Technology

lundi 30 octobre 2023

Sam Bankman-Fried doesn’t recall

Photo Illustration of Sam Bankman-Fried in front of a graphic background of pixels and handcuffs. — Photo Illustration by Cath Virginia / The Verge | Photo by Bloomberg, Getty Images

Bankman-Fried gets a shot at his side of the FTX story — then promptly shreds his own credibility with the jury.

Midway through Sam Bankman-Fried’s cross examination, as prosecutor Danielle Sassoon went through a brutal line of questioning like a hot buzzsaw through a butter cow, I found myself reflecting on how smart the average person is. Maybe they don’t know calculus. Maybe they’ll never read Ulysses. Maybe they can’t code. But they definitely know how to identify bullshit when they see it.

So if you, like Bankman-Fried, have moved into the Clintonian territory of “it depends on how you define ‘trading’’” you done fucked up, son. Make whatever “sophisticated” argument you like; even the stupid will see through it.

At various points during Sam Bankman-Fried’s cross examination, I saw jurors shake their heads, frown so hard their lips disappeared, and make prolonged eye contact with each other. Personally, I now have a Pavlovian fear response to the phrase “Is it your testimony that…”

On the stand, Bankman-Fried’s demeanor suggested a spoiled child complaining he didn’t get the biggest scoop of ice cream at his birthday party. He didn’t want to answer the prosecutor’s questions, or his lawyer’s questions — he wanted to answer his own questions, which he liked better. He often replied to yes-or-no questions with nonsense.

The day did not start as an actual disaster. In fact, Bankman-Fried’s direct testimony was the strongest he’d given so far: clear, coherent, believable. On his home territory of direct examination rather than cross-examining the prosecutors’ witnesses, defense lawyer Mark Cohen improved his ability to order a simple chronological narrative, even if he once instructed us to move forward in time from November 7th to… November 7th.

But Bankman-Fried’s recounting of events was supported by very little other evidence. Gary Wang, Caroline Ellison, and Nishad Singh all had text messages, documents, code snippets, and so on to corroborate their versions of events. Bankman-Fried’s testimony had very little of that, and what little it did have was pretty thin.

We were introduced to a document where Bankman-Fried listed his priorities, including: “getting accounting right on FTX.” Cohen and Bankman-Fried used this to show how devoted Bankman-Fried was to getting to the bottom of the general fiasco with Alameda’s money. The idea was to display in real time FTX’s revenue and expenses, where its bank accounts were, how much investor money it had, and so on. This did reveal Bankman-Fried’s priorities: getting accounting right was ranked ninth.

So for the stories Bankman-Fried wanted to tell, we had to rely on… Bankman-Fried. We moved on to Bankman-Fried’s argument about hedging, which I still do not understand except as a way for him to say he’s a smarter trader than his ex-girlfriend, the former Alameda Research CEO Caroline Ellison. The actual evidence suggests Ellison is both a better trader and much savvier than Bankman-Fried. She modeled out a risk scenario that matched almost exactly what happened at FTX, for instance, to try to keep him from sinking $2 billion in venture investing. She is also cooperating with the government. If I have to bet on one of them, I’m going “balls long” Ellison.

Bankman-Fried’s other major task was figuring out how to explain away the infamous “FTX is fine. Assets are fine.” tweet. According to Bankman-Fried, he really believed that at the time. On November 8th, Alameda was still solvent, he said. Almost all of the other events occurred exactly as others had described them, except that Bankman-Fried sounded more heroic in this telling, if only because he was not doing any crimes this time.

Then we got what I had been waiting for: the cross examination. And while Bankman-Fried’s direct testimony was short on contemporaneous evidence, the cross was not. Sassoon was a matador in kitten heels, baiting Bankman-Fried before driving her sword through his shoulders.

Bankman-Fried claimed to have been “not involved as a general principle in day to day trading,” but this turned out to depend highly on how one defines trading. Sassoon quickly introduced the “Vertex” Signal groupchat for discussing Alameda’s trading. In it, we saw messages where Bankman-Fried asked the group how much of two tokens, OXY and MAPS, the group had bought. He then suggested Alameda should buy $1 million to $2 million of each over the next few days. (Bankman-Fried denied that this was him giving instructions, which depends highly on how one defines giving instructions.)

