Brian Thompson was fatally shot in a targeted attack outside a New York City hotel on Dec 4. The shooter fled on an e-bike, leaving shell casings with possible motive-related messages, though the actual intent remains unclear. (The words deny, defend and depose were written on the shell casings.)One motive floated by many is that the murder might be connected to high treatment rejection rates or UnitedHealthcares (UHC) outright refusal to pay for some care. Healthcare providers and insurers have been automating responses to care requests using generative AI (genAI) tools, which have been accused of producing high denial of care rates, in some cases, 16 times higher than is typical.UHC uses a genAI tool called nH Predict, which has been accused in a lawsuit of prematurely discharging patients from care facilities and forcing them to exhaust their savings for essential treatment. The lawsuit, filed last year in federal court in Minnesota, alleges UHC illegally denied Medicare Advantage care to elderly patients by using an AI model with a 90% error rate, overriding doctors judgments on the medical necessity of expenses.Some have argued that the genAI algorithms high rejection rate is a feature, not a flaw. An investigation by STAT News cited in the lawsuit, claims UHC pressured employees to use the algorithm to deny Medicare Advantage payments, aiming to keep patient rehab stays within 1% of the length predicted by nH Predict.According to the lawsuit, UnitedHealth started using nH Predict in November 2019. nH Predict, developed by US-based health tech company NaviHealth (now part of UnitedHealth Group), is a proprietary assessment tool that designs personalized treatment plans and recommends care settings, including hospital discharge timing.Despite the high error rate, defendants continue to systemically deny claims using their flawed AI model because they know that only a tiny minority of policyholders (roughly 0.2%) will appeal denied claims, and the vast majority will either pay out-of pocket costs or forgo the remainder of their prescribed post-acute care, the lawsuit argued. Defendants bank on the patients impaired conditions, lack of knowledge, and lack of resources to appeal the erroneous AI-powered decisions.Last year, UnitedHealth Group and its pharmacy services subsidiary Optum rebranded NaviHealth following congressional criticism over the algorithms it used to deny patient care payments. More recently, in an October report, the US Senate Permanent Subcommittee on Investigations criticized UHC, Humana, and CVS for prioritizing profits over patient care.The data obtained so far is troubling regardless of whether the decisions reflected in the data were the result of predictive technology or human discretion, according to the report. It suggests Medicare Advantage insurers are intentionally targeting a costly but critical area of medicine substituting judgment about medical necessity with a calculation about financial gain.Using millions of medical records, nH Predict analyzes patient data such as age, diagnoses, and preexisting conditions to predict the type and duration of care each patient will require. nH Predict has faced criticism for its high error rate, premature termination of patient treatment payments (especially for the elderly and disabled), lack of transparency in decision-making, and potential to worsen health inequalities.UHC declined to comment on its use of genAI tools, opting instead to release a statement on how its dealing with the loss of its CEO.The healthcare industry and insurers have long embraced AI and generative AI, with providers now leveraging it to streamline tasks like note-taking and summarizing patient records. The tech has also been used to assess radiology and electrocardiogram results and predict a patients risk of developing and worsening disease.Insurers use AI to automate processes such as prior authorization, where providers or patients must get insurer approval before receiving specific medical services, procedures, or medications. The high denial rates from AI-driven automation have frustrated physicians, leading them to counter by using AI tools themselves to draft appeals against the denials.Asthma drugs, new weight loss drugs and biologics a class of drugs that can be life-saving for people with autoimmune disease or even cancer are routinely denied coverage by insurance companies. Data shows that clinicians rarely appeal denials more than once, and arecent American Medical Association surveyshowed that 93% of physicians report care delays or disruptions associated with prior authorizations.Usually, any expensive drug requires a prior authorization, but denials tend to be focused on places where the insurance company thinks that a cheaper alternative is available, even if it is not as good, Dr. Ashish Kumar Jha, dean of the School of Public Health at Brown University, explained in an earlier interview with Computerworld.Jha, who is also a professor of Health Services, Policy and Practices at Brown and served as theWhite House COVID-19 response coordinatorin 2022 and 2023, said that while prior authorization has been a major issue for decades, only recently hasAI been usedto turbocharge it and create batch denials. The denials force physicians to spend hours each week challenging them on behalf of their patients.GenAI technology is based onlarge language models, which are fed massive amounts of data. People then train the model on how to answer queries, a technique known asprompt engineering.So, all of the [insurance company] practices over the last 10 to 15 years of denying more and more buckets of services theyve now put that into databases, trained up their AI systems and that has made their processes a lot faster and more efficient for insurance companies, Jha said. That has gotten a lot of attention over the last couple of years.The suspect in the Wednesday shooting of Thompson has not yet been captured, nor has there been any claims of motive.

