We've reached the end of another week, so it's time for a roundup of the best deals we could find on tech we like. Happily, a couple of the items we saw at CES (and named among the best) are already seeing decent discounts. So those who want to smoke meat indoors or invest in home back up batteries are in luck. As for smaller gadgets, we saw a deal on our favorite Bluetooth tracker, found a sale on Anker batteries and chargers and noted that Apple's AirPods Pro (which will work with the upcoming Vision Pro) are still hovering at their all-time low price. This also may be the last week to take advantage of pre-order promotions on the soon-to-be-released Samsung Galaxy S24 smartphones. Here are the best tech deals from this week that you can still get today.

Follow @EngadgetDeals on Twitter and subscribe to the Engadget Deals newsletter for the latest tech deals and buying advice.

George Carlin’s estate has filed a lawsuit against the makers of an hour-long comedy special featuring an AI replica of the comedian, as reported by NBC News. The late comedian’s estate, including his daughter Kelly Carlin, filed the suit in a Los Angeles federal court last night. It claims the online media company that posted the video, Dudesy, violated the performer’s right to publicity and infringed on a copyright.

The video’s called “George Carlin: I'm Glad I'm Dead” and features an hour of new “material” by the comedian, who died in 2008. As AI replications go, it’s certainly not going to break any records. It’s audio only and, honestly, doesn’t even sound that much like Carlin. It sounds like a below average impression of the comedian. Also, it’s very, very bad. Carlin had an extremely unique voice and this video is mostly basic punchlines you can see coming from a mile away. There’s very little outlandish wordplay. There’s no righteous fury. There are, however, a lot of jokes comparing Donald Trump to poop.

"I understand and share the desire for more George Carlin. I, too, want more time with my father. But it is ridiculous to proclaim he has been ‘resurrected’ with AI,” Kelly Carlin wrote in a statement. She went on to write that the Carlin in that video is a “poorly-executed facsimile cobbled together by unscrupulous individuals.”

The estate's attorney, Josh Schiller, went on to warn that AI risked becoming "a tool that allows bad-faith actors to replace creative expression, to exploit the already existing work of creators, and to get rich at the expense of others."

Dudesey, the channel that created and posted the video, is actually run by the popular comedian Will Sasso and author Chad Kultgen. They didn’t write the material here. The AI was trained on thousands of hours of Carlin routines to create the facsimile, according to a report by NPR. Sasso and Kultgen are, however, named in the suit. The pair behind Dudesy liken the AI-created Carlin to an impressionist who impersonates a public figure.

Sasso suggested in a podcast last week that the AI version was no replacement for the real thing, going on to say that it was “interesting how heated people get about it.” The lawsuit calls the video a “piece of computer-generated click-bait which detracts from the value of Carlin’s comedic works and harms his reputation.”

The complaint seeks unspecified damages and the immediate removal of “any video or audio copies” of the hour-long special. So, if you’re curious to hear a pretty bad Carlin impression make obvious jokes about Taylor Swift, you had better get on that while you have the chance.

Of course, this is just the latest salvo in the ongoing war between AI algorithms and humans that create works of value. This issue was at the very heart of last year’s Hollywood writers’ strike and the recent spate of AI-created celebrities used to scam consumers. This is just the beginning. It’s an election year, after all, and bad actors have already used an AI replication of President Biden’s voice to urge New Hampshire residents not to vote in last week’s primary.

Swedish electric car company Polestar is slashing its workforce by 15 percent globally. About 450 employees are expected to be let go due to “challenging market conditions.” The news comes despite its six percent increase in global car deliveries compared to 2022, according to its recent fourth quarter global fiscal report.

The company did, however, warn that it would reduce its headcount back in May 2023 which was around the same time it announced its production goals were disappointingly off by 10,000 to 20,000 cars from its initial goal. Polestar defended its decisions and explained it was “intensifying its focus” on cutting costs to make the business more efficient.