We also saw Bankman-Fried giving directions on Japanese government bonds. Hedging is a form of trading, and Bankman-Fried had just testified about giving Ellison directions on hedging. Then, switching to a December 2022 recording of a Twitter space, we heard Bankman-Fried say he was “not at all involved in the trading and hadn’t been for years” at Alameda. “I was intentionally not getting involved in it” because of possible conflicts of interest. Also in an interview with the Financial Times, Bankman-Fried said he’d walled himself off from trading.

I glanced over at Barbara Fried, the defendant’s mother. She appeared to be biting her nails.

Sassoon moved on to a series of questions about whether Bankman-Fried recalled extremely specific things. Did he recall saying that FTX had reformed how crypto exchanges worked? That he had built a responsible system? That FTX was a thoughtful exchange? That FTX was providing clarity and transparency to the crypto system?

To each of those questions, Bankman-Fried replied, “No, but I might have.”

And then Sassoon played a clip from FTX’s official podcast. You are never going to guess what he said on the pod.

There was more. Had Bankman-Fried referred to FTX as safe in numerous public statements? He couldn’t recall. Did he remember tweeting that the acceptable number of issues when it comes to a client’s money is zero? He couldn’t recall. Had he tweeted that lying to customers broke “sacred rules of conduct” everyone knows to follow? He couldn’t recall. Did he remember that he supported regulation only as long as it protected consumers? He couldn’t recall.

Every single question was followed by evidence of Bankman-Fried publicly using the precise language Sassoon had offered. Several journalists were in the courthouse — some even in the courtroom — as their articles about Bankman-Fried were read aloud. It was obvious to everyone in the courtroom what was going on. Bankman-Fried stuck to his “do not recalls” anyway.

Still, he hadn’t made any of those statements under legal oath, had he? Well… that remained true until we reached his Congressional testimony. Bankman-Fried read aloud testimony he’d submitted to Congress: that trading platforms’ obligations included maintaining sufficient liquid assets that customers could withdraw on request. That platforms should ensure appropriate bookkeeping to prevent misuse of customer assets. Ensuring appropriate management of risks. Avoiding conflicts of interest.

Sassoon immediately followed this with direct messages Bankman-Fried had sent to Kelsey Piper, in which he said this was all just public relations, and “fuck regulators.”

At this point, my notes simply read “Jesus fucking Christ” in all caps.

As the day wore on, I saw the mood in the jury box darken. At least three jurors were visibly fed up with Bankman-Fried’s “don’t recall” followed by the exact statement he’d been asked about. The remarkable thing was not that Sassoon had used Bankman-Fried’s many public statements to make him sound like a liar. It was that by denying he remembered making them so consistently, Bankman-Fried made himself sound like a liar.

Once that was accomplished, Sassoon moved on to the meat of the cross. Multiple statements, from multiple interviews, featured Bankman-Fried saying that Alameda had some special privileges — which was, of course, something Bankman-Fried had also denied to customers.

Then there’s the massive loss from MobileCoin that Alameda took on while under Bankman-Fried’s official leadership. None of the other backstop liquidity providers wanted the position, Sassoon asked, because it was a “loser” — right? As Bankman-Fried hemmed and hawed, Sassoon pointed out the loss cost Alameda hundreds of millions. Had Alameda not taken it on, FTX would have booked the loss itself, she pointed out. And FTX, unlike Alameda, had to share its books with investors. Scapegoating Alameda let FTX remain pristine, which made it much easier to attract VC funds.

But Sasson didn’t leave it there. The loss occurred because someone had exploited several loopholes on FTX in order to make the trade, she noted. Did Bankman-Fried tell investors about this exploit? He did not. Did he tell customers? Nope.

Behind all the finance sheets and code bases, the fall of FTX was in a way incredibly childish: a nerd posse running away with a bunch of other people’s money in the stupidest and simplest way possible. I understand Bankman-Fried went to MIT and majored in physics and was a successful trader on Wall Street. But sitting on the stand, he resembled nothing so much as a child who’s broken a family heirloom and is insisting his invisible friend did it.