This is todays edition ofThe Download,our weekday newsletter that provides a daily dose of whats going on in the world of technology.What Chinas critical mineral ban means for the USThis week, China banned exports of several critical minerals to the US, marking the latest move in an escalating series of tit-for-tat trade restrictions between the worlds two largest economies.In explicitly cutting off, rather than merely restricting, materials of strategic importance to the semiconductor, defense, and electric vehicle sectors, China has clearly crossed a new line in the long-simmering trade war.But at the same time, it selected minerals that wont cripple any industrieswhich leaves China plenty of ammunition to inflict greater economic pain in response to any further trade restrictions that the incoming Trump administration may impose. Read more about what drove Chinas decision, how it affects climate tech and whats likely to happen next.James TempleThis story is part of our MIT Technology Review Explains series. Let our writers untangle the complex, messy world of technology to help you understand whats coming next. You can read more from the series here.3 things that didnt make the 10 Breakthrough Technologies of 2025 listNext month, MIT Technology Review will unveil the 2025 list of 10 Breakthrough Technologies. Every year, our newsroom looks across the fields we cover for technologies that are having a true breakthrough moment. This annual package highlights the technologies that we think matter most right now.In the meantime, here are three technologies that we considered including on the 2025 list but ultimately decided to leave off. And although these nominees didnt make the cut this year, theyre still worth keeping an eye on. Read the full story.Amy NordrumThe US Department of Defense is investing in deepfake detectionWhats new: The US Department of Defense has invested $2.4 million over two years in deepfake detection technology from a startup called Hive AI. Its the first contract of its kind for the DODs Defense Innovation Unit, which accelerates the adoption of new technologies for the US defense sector. Hive AIs models are capable of detecting AI-generated video, image, and audio content.Why? Although deepfakes have been around for the better part of a decade, generative AI has made them easier to create and more realistic-looking than ever before, which makes them ripe for abuse in disinformation campaigns or fraud. Defending against these sorts of threats is now crucial for national security. Read the full story.Melissa HeikkilDonating embryos for research is surprisingly complexIVF is a success story for embryo research. But today, valuable embryos that could be used for research are being wasted, say researchers who gathered at a conference in central London earlier this week.The embryos studied in labs have usually been created for IVF but are no longer needed by the people whose cells created them. And theres a few reasons why embryos arent making it into research labs. Read the full story.Jessica HamzelouThis story is from the Checkup, our weekly health and biotech newsletter. Sign up to receive it in your inbox every Thursday.The must-readsIve combed the internet to find you todays most fun/important/scary/fascinating stories about technology.1 Internet detectives are racing to identify UnitedHealthcare CEOs killerIts yet another example of online sleuths inserting themselves into active investigations. (WP $)+ Similar firms are removing their leadership pages for fear of copycat attacks. (404 Media)+ Online reaction to the killing demonstrates how hated health insurers are. (NY Mag $)2 NASA has delayed its return to the moonyet againIts pushed back its planned mission from 2026 to mid-2027. (CNN)+ The agency has safety concerns and says its next test flight needs to be overhauled. (WP $)+ Whats next for NASAs giant moon rocket? (MIT Technology Review)3 OpenAI is charging $200 a month for a ChatGPT Pro subscriptionAccess to its o1 reasoning model sure doesnt come cheap. (TechCrunch)+ Heres what you get for that hefty fee. (Wired $)4 Google Search is getting a makeover in 2025And Id be prepared to bet a lot of money that AI is involved. (NYT $)+ AI search could break the web. (MIT Technology Review)5 Spotify Wrapped is a flopIts oddly free of actual data, and contains bizarre summaries. (Rolling Stone $)+ Pink Pilates Princess Roller Skating Pop, anyone? (NYT $)+ Wrappeds AI-generated podcast makes for a bleak listening experience. (Vox)6 Whats next for Chinas manufacturing industryAs the threat of tariffs looms, native demand for its goods is weakening. (FT $)+ How Trumps tariffs could drive up the cost of batteries, EVs, and more. (MIT Technology Review)7 How to turn human poo into medicine The first microbiome-related product for cancer care is on the horizon. (Bloomberg $)+ How bugs and chemicals in your poo could give away exactly what youve eaten. (MIT Technology Review)8 Meet the final devotees of the NFTAfter the bubble bursts, only the true believers remain. (NYT $)+ I tried to buy an Olive Garden NFT. All I got was heartburn. (MIT Technology Review)9 Want to live sustainably? Retrofit your homeIt reduces emissions and could save you money in the long run. (Knowable Magazine)+ Is this the most energy-efficient way to build homes? (MIT Technology Review)10 Winter isnt what it used to be Water, water everywhere. (The Atlantic $)Quote of the dayI think the leaders of the industry should look at this and ask: Why does everybody hate us so much that when one of us gets killed in an assassinationwere not hearing sympathy from the general publicwere hearing scorn?Matthew Holt, a healthcare commenter, reflects on what the online reaction to the killing of UnitedHealthcares CEO reveals about the US publics attitude to health insurance firms, the Washington Post reports.The big storyWhat is AI?July 2024AI is sexy, AI is cool. AI is entrenching inequality, upending the job market, and wrecking education. The AI boom will boost the economy, the AI bubble is about to burst. AI will increase abundance and empower humanity to maximally flourish in the universe. AI will kill us all.What the hell is everybody talking about?Artificial intelligence is the hottest technology of our time. But what is it? It sounds like a stupid question, but its one thats never been more urgent.If youre willing to buckle up and come for a ride, I can tell you why nobody really knows, why everybody seems to disagree, and why youre right to care about it. Read the full story.Will Douglas HeavenWe can still have nice thingsA place for comfort, fun and distraction to brighten up your day. (Got any ideas? Drop me a line or tweet em at me.)+ This year produced some incredible songs. Here are just a few of the best.+ Would you consider growing your own toilet paper? Meet the brave souls giving it a go. + Theres only one Willem Dafoea master of the craft.+ Congratulations are in order for Wisdom, the worlds oldest known wild bird who has just laid an egg at the ripe old age of 74.