Despite delays in shipments last year, the 2024 Polestar 2 lineup is coming in strong with a suite of new upgrades, including longer mileage and faster charging. However, the company is faced with the issue that buyers might be turned off by its nearly $50,000 price tag when they can get newer models produced by rivals like Tesla for more than $10,000 less.

Job cuts across the EV sector have become commonplace, with rivals like Lucid Motors’ announcement to cut 18 percent of its workforce last year and Rivian slashing six percent. These trends might be due to the fact that supply chain issues are a huge problem in the EV industry, coupled with buyer hesitancy to invest in electric cars.

Last year, the European Union implemented new laws to make big tech open up its platforms to competitors. The deadline for compliance is March, and all eyes were on how Apple, which is famous for not playing nicely with others, would react. Now the company has set out how it will comply with the law, and the result is the sort of malicious compliance everyone was expecting. Similarly, the reaction from the coalition of well-heeled critics who were all hoping to get a slice of Apple’s pie for free has been similarly predictable.

The Digital Markets Act

In 2023, the EU laid down a new regime to prevent big tech throwing all of its weight around in the bloc. The Digital Markets Act and Digital Services Act govern what it calls “gatekeepers,” the big platforms who get between users and businesses. That includes Meta, Alphabet, Apple, Amazon and (TikTok owner) ByteDance, who all have big user bases, deep pockets and a lot of power. One key provision of the law was to get platform holders like Apple and Google to open their systems and allow competing services, such as alternative app stores, a topic we covered in depth back in 2020.

On January 25, Apple published a statement explaining how the DMA would impact iOS, Safari and the App Store. The document is laced with references to how the law makes iOS less secure and that Apple needs to take steps to mitigate those risks. And while Apple does not say how much each part of its business makes specifically, the App Store is a key part of its services division which earned a combined $22 billion in its most recent quarter. Consequently, Apple will happily let you set up a competing iOS app store, but in order to do so, you will have to vault Mount Everest, dig a tunnel to the center of the Earth and front a million dollars in cash.

Okay, not quite that.

You can compete, but you won’t want to

The creators of a would-be rival app store can’t simply turn up and sell their wares without any oversight. It was obvious from the get-go that even if Apple did open up its platforms, no third party app store would be allowed to do an end-run around the company’s basic rules. If you were hoping to run Honest Doug’s App Store (Not A Scam) and take the world for a ride, then you’re out of luck.

Would-be rivals will still need to meet Apple’s Notarization requirements and have tight rules and moderation tools governing quality, piracy, fraud and payment disputes. (Notarization will mean these apps will be checked by Apple to look for “known malware”, with the ability to shut the app down if any is detected.) They will need key rules around data collection and to offer users the same level of control they enjoy in the App Store proper. Not to mention complying with the Digital Services Act, GDPR and a number of other acronym-heavy EU regulations around digital services and online privacy. Essentially, if you want to run your own App Store, you’ll need to do it to the same level that Apple does.

Apple has also said app stores need to ensure they can meet their obligation to pay app developers. In this case, it means sharing a letter from a top financial institution with proof they have access to a minimum of €1,000,000 (around $1.1 million) in credit. And to avoid third party app stores taking advantage of Apple’s platform without Apple benefiting, developers will need to pay a Core Technology Fee once an app has been downloaded more than a million times. This is a per-install fee of €0.50 (around 54 cents) which renews every 12 months the app is installed for. You can decide for yourself if this reminds you of Unity’s aborted Runtime Fee payment scheme.

At the present time, Apple charges developers either $99 or $299, depending on if they are for an individual or a company. Apple then takes a flat commission on any transaction, either to buy the app itself or with an in-app purchase. For small developers making less than $1 million per year, Apple takes a 15 percent cut, while bigger names pay 30 percent. There are exceptions, including “reader” apps which are downloaded for free and tie to subscriptions elsewhere. So far it's not clear under what circumstances the sideloading fees might be preferable (if ever) to the vanilla "Apple tax" through its proprietary storefront.

The Expected Response

Naturally, Apple’s statement and all of the explanatory detail in its developer notes was controversial. Its critics, many of whom feel that Apple has too much power over its platform, were incensed.