All the news from Apple’s ‘Scary Fast’ Mac event

Apple’s Finder two-face icon in black with a ghostly white outline — *Dark Finder.* | Image: Apple

Apple plays a pre-Halloween trick with new product announcements coming in from a surprise nighttime livestream event.

Apple is hosting a dark and spooky-themed product launch event for the evening of October 30th, 2023, the night before Halloween. The company has, uncharacteristically, scheduled the online livestream presentation for 8PM ET / 5PM PT, meaning it’s not for the morning or lunchtime crowd this time.

Like every Apple event, the invitation comes with a title and themed logo that provides a clue as to what will be announced. This event is called “Scary Fast” and features a shadowy Apple logo that morphs into an only-slightly-creepy Mac Finder icon. Apple is expected to focus on new Mac products at this event, including its next-generation M3 chips and an updated iMac.

Apple could have other surprises in store as well, including but not limited to M3 processor upgrades for the higher-end MacBook Pros. Less likely are new iPads since Apple just released a new Pencil, but who’s to say Apple can’t give us more Halloween treats?

Find out all the news right here. We’ll be live blogging the event below.

Samsung’s Galaxy S23 starts receiving Android 14 update

Samsung S23 and S23 Plus on a tabletop lying horizontally with rear camera array showing — *The Galaxy S23 and S23 Plus.* | Photo by Allison Johnson / The Verge

Samsung’s Android 14-based One UI 6 update is leaving beta on the Galaxy S23 in Europe, users in the region are reporting. Across X, the platform formerly known as Twitter, and Reddit, people are seeing the update roll out in countries including the UK, France, Germany, Poland, Spain, and Portugal. While many are updating from the beta version of One UI 6, some are receiving a larger 3GB patch to update from the stable version of One UI 5.1.

While the update appears to be limited to S23 users in European markets for now, Samsung’s approach in previous years suggests that a stable US release won’t be far behind. The company’s Android 13-based One UI 5 update started rolling out for European Galaxy S22 owners around October 24th last year, which was followed by a US release roughly a week later.

One UI 6 has been available in beta for around two and a half months, and Android Police has a nice writeup of the biggest changes to look forward to with the update. These include updated camera and weather apps and widgets, a redesigned quick settings menu, and overhauled emoji designs. Here’s a complete changelist if you want to dive deeper.

To download the update, head into your Galaxy S23’s Settings app, tap Software update, and then Download and install.

Microsoft starts blocking ‘unauthorized’ Xbox controllers and accessories

Illustration by Alex Castro / The Verge

Microsoft is starting to block “unauthorized” Xbox controllers and accessories from being used on Xbox consoles. Resetera posters spotted a warning about the block last week, with some third-party Xbox controllers now throwing up a “connected accessory is not authorized” warning when connected to an Xbox console. It’s unclear if Microsoft is trying to target cheat devices, or whether the Xbox maker is trying to push its official partner program.

An error has now started appearing for some third-party Xbox controllers, alongside a warning that notes the accessory will be blocked from further use after two weeks. “From the moment you connect an unauthorized accessory and receive error code 0x82d60002, you’ll have two weeks to use the accessory, after which time it will then be blocked from use with the console,” says Microsoft in a support note. “At that time, you’ll receive error code 0x82d60003. We encourage you to contact the store or manufacturer where you obtained the accessory to get help with returning it.”

Third-party Xbox controllers that are part of the “designed for Xbox” hardware partner program are unaffected, but any that haven’t been officially authorized by Microsoft run the risk of generating this error and being blocked for use. This might also block third-party cheat devices like XIM, Cronus Zen, and ReaSnow S1 from working on an Xbox console.

These adapters are commonly used on PC to spoof controller inputs, so mouse and keyboard users can get the benefits of aim assist and reduced recoil from controller mixed with the benefits of movement from mouse and keyboard. Activision, Bungie, and Ubisoft have all been trying to block these hardware spoofing devices, with restrictions and bans in Call of Duty, Destiny 2, and Rainbow Six Siege.