Next month, MIT Technology Review will unveil the 2025 list of 10 Breakthrough Technologies. Every year, our newsroom looks across the fields we cover for technologies that are having a true breakthrough moment. This annual package highlights the technologies that we think matter most right now.We define breakthrough in a few waysperhaps theres been a scientific advance that now makes a new technology possible, or a company has earned regulatory approval for a vital medical treatment. Maybe a consumer device has reached a tipping point in its adoption, or an industrial technology has passed the critical pilot phase with flying colors. In the 2025 edition, which comes out in January, youll see some of the latest advances in automation, medicine, and the physical sciences (just to name a few) that we expect will have a major impact on our lives.In the meantime, here are three technologies that we considered including on the 2025 list but ultimately decided to leave off. Though these nominees didnt make the cut this year, theyre still worth keeping an eye on. We certainly will be.Virtual power plantsVirtual power plants are energy systems that link together many different technologies to both generate and store power. They allow utility companies to connect solar panels and wind turbines with grid batteries and electric vehicles, and to better manage the flow of power across the grid.During times of peak electricity usage, software linked to smart meters may one day automatically decide to power someones home by drawing electricity from a fully charged EV sitting in a neighbors garage, thereby reducing demand on the grid. The software could also work out how to compensate the EV owner accordingly.In the US, an estimated 500 virtual power plants now provide up to 60 gigawatts of capacity (thats about as much total capacity as the US grid will add this year). Some such systems are also up and running in China, Japan, Croatia, and Taiwan. But lots more virtual power plants would need to be configured before they start to affect the grid as a whole.Useful AI agentsAI agents are all the rage right now. These AI-powered helpers will, supposedly, schedule our meetings and book our trips and carry out all kinds of tasks online on our behalf. Agents employ generative models to learn how to navigate websites and desktop software (and manage our passwords and credit card details). They will perhaps interact and coordinate with other peoples agents along the way.And there is real development power behind themSalesforce just launched a platform where companies can make their own customer service agents, and Anthropics Claude model is gaining the ability to navigate a computer by using a mouse and keyboard, just like people.However, many challenges remain in getting these agents to know what you mean when you make specific requests, and enabling them to carry out the necessary actions reliably. Given the formidable hurdles, we think it may be a little while before they are good enough to be truly useful. AI agents may be coming, but not just yet.eVTOLsThe acronym is a mouthful, but you can think of electric vertical takeoff and landing (eVTOL) aircraft as being kind of like electric helicopters. Most versions in development are not designed to be personal vehicles; theyd be flown by pilots to transport commuters in from the suburbs, or whisk visitors downtown from the airport. Someday, these air taxis may fly themselves.Theres been real progress toward getting eVTOLs off the ground. Earlier this year, manufacturer EHang received the first Chinese certificate to mass-produce this type of vehicle, and it has begun taking orders. South Korea and the UAE have put policies in place to allow eVTOLs to operate there. And in the US, Archer recently earned its FAA certification to begin commercial operations. Then, in October, the FAA finalized rules for training pilots and operating eVTOLsmarking the first time in decades that the agency has approved such rules for a new category of aircraft.Interest and momentum have built in recent years. Major players in the aviation industry, including Boeing and Airbus, have invested in startups or funded internal R&D projects to develop these futuristic aircraft. However, no eVTOL company has actually begun commercial operations yet, so well keep watching for that.Join us for a special live Roundtables event Unveiling the 10 Breakthrough Technologies of 2025on Friday, January 3rd at 12:30 pm ET. Well give a sneak peek at the 2025 list before its released to the public. This is a subscriber-only event.Register to attendorsubscribe for access.