Epic Games CEO Tim Sweeney, who has previously sued the company about this matter, was quick to denounce the changes. He said the new rules were “a devious new instance of malicious compliance.” Adding that it is forcing app developers to pick between App Store exclusivity or an “anticompetitive scheme rife with new junk fees on downloads and new Apple taxes on payments they don’t process.”

The Coalition for App Fairness, a lobby group backed by Epic, Spotify and Match Group, was quick to support one of its biggest backers. Executive director and former Republican spokesperson Rick Vanmeter said Apple had “no intention” to comply with the DMA. And added the move was a “shameless insult to the European Commission and the millions of European consumers they represent,” and urged officials to reject the move.

Despite Sweeney’s personal objection and that of his lobbyists, Epic Games has already said Fortnite – which was pulled from the Apple Store when Epic deliberately violated Apple’s Terms of Service – will return to iOS. The company said it would launch its own Epic Games Store for iOS in 2024, through which it would distribute its own titles. It added in the announcement tweet it would continue to “argue to the courts and regulators that Apple is breaking the law.”

But it’s not just Apple’s well-heeled rivals who feel the company is thumbing its nose at the EU with these changes. Andy Yen, the founder of privacy service Proton, told Engadget that Apple’s compliance with the DMA is “done in bad faith,” and that the iPhone maker is “fighting tooth and nail to maintain its profits and monopoly." Yen added that the “strings attached to Apple’s new policies mean that in practice it will be impossible for developers to benefit from them.” And that the moves erode “the fundamental rights of users by giving Apple the ability to review apps downloaded outside the App Store.” He added that the “European Commission can’t let this blatant bending of the rules fly.”

But despite the chorus of calls demanding the European Commission to Do Something, the body hasn’t budged just yet. “We take note of Apple’s announcements ahead of the compliance deadline,” a commission spokesperson told Engadget “We do not comment on these announcements.” The spokesperson added they “strongly encourage designated gatekeepers to test their proposals with third parties.” And that these comments were “without prejudice to the Commission’s own assessment of these proposals.”

At the time of writing, there has not yet been a comment from any high-profile EU figures about the matter. European Commission President Ursula von der Leyen and Margrethe Vestager, who handles technology and competition matters, have been active on social media but not about this topic. Similarly, we are waiting to hear back from Spotify and Deezer, who have both previously urged the European Union to act. Not to mention that, before Apple’s announcement, Spotify published its own announcement saying it will offer app downloads directly from its site.

The National Security Agency’s director has confirmed that the agency buys Americans’ web browsing data from brokers without first obtaining warrants. Senator Ron Wyden (D-OR) blocked the appointment of the NSA’s inbound director Timothy Haugh until the agency answered his questions regarding its collection of Americans’ location and Internet data. Wyden said he’d been trying for three years to “publicly release the fact that the NSA is purchasing Americans’ internet records.”

In a letter dated December 11, current NSA Director Paul Nakasone confirmed to Wyden that the agency does make such purchases from brokers. "NSA acquires various types of [commercially available information] for foreign intelligence, cybersecurity, and other authorized mission purposes, to include enhancing its signals intelligence (SIGINT) and cybersecurity missions," Nakasone wrote. "This may include information associated with electronic devices being used outside and, in certain cases, inside the United States."

Nakasone went on to claim that the NSA "does not buy and use location data collected from phones known to be used in the United States either with or without a court order. Similarly, NSA does not buy and use location data collected from automobile telematics systems from vehicles known to be located in the United States."

An NSA spokesperson told Reuters that the agency uses such data sparingly but that it has notable value for national security and cybersecurity purposes. "At all stages, NSA takes steps to minimize the collection of US [personal] information, to include application of technical filters," the spokesperson said.

Wyden has called the practice unlawful. "Such records can identify Americans who are seeking help from a suicide hotline or a hotline for survivors of sexual assault or domestic abuse," he said.