Xbox console-related product Issue Update pic.twitter.com/QK0N41LmHW
— Brook Gaming (@brookgamingfans) October 20, 2023

Cronus devices have been growing in popularity on console, and Microsoft’s block will likely affect some of these unauthorized adapters. Brook Gaming, which manufactures an adapter that supports PlayStation controllers on Xbox says their device is affected by Microsoft’s block. In a post on X (Twitter), Brook Gaming warns of “functional disruptions in the near future” for a variety of its products, including a fighting board, controller adapter, and steering wheel adapter.

Most third-party Xbox controllers are wired, as Microsoft hasn’t typically licensed its Xbox Wireless protocol to other vendors. PowerA launched the first officially licensed third-party wireless controller for Xbox consoles earlier this year, and Windows Central speculates Microsoft’s latest ban could be related to the company working to expand approval for third-party wireless Xbox controllers.

We’ve reached out to Microsoft to comment on the “unauthorized” Xbox controller block and we’ll update you accordingly.

Biden releases AI executive order directing agencies to develop safety guidelines

An illustration of a cartoon brain with a computer chip imposed on top. — Illustration by Alex Castro / The Verge

President Joe Biden signed an executive order providing rules around generative AI, ahead of any legislation coming from lawmakers.

The order has eight goals: to create new standards for AI safety and security, protect privacy, advance equity and civil rights, stand up for consumers, patients, and students, support workers, promote innovation and competition, advance US leadership in AI technologies, and ensure the responsible and effective government use of the technology.

Several government agencies are tasked with creating standards to protect against the use of AI to engineer biological materials, establish best practices around content authentication, and build advanced cybersecurity programs.

The National Institute of Standards and Safety (NIST) will be responsible for developing standards to “red team” AI models before public release, while the Department of Energy and Department of Homeland Security are directed to address the potential threat of AI to infrastructure and the chemical, biological, radiological, nuclear and cybersecurity risks. Developers of large AI models like OpenAI ‘s GPT and Meta’s Llama 2 are required to share safety test results.

A senior Biden administration official told reporters in a briefing that the safety guidelines apply mainly to future AI models.

“We’re not going recall publically available models that are out there,” the official said. “Existing models are still subject to the anti-discrimination rules already in place.”

To protect users’ privacy, the White House called on Congress to pass data privacy regulations. The order also seeks federal support for the development of “privacy-preserving” techniques and technologies.

Part of the order plans to prevent the use of AI to discriminate, including addressing algorithmic discrimination and ensuring fairness when utilizing the technology for sentencing, parole, and surveillance. It also orders government agencies to provide guidelines for landlords, Federal benefits programs, and contracts on how to prevent AI from exacerbating discrimination.

Agencies are directed to address job displacement and produce a report on the impact of AI on the labor market. The White House also wants to encourage more workers to work in the AI ecosystem and ordered the launch of a National AI Research Resource to provide key information to students and AI researchers and access to technical assistance for small businesses. It also directed the rapid hiring of AI professionals for the government.

The Biden administration first released an AI Bill of Rights outlining a set of principles developers of AI models should follow. These were later turned into a series of agreements between the White House and several AI players, including Meta, Google, OpenAI, Nvidia, and Adobe.

But, an executive order is not a permanent law and generally only lasts the length of Biden’s administration. Lawmakers are still discussing how to regulate AI, though some politicians said they want to pass laws around AI before the end of the year.

Industry observers said the executive order is at least a step forward in providing standards around generative AI.

Navrina Singh, founder of Credo AI and a member of the National Artificial Intelligence Advisory Committee, said an executive order is a strong signal that the US is serious about generative AI.

“It’s the right move for right now because we can’t expect policies to be perfect at the onset while legislation is still discussed,” Singh said. “I do believe that this really shows AI is a top priority for government.”