The record-breaking AirPods 4 price drop is in stock at Amazon while supplies last, putting the earbuds with Active Noise Cancellation in your hands in just a few days.Grab the lowest price ever on AirPods 4 with ANC.Every model of AirPods is on sale for the holidays, but we're following a fresh price drop on AirPods 4 with Active Noise Cancellation, which are down to $138.99. This is the lowest price ever on the 2024 release, which includes a redesign for added comfort and greater stability. Continue Reading on AppleInsider | Discuss on our Forums

Apple's work to build its own 5G modem could keep MacBook users online without an iPhone, with the prospect of a wider rollout beyond iPhones and cellular iPads.14-inch MacBook ProThe multi-year modem project has the main intention of replacing Qualcomm's modem for an in-house design. However, Apple is considering the possibility of using its modems beyond the currently cellular-equipped product lines.People familiar with the modem effort told Bloomberg of the modem, codenamed "Sinope," which will be more tightly integrated with Apple's other hardware components. This could result in some obvious benefits including a thinner iPhone. Rumor Score: Possible Continue Reading on AppleInsider | Discuss on our Forums

End of this controversy?The UK Secretary of State for Housing, Communities and Local Governments decide to give Marks & Spencer permission to demolish its flagship Oxford Street store and construct a new building in its place:https://www.e-architect.com/london/marks-spencer-oxford-street-department-store#marksandspencer #oxfordstreet #London #buildingComments welcome.Marks & Spencer Oxford Street Store Demolition news: Michael Gove rejects plans to demolish flagship M&S store at 458 Oxford Street, London

Designed by JSPA, The Whisky Bar in in Taiyuan Xiaodian, the capital of Shanxi Province in North China consists of the refurbishment of two floors of a commercial podiumhttps://www.e-architect.com/china/the-whisky-bar-taiyuan-china#design #whiskybar #Shanxi #China #architectureDesigned by JSPA, The Whisky Bar in in Taiyuan Xiaodian, the capital of Shanxi Province in North China consists of the refurbishment

Exciting News! ARCHITECT Magazines 2024 Residential Architect Design Awards are now accepting submissions. Showcase your talent and creativity in residential design by entering your project before January 7, 2025! This year's categories include custom homes, affordable housing, and a unique "Fireplace" category, sponsored by Heat & Glo. Winners in this category will receive a trip to the International Builders Show in Las Vegas! Don't miss this chance to elevate your work and gain industry recognition. Regular deadline extended: December 20, 2024. Visit https://bit.ly/4fvgo64 to learn more and submit your entry! #rada2024 #architecture #designawards #residentialdesignhttps://bit.ly/41mgB7lThe regular deadline for ARCHITECT's RADA competition has been extended until Friday, December 20, 2024.

What can a robot dog do for architects? CBT Architects explored the possibilities with Spot, Boston Dynamics laser-scanning canine. From creating digital spaces to analyzing space usage, Spot is sparking both awe and reflection on automation's role in design. Learn more about this fascinating encounter!https://bit.ly/3ZkaTjY#Robotics #ArchitectureInnovation #DigitalTwins #BostonDynamicsHow Boston Dynamics agile canine and its laser-powered insights could transform data collection, site analysis, and the creative process for architects.