The senator urged Director of National Intelligence Avril Haines to order US intelligence agencies to stop buying Americans’ private data without consent. He also asked Haines to direct intelligence agencies to "conduct an inventory of the personal data purchased by the agency about Americans, including, but not limited to, location and internet metadata." Wyden said that any data that does not comply with Federal Trade Commission standards regarding personal data sales should be deleted.

Wyden pointed to an FTC settlement that this month banned a data broker from selling location data. The agency alleged that the information, which it claimed was sold to buyers including government contractors, "could be used to track people’s visits to sensitive locations such as medical and reproductive health clinics, places of religious worship and domestic abuse shelters."

The FTC stated in its complaint against the broker, formerly known as X-Mode Social, that by "failing to fully inform consumers how their data would be used and that their data would be provided to government contractors for national security purposes, X-Mode failed to provide information material to consumers and did not obtain informed consent from consumers to collect and use their location data."

The settlement was the first of its kind with a data broker. In a statement, Wyden, who has been investigating the data broker industry for several years, said he was "not aware of any company that provides such a warning to users [regarding their consent] before collecting their data."

The issue of US federal agencies buying phone location data isn't exactly new. In 2020, it emerged that Customs and Border Protection had been doing so. The following year, Wyden claimed the Defense Intelligence Agency and the Pentagon bought and used location data from Americans’ phones.

Tesla is recalling 200,000 vehicles in the US due to a malfunctioning backup camera. There were reports that the cameras wouldn’t engage when the cars were in reverse, which is a pretty big safety issue and the whole point of those cameras in the first place. Tesla has processed 81 warranty claims potentially related to the issue, according to Autoblog.

The recall includes certain Model Y, Model S, and Model X vehicles from 2023. Tesla says it delivered 1.8 million vehicles in 2023, so this recall accounts for more than 10 percent of the company’s yearly output. The US National Highway Traffic Safety Administration (NHTSA) released a statement on the matter and said that a software issue was to blame for the problem, according to Reuters.

To that end, all of the recalled vehicles feature Tesla’s “Full Self-Driving” computer 4.0 and run software version 2023.44.30 through 2023.44.30.6, or 2023.44.100. Tesla owners can check to see what software versions they’re running. The company has released an over-the-air (OTA) software update to fix the glitch, according to the NHTSA.

Tesla became aware of the problem in December and decided on a recall on January 12. Customers will receive a letter alerting them to the problem by March 22. The company says that it’s not aware of any crashes, injuries or deaths associated with the malfunction.

This latest recall comes just six weeks after Tesla recalled over two million vehicles after serious safety issues regarding its Autopilot advanced driver-assistance system. That was also addressed via an OTA software update.

A woman has a text chat with her long-dead lover. A family gets to hear a deceased elder speak again. A mother gets another chance to say goodbye to her child, who died suddenly, via a digital facsimile. This isn't a preview of the next season of Black Mirror — these are all true stories from the Sundance documentary Eternal You, a fascinating and frightening dive into tech companies using AI to digitally resurrect the dead.

It's yet another way modern AI, which includes large language models like ChatGPT and similar bespoke solutions, has the potential to transform society. And as Eternal You shows, the AI afterlife industry is already having a profound effect on its early users.

The film opens on a woman having a late night text chat with a friend: "I can't believe I'm trying this, how are you?" she asks, as if she's using the internet for the first time. "I'm okay. I'm working, I'm living. I'm... scared," her friend replies. When she asks why, they reply, "I'm not used to being dead."

Beetz Brothers Film Production

It turns out the woman, Christi Angel, is using the AI service Project December to chat with a simulation of her first love, who died many years ago. Angel is clearly intrigued by the technology, but as a devout Christian, she's also a bit spooked out by the prospect of raising the dead. The AI system eventually gives her some reasons to be concerned: Cameroun reveals that he's not in heaven, as she assumes. He's in hell.

"You're not in hell," she writes back. "I am in hell," the AI chatbot insists. The digital Cameroun says he's in a "dark and lonely" place, his only companions are "mostly addicts." The chatbot goes on to say he's currently haunting a treatment center and later suggests "I'll haunt you." That was enough to scare Angel and question why she was using this service in the first place.