Biden to Issue First Regulations on Artificial Intelligence Systems

Biden to Issue First Regulations on Artificial Intelligence Systems In an order to be issued on Monday, the White House will outline requirements that the most advanced A.I. products be tested to assure they cannot be used to produce weapons, among other regulations.

dimanche 29 octobre 2023

Android 14 will make it easier for apps to support passkeys soon

Illustration by Alex Castro / The Verge

Android apps are about to get better built-in passkey support. Google announced in a developer blog post last week that Credential Manager, a new Android-specific API for storing credentials like username and password combinations and passkeys, is going public on November 1st. Credential Manager, which has been in developer preview for months, houses biometric authentication of passkeys, traditional passwords, and federated identity login under one roof in Android phones.

Ultimately, the change should allow apps to offer better authentication support in Android 14. Using Credential Manager, apps can offer users easy biometric logins through passkeys. That should mean a more friction-free sign-in experience since people who use that method wouldn’t have to worry about keeping login information in their heads. Third-party password managers like 1Password can also integrate the API for a more streamlined experience when defaulting to such an alternative instead of Google Password Manager.

Google explained why it’s pushing for passkeys in Android at Google I/O this year.

Google wrote in another blog post last week that it’s deprecating several authentication APIs so that developers will only have to call on the Credential Manager for authenticating users. Hopefully, that should make it much simpler and, therefore, much more likely to be used by third-party apps, as others, like WhatsApp and Uber, have already done.

Apple is getting ready for Prime Time

Image: Apple

I’m trying to remember the last time a company’s big event came for prime time TV the way Apple’s 8pm ET keynote is trying to lock down our attention tomorrow. The only company I can think of that’s done it successfully lately is Victoria’s Secret with its Fashion Show. While I suspect there’s some overlap in audience between that and an Apple Keynote, I also suspect Bella Hadid is a bigger draw than a new iMac.

But Apple is still going for it! Promising a keynote that will air on its website, YouTube, and over Apple TV while things like Monday Night Football airs on broadcast TV. The company is ready to take all the cultural cache its phones, Bluetooth headsets, and computers have earned it and convert that fanbase into a bigger audience to advertise its products.

This big prime time Apple event feels like the natural next step for the company. No one else in the tech space has had the same success as Apple at getting people to treat their announcements as big events. Nearly every single major tech company has tried. Sony had Taylor Swift at a CES keynote, and Samsung marched out a member of BTS to applause at a Galaxy Unpacked event. Google had the Slo-Mo guys. Intel had dancers and acrobats festooned in LEDs. But something about an Apple event seems to resonate more with folks.

And after the iPhone event, it feels pretty clear the company has come close to the pinnacle of what it can do with an hour-long mid-day infomercial. One of the only ways to get bigger, grab more attention, and become a more consistent part of the conversation is to go prime time (or buy a social media company and run it into the ground). Take that impeccably produced product that captures the attention of techies and a few of their closest friends and move it to a time of night where a whole lot more of those close friends can watch.

And Apple is probably doing it now, instead of with the iPhone, because the stakes are lower. Fewer people care about a Mac event than an iPhone one. My brother calls me before an iPhone event to chat about the phones. He doesn’t do that nearly as often for a Mac one. Plus, the iPhone event can dramatically affect Apple’s share price. A spec bump for the MacBook Pro, while welcome, probably won’t move the needle as dramatically.

I can’t guarantee fewer people will tune into this event than the iPhone event last month, but I expect Apple’s less concerned with breaking its viewership records (the company doesn’t make those public). Instead, I’d hazard Apple is thinking about next year’s keynotes — especially any centered around the Apple Vision Pro. Apple’s going to need every tool at its disposal to make people care about a $3,500 AR and VR headset. If there’s a chance an evening keynote nets more attention than the traditional morning keynote, then Apple’s going to want to take it when it's trying to sell people on why AR and VR are the future of computing.

So why not do a test run with a series of Mac updates that wouldn’t really benefit from the traditional keynote where hundreds of reporters and analysts fly to Cupertino to sit in a theater and watch a video before going hands-on with the products?

But it will need to be one heckuva show. Tim Cook is going to need to do more than come out an an AFC Richmond kit with a Ted Lasso mustache spirit-gummed to his upper lip. The announcements (at least the rumored ones), aren’t going to be enough. I suspect that in addition to those rumored Mac updates, we’ll also see more skits like that one from the iPhone event starring Octavia Spencer as Mother Earth.

It probably won’t be as star-studded, given SAG-AFTRA is currently on strike and bargaining with Apple, among other studios. So there probably won’t be a series of cameos from the actors of Apple’s best-known shows and films. Which means no Jennifer Anniston and Reese Witherspoon doing the Apple equivalent of an SNL digital short with Eddy Cue or Adam Scott and Patricia Arquette staring at each other from across a table for an unsettling amount of time before Craig Federighi interrupts as the newest member of the Severance program and shows off an iMac. But maybe Martin Scorsese will show off how easy it is to access his new Letterboxd account on a MacBook Pro. Or maybe half the keynote will be rendered in the video game Resident Evil Village (naturally made on a Mac).

Regardless, I’m still assuming Apple will go in a more creative direction, because besides the company wanting to test run keynotes at new times, it's also increasingly showing off its Hollywood ambitions. We heard rumors of its ambitions earlier this year when Bloomberg reported the company was looking to spend approximately one billion dollars on new film programming year. Now its films like Killers of the Flower Moon and Napoleon are contenders for the Oscars (it snagged its first Academy Awards last year for Coda). Earlier this week Apple TV Plus even saw its second price hike since launch (alongside a slew of other Apple services).

The company wants people associating it with entertainment. It would be wild for its very first prime time event to ignore that entire side of its business just to show off some new M3 iMacs. Apple can’t just give us a time-shifted keynote. Well, it can... but that won’t really match its ambitions.

Your guide to the internet’s most (and least) important moments

Image: William Joel / The Verge

Hi, friends! Welcome to Installer No. 12, your guide to the best and Verge-iest stuff in the world. (If you’re new here, welcome. I’m so psyched you found us, and also, you can read all the old editions at the Installer homepage.)

This week, I’ve been testing the Superlist tasks app, reading about the high-stakes race to crack crypto wallets, rehearsing my case for buying an e-bike, taking voice notes like a boss with Whisper Memos, and avoiding all my problems by watching Between Two Ferns bloopers on YouTube.

I also have for you a new pair of AR glasses, a Netflix thriller, all the great sci-fi you could ever need, a great parenting app, a super-deep interview about Windows, a new puzzling platform, and a bunch more.

As always, of course, the best part of Installer is your ideas and tips. What app should everyone be using? What movie / show / podcast / book / spaghetti recipe does everyone need to know about? Tell me everything: installer@theverge.com. And if you want to get Installer in your inbox and a day early every week, subscribe here.

So much good stuff this week — let’s get into it.

The Drop

Internet Artifacts. Neal Agarwal’s whole website is an endless string of delightful games and other silly things, and this museum of important internet things — the first MP3, the first-ever “LOL,” footage from the web’s first webcam, Pizzanet, and much more — is some of his best work yet. So much cool stuff here I’d never seen before. (A lot of people sent this in this week, and with good reason — thanks to all who mentioned it!)
The 2023 Hugo Awards. If you’re a science fiction fan, here’s your reading / viewing list for the next year or so. Nettle & Bone by T. Kingfisher (who is just Ursula Vernon under a pen name) won best novel; Samantha Mills’ Rabbit Test won best short story; The Expanse and Everything Everywhere All At Once and Dune and Andor are all among the nominees and winners. There’s so much good stuff here.
Apple’s Journal app. iOS 17.2 is starting to roll out now, bringing with it the private-diary app Apple announced at WWDC this year. I’ve been playing with the app for a bit, and it’s… fine? Like, it is, in fact, a journal app. But I’m not quitting Day One yet.
The Killer. You already know this about me, but I’m here for anything about spies and assassins and, in general, people who are both good and bad and don’t know where the lines are. I am also an unabashed David Fincher fan. This one’s in theaters now and on Netflix in two weeks, so, yeah, I’ll be keeping my outrageously expensive Netflix subscription a bit longer.
Google Maps. I know, new app, super exciting, right? But Maps actually got a really useful update this week. I don’t care much about the immersive views, personally, but the better tools for EV charging and especially the improvements to the “what cool stuff is happening nearby” features are going to be super useful.
The Leica M11-P. It’s $9,195, and I’m sure it’s fantastic because it’s a Leica. But I’m more interested in the built-in support for Adobe’s Content Authenticity tools, which sign and verify each image. In this increasingly messy AI-filled world, I love this idea.
Alan Wake 2. A horror-story action game that’s also a detective thriller with a lot of super-weird storytelling that seems to add up to a coherent, fun, intense package? Sign me up. (Also, I officially have too many great games to play and not enough time. Please consider this my vacation request.)
The Fight Over AI Music. The music industry is having to figure out its rules and norms about AI faster than just about anybody. Cleo Abram does a good job digging into how it all works and why it’s so complicated. And she asks a really important question that not a lot of people do: is any of this AI-generated stuff actually something we want?
Xreal Air 2. There are basically two kinds of AR glasses right now. One type is the Ray-Ban Meta smart glasses, which look cool but are all audio. These are the other kind: they’re a screen that goes on your face. That’s it. That’s the whole $400 idea. But… I kinda get it.
“How Does AI Actually Work?” From our friends over at the Waveform podcast, this is a super accessible, thoughtful dive into what we mean when we talk about AI — and intelligence in general. It gets kinda deep!