While Angel was aware she was talking to a digital recreation of Cameroun, which was based on the information she provided to Project December, she interacted with the chatbot as if she was actually chatting with him on another plane of existence. That's a situation that many users of AI resurrection services will likely encounter: Rationality can easily overwhelm your emotional response while "speaking" with a dead loved one, even if the conversation is just occurring over text.

In the film, MIT sociologist Sherry Turkle suggests that our current understanding of how AI affects people is similar to our relationship with social media over a decade ago. That makes it a good time to ask questions about the human values and purposes it's serving, she says. If we had a clearer understanding of social media early on, maybe we could have pushed Facebook and Twitter to confront misinformation and online abuse more seriously. (Perhaps the 2016 election would have looked very different if we were aware of how other countries could weaponize social media.)

Beetz Brothers Film Production

Eternal You also introduces us to Joshua Barbeau, a freelance writer who became a bit of an online celebrity in 2021 when The San Francisco Chronicle reported on his Project December chatbot: a digital version of his ex-fiancee Jessica. At first, he used Project December to chat with pre-built bots, but he eventually realized he could use the underlying technology (GPT-3, at the time) to create one with Jessica's personality. Their conversations look natural and clearly comfort Barbeau. But we're still left wondering if chatting with a facsimile of his dead fiancee is actually helping Barbeau to process his grief. It could just as easily be seen as a crutch that he feels compelled to pay for.

It's also easy to be cynical about these tools, given what we see from their creators in the film. We meet Jason Rohrer, the founder and Project December and a former indie game designer, who comes across as a typical techno-libertarian.

"I believe in personal responsibility," he says, after also saying that he's not exactly in control of the AI models behind Project December, and right before we see him nearly crash a drone into his co-founders face. "I believe that consenting adults can use that technology however they want and they're responsible for the results of whatever they're doing. It's not my job as the creator of the technology to prevent the technology from being released, because I'm afraid of what somebody might do with it."

But, as MIT's Turkle points out, reanimating the dead via AI introduces moral questions that engineers like Rohrer likely aren't considering. "You're dealing with something much more profound in the human spirit," she says. "Once something is constituted enough that you can project onto it, this life force. It's our desire to animate the world, which is human, which is part of our beauty. But we have to worry about it, we have to keep it in check. Because I think it's leading us down a dangerous path."

Beetz Brothers Film Production

Another service, Hereafter.ai, lets users record stories to create a digital avatar of themselves, which family members can talk to now or after they die. One woman was eager to hear her father's voice again, but when she presented the avatar to her family the reaction was mixed. Younger folks seemed intrigue, but the older generation didn't want any part of it. "I fear that sometimes we can go too far with technology," her father's sister said. "I would just love to remember him as a person who was wonderful. I don't want my brother to appear to me. I'm satisfied knowing he's at peace, he's happy, and he's enjoying the other brothers, his mother and father."

YOV, an AI company that also focuses on personal avatars, or "Versonas," wants people to have seamless communication with their dead relatives across multiple channels. But, like all of these other digital afterlife companies, it runs into the same moral dilemmas. Is it ethical to digitally resurrect someone, especially if they didn't agree to it? Is the illusion of speaking to the dead more helpful or harmful for those left behind?

The most troubling sequence in Eternal You focuses on a South Korean mother, Jang Ji-sun, who lost her young child and remains wracked with guilt about not being able to say goodbye. She ended up being the central subject in a VR documentary, Meeting You, which was broadcast in South Korea in early 2020. She went far beyond a mere text chat: Jang donned a VR headset and confronted a startlingly realistic model of her child in virtual reality. The encounter was clearly moving for Jang, and the documentary received plenty of media attention at the time.

"There's a line between the world of the living and the world of the dead," said Kim Jong-woo, the producer behind Meeting You. "By line, I mean the fact that the dead can't come back to life. But people saw the experience as crossing that line. After all, I created an experience in which the beloved seemed to have returned. Have I made some huge mistake? Have I broken the principle of humankind? I don't know... maybe to some extent."

Eternal You paints a haunting portrait of an industry that's already revving up to capitalize on grief-stricken people. That's not exactly new; psychics and people claiming to speak to the dead have been around for our entire civilization. But through AI, we now have the ability to reanimate those lost souls. While that might be helpful for some, we're clearly not ready for a world where AI resurrection is commonplace.

If there’s one thing we can all agree upon, it’s that the 21st century’s captains of industry are trying to shoehorn AI into every corner of our world. But for all of the ways in which AI will be shoved into our faces and not prove very successful, it might actually have at least one useful purpose. For instance, by dramatically speeding up the often decades-long process of designing, finding and testing new drugs.

Risk mitigation isn’t a sexy notion but it’s worth understanding how common it is for a new drug project to fail. To set the scene, consider that each drug project takes between three and five years to form a hypothesis strong enough to start tests in a laboratory. A 2022 study from Professor Duxin Sun found that 90 percent of clinical drug development fails, with each project costing more than $2 billion. And that number doesn’t even include compounds found to be unworkable at the preclinical stage. Put simply, every successful drug has to prop up at least $18 billion waste generated by its unsuccessful siblings, which all but guarantees that less lucrative cures for rarer conditions aren’t given as much focus as they may need.

Dr. Nicola Richmond is VP of AI at Benevolent, a biotech company using AI in its drug discovery process. She explained the classical system tasks researchers to find, for example, a misbehaving protein – the cause of disease – and then find a molecule that could make it behave. Once they've found one, they need to get that molecule into a form a patient can take, and then test if it’s both safe and effective. The journey to clinical trials on a living human patient takes years, and it’s often only then researchers find out that what worked in theory does not work in practice.

The current process takes “more than a decade and multiple billions of dollars of research investment for every drug approved,” said Dr. Chris Gibson, co-founder of Recursion, another company in the AI drug discovery space. He says AI’s great skill may be to dodge the misses and help avoid researchers spending too long running down blind alleys. A software platform that can churn through hundreds of options at a time can, in Gibson’s words, “fail faster and earlier so you can move on to other targets.”

CellProfiler / Carpenter-Singh laboratory at the Broad Institute

Dr. Anne E. Carpenter is the founder of the Carpenter-Singh laboratory at the Broad Institute of MIT and Harvard. She has spent more than a decade developing techniques in Cell Painting, a way to highlight elements in cells, with dyes, to make them readable by a computer. She is also the co-developer of Cell Profiler, a platform enabling researchers to use AI to scrub through vast troves of images of those dyed cells. Combined, this work makes it easy for a machine to see how cells change when they are impacted by the presence of disease or a treatment. And by looking at every part of the cell holistically – a discipline known as “omics” – there are greater opportunities for making the sort of connections that AI systems excel at.

Using pictures as a way of identifying potential cures seems a little left-field, since how things look don’t always represent how things actually are, right? Carpenter said humans have always made subconscious assumptions about medical status from sight alone. She explained most people may conclude someone may have a chromosomal issue just by looking at their face. And professional clinicians can identify a number of disorders by sight alone purely as a consequence of their experience. She added that if you took a picture of everyone’s face in a given population, a computer would be able to identify patterns and sort them based on common features.

This logic applies to the pictures of cells, where it’s possible for a digital pathologist to compare images from healthy and diseased samples. If a human can do it, then it should be faster and easier to employ a computer to spot these differences in scale so long as it’s accurate. “You allow this data to self-assemble into groups and now [you’re] starting to see patterns,” she explained, “when we treat [cells] with 100,000 different compounds, one by one, we can say ‘here’s two chemicals that look really similar to each other.’” And this looking really similar to each other isn’t just coincidence, but seems to be indicative of how they behave.