Pro tips

A couple of weeks ago, a lot of you told me you like the app Sequel for tracking all the movies, books, TV shows, and everything else you want to get to. Romain Lefebvre, the app’s developer, just launched Sequel 2.1 — which integrates with the discovery and scrobbling app Trakt, has a bunch more cast and crew info, and adds spoiler-avoidance features.

I asked Romain to give us some tips on how to use Sequel and, in general, how to make the most of a stuff-tracker app like his. Here’s what he said:

Start with what you’re actually excited about. “Avoid the temptation to catalog everything you’ve ever consumed, as this can be overwhelming. First, focus on adding recent or upcoming releases you’re looking forward to. This will help you gauge if the app is right for you. Coming up blank? Explore works from the actors and directors of your favorite movies and series! If you use a service like Trakt, check if the app can sync with your account, as this will help you evaluate it even more easily.”
Put the app one tap away. “Place the app where you can quickly add your friends’ recommendations or mark items as watched as soon as you finish them. Widgets are also a great way of surfacing recent releases and can serve as a quick entry point, too. (Sequel has a quick access widget to allow you to open up the search from your homescreen and lockscreen.)”
Enable notifications. “Turn on notifications for new releases from your tracked list. These reminders not only keep you updated on exciting new content but also encourage you to track what you’ve recently consumed.”
Regularly trim your list. “It’s easy to accumulate more than you can realistically consume. Avoid an overwhelming and stale backlog by periodically reviewing and removing items you’re unlikely to finish. (Sequel, for instance, allows you to archive a series and ‘abandon’ a book or game you don’t intend to finish, keeping your list realistic and up to date.)”

Be intentional with your choices. “Your tracker is a valuable tool to guide your media consumption. Avoid jumping into the first recommendation from a streaming service. Instead, take the time to browse your tracker and choose something that you really care about. This can also help you curb impulsive purchases by checking what games or books you already own before buying the latest bestseller.”

Screen share

Alex Cranz, The Verge’s managing editor, loves terrible TV. Like, earnestly loves bad shows in a way that almost makes you forget that they’re bad — even though they’re definitely for sure bad. That is just one of many delightful things about Alex Cranz.

In addition to covering the streaming biz and spending an alarming amount of time defending The Morning Show to haters everywhere, Alex also keeps a lot of The Verge running in the right direction and on time. So I’ve always been curious how she manages the many things going on, seems to have always watched and played everything, and keeps it all straight. So I asked her to share with the class!

Here’s Alex’s homescreen, plus some info on the apps she uses and why:

Three panels of a homescreen from an iPhone

The phone: iPhone 14 Pro

The wallpaper: Remember those photos that went viral after a woman found them on an old roll of film? I thought they looked neat, so I made them my wallpaper. Turns out they were rejected images for an advertisement shoot.