In one example, Carpenter cited that two different compounds could produce similar effects in a cell, and by extension could be used to treat the same condition. If so, then it may be that one of the two – which may not have been intended for this purpose – has fewer harmful side effects. Then there’s the potential benefit of being able to identify something that we didn’t know was affected by disease. “It allows us to say, ‘hey, there’s this cluster of six genes, five of which are really well known to be part of this pathway, but the sixth one, we didn’t know what it did, but now we have a strong clue it’s involved in the same biological process.” “Maybe those other five genes, for whatever reason, aren’t great direct targets themselves, maybe the chemicals don’t bind,” she said, “but the sixth one [could be] really great for that.”

FatCamera via Getty Images

In this context, the startups using AI in their drug discovery processes are hoping that they can find the diamonds hiding in plain sight. Dr. Richmond said that Benevolent’s approach is for the team to pick a disease of interest and then formulate a biological question around it. So, at the start of one project, the team might wonder if there are ways to treat ALS by enhancing, or fixing, the way a cell’s own housekeeping system works. (To be clear, this is a purely hypothetical example supplied by Dr. Richmond.)

That question is then run through Benevolent’s AI models, which pull together data from a wide variety of sources. They then produce a ranked list of potential answers to the question, which can include novel compounds, or existing drugs that could be adapted to suit. The data then goes to a researcher, who can examine what, if any, weight to give to its findings. Dr. Richmond added that the model has to provide evidence from existing literature or sources to support its findings even if its picks are out of left-field. And that, at all times, a human has the final say on what of its results should be pursued and how vigorously.

It’s a similar situation at Recursion, with Dr. Gibson claiming that its model is now capable of predicting “how any drug will interact with any disease without having to physically test it.” The model has now formed around three trillion predictions connecting potential problems to their potential solutions based on the data it has already absorbed and simulated. Gibson said that the process at the company now resembles a web search: Researchers sit down at a terminal, “type in a gene associated with breast cancer and [the system] populates all the other genes and compounds that [it believes are] related.”

“What gets exciting,” said Dr. Gibson, “is when [we] see a gene nobody has ever heard of in the list, which feels like novel biology because the world has no idea it exists.” Once a target has been identified and the findings checked by a human, the data will be passed to Recursion’s in-house scientific laboratory. Here, researchers will run initial experiments to see if what was found in the simulation can be replicated in the real world. Dr. Gibson said that Recursion’s wet lab, which uses large-scale automation, is capable of running more than two million experiments in a working week.

“About six weeks later, with very little human intervention, we’ll get the results,” said Dr. Gibson and, if successful, it’s then the team will “really start investing.” Because, until this point, the short period of validation work has cost the company “very little money and time to get.” The promise is that, rather than a three-year preclinical phase, that whole process can be crunched down to a few database searches, some oversight and then a few weeks of ex vivo testing to confirm if the system’s hunches are worth making a real effort to interrogate. Dr. Gibson said that it believes it has taken a “year’s worth of animal model work and [compressed] it, in many cases, to two months.”

Of course, there is not yet a concrete success story, no wonder cure that any company in this space can point to as a validation of the approach. But Recursion can cite one real-world example of how close its platform came to matching the success of a critical study. In April 2020, Recursion ran the COVID-19 sequence through its system to look at potential treatments. It examined both FDA-approved drugs and candidates in late-stage clinical trials. The system produced a list of nine potential candidates which would need further analysis, eight of which it would later be proved to be correct. It also said that Hydroxychloroquine and Ivermectin, both much-ballyhooed in the earliest days of the pandemic, would flop.

And there are AI-informed drugs that are currently undergoing real-world clinical trials right now. Recursion is pointing to five projects currently finishing their stage one (tests in healthy patients), or entering stage two (trials in people with the rare diseases in question) clinical testing right now. Benevolent has started a stage one trial of BEN-8744, a treatment for ulcerative colitis that may help with other inflammatory bowel disorders. And BEN-8744 is targeting an inhibitor that has no prior associations in the existing research which, if successful, will add weight to the idea that AIs can spot the connections humans have missed. Of course, we can’t make any conclusions until at least early next year when the results of those initial tests will be released.

Yuichiro Chino via Getty Images

There are plenty of unanswered questions, including how much we should rely upon AI as the sole arbiter of the drug discovery pipeline. There are also questions around the quality of the training data and the biases in the wider sources more generally. Dr. Richmond highlighted the issues around biases in genetic data sources both in terms of the homogeneity of cell cultures and how those tests are carried out. Similarly, Dr. Carpenter said the results of her most recent project, the publicly available JUMP-Cell Painting project, were based on cells from a single participant. “We picked it with good reason, but it’s still one human and one cell type from that one human.” In an ideal world, she’d have a far broader range of participants and cell types, but the issues right now center on funding and time, or more appropriately, their absence.

But, for now, all we can do is await the results of these early trials and hope that they bear fruit. Like every other potential application of AI, its value will rest largely in its ability to improve the quality of the work – or, more likely, improve the bottom line for the business in question. If AI can make the savings attractive enough, however, then maybe those diseases which are not likely to make back the investment demands under the current system may stand a chance. It could all collapse in a puff of hype, or it may offer real hope to families struggling for help while dealing with a rare disorder.

Apple’s Mac just turned 40 years old! This week, Devindra chats with Deputy Editor Nathan Ingraham about his Mac retrospective. We focus on how much has changed since Apple’s disastrous 2016 lineup, why the Apple Silicon chips feel so revolutionary, and look back at our earliest Mac experiences. Also, we review the Framework Laptop 16, a wonderfully modular miracle of a laptop, but one that we wish had more graphics power for gaming. (But hey, at least you can replace the GPU eventually!).

Listen below or subscribe on your podcast app of choice. If you've got suggestions or topics you'd like covered on the show, be sure to email us or drop a note in the comments! And be sure to check out our other podcast, Engadget News!

Topics

• Framework Laptop 16 review: Amazingly modular, but not so great at gaming – 1:17

• The Mac turns 40 – 19:27

• More tech layoffs at Blizzard/Activision, Riot, eBay and others – 49:58

• Apple’s Car concept is allegedly still alive – 52:44

• Apple overhauls App Store rules in response to European Union regulation – 58:25

• Working on – 1:09:30

• Pop culture picks – 1:13:40

Subscribe!

• iTunes

• Spotify

• Pocket Casts

• Stitcher

• Google Podcasts

Credits
Hosts: Devindra Hardawar and Nathan Ingraham
Producer: Ben Ellman
Music: Dale North and Terrence O'Brien

A large downside to Windows PCs with Arm64 processors like Microsoft's own Surface Pro 9 5G has been a lack of native support for Chrome, the world's most popular browser. Now, Google has finally released a Chrome Canary beta version that fully supports the Arm64 architecture, Windows Central has reported. 

The new version should significantly accelerate Chrome performance on Arm64 PCs, negating the need to run Chrome in emulation mode. The download can be installed on PCs running recent versions of Windows 11 for Arm processors, with one user confirming it runs on a seven-year-old Snapdragon 835 SoC. 

Chrome has been available for some time on Google's Chromium on Arm64 and even Linux for Arm64, along with iOS and Mac. On top of that, Microsoft's Edge browser (which is based on Chrome) has run natively on Arm64 for years. So why the delay for Windows on Arm64? It may be because there aren't that many Arm64 Windows PCs and those that do exist are relatively expensive, especially compared to Chromebooks. 

Google might be reasoning that now is a good time to introduce the feature, since Qualcomm is set to release its Snapdragon X Elite chip, a successor to the Snapdragon 8cx Gen 3. Based on TSMC's latest 4-nanometer tech, it's promising performance double that of some 13th-gen Intel Core i7 CPUs with a third the power draw, allowing it to better compete with Apple's latest M-series silicon. 

If Windows laptops using the chip can finally deliver performance that's sadly been lacking in models to date, we may finally see them arrive in decent numbers. Snapdragon Elite X models are supposed to launch in mid-2024, so hopefully Google will be ready with a stable version of Chrome. If you have an Arm64 PC, you can download the Canary version here.