The apps: I love flicking between pages on my phone, so my most used apps don’t necessarily mean the ones on the front page. Slack is where the work and group chats live, Outlook is where the email and calendars live, Carrot is where the weather lives, and Ulysses is there to remind me I should write more. I try to keep everything else in folders, which means I forget to open other texting and social media apps unless the notification pops when I’m looking at my phone. The only other apps I keep at the ready are Parcel, a great package tracking app; Storygraph, which is like if Goodreads was actually good; Wikipanion, because I like browsing Wikipedia articles when I’m bored; and Google Authenticator, because work.

As always, I also asked Alex to share a few things she’s into right now — bad TV shows ideally not included. Here’s what she came back with:

Ugly Betty on Netflix. I’ve been rewatching it, because America Ferrera, and really enjoying how it accidentally explores two enormous transitions that happened in the late 2000s: the decline of print media and the rise of the phone as a do-everything device.
He Who Drowned the World by Shelley Parker-Chan. It’s the sequel to She Who Became the Sun and is a queer fantasy retelling of the founding of the Ming dynasty. Parker-Chan is a part of this new class of fantasy writers that all seem to be asking, “What if more big violent fantasy sagas were centered on the experience of women and queer people?”
For All Mankind. I know the fourth season of this Apple TV show isn’t available to most people yet, but I’m obsessed with the new season, which is set in 2003 on a rapidly commercializing base on Mars. If you’ve struggled to envision a near future where space mining matters, this show will help.

Crowdsourced

Here’s what the Installer community is into this week. I want to know what you’re into right now as well! Email installer@theverge.com with your recommendations for anything and everything, and we’ll feature some of our favorites here every week.

“The Quiche Browser is really great! You can rearrange the whole search bar area to your liking and also add / remove buttons to / from the search bar area in the settings and change the layout of all the buttons, too. You can also add a ‘reading time’ to the search bar area, which shows approximately how much time it would take you to read the page. You can edit every pop-up menu in the settings. You can customize the color of the search bar area. It’s a really great browser!” — Harun

“Not a new feature but one I just remembered and utilized: sometimes I use the iPad as a white noise machine for the kids when we’re traveling, but I was using a free app that would cut off after eight hours and wake them up early. I just remembered iOS 15 introduced ‘Background Sounds,’ so now I use that as a built-in white noise machine!” — Thaddaeus

“Bodies on Netflix so far has been rather interesting.” — James

“I decided to move away from the native Google keyboard (GBoard), seeking something new and refreshing. I remembered SwiftKey, the pioneer in gesture / flow input before Microsoft acquired it. Recently, Microsoft added Bing Chat to it, which was a pleasant surprise! It’s now effortless to modify tone and compose messages with various parameters like format and length without ‘leaving’ the keyboard. It feels revitalized, and I’ve adopted it as my main keyboard, much like a decade ago.” — Andriy

“I just recently learned about the TP-Link AV-2000 Powerline Adapter, and I think it’s exactly what I need to get a wired internet connection from my horribly placed router to my home entertainment center!” — Charles

“Worth a watch: a three-hour interview with Dave Cutler, ‘The Mind Behind Windows.’” – Michael

“The app that helped us through those first few months of parenthood was Nara Baby. Great design, highly customizable to choose what you want to track, and you can use it like a power user or more casual one. I still use it to keep track of my daughter’s height and weight and sometimes to track medicine use when she is sick.” — Jasper

“Apple’s Reminders app: now with columns! I just switched from paper to-do lists to this, and it’s kind of blowing my mind. I can add things from any device! I can mark them off anywhere. This is not an absurdly techie thing, but it’s made things easier for me.” — Will

“I’m absolutely loving Pikmin 4. During a crowded time for games, I keep going back to it over others. It’s great for newcomers and series fans alike. It’s gorgeous and has very chill vibes. A lot of time pressure from previous games has been removed and allows the player more freedom to complete levels and challenges.” — Bobby

Signing off

Y’all. A new Taylor Swift album came out this week. CAN YOU EVEN BELIEVE IT? I’ve spent the last 24-ish hours mostly streaming “Say Don’t Go” on repeat while also continuing to read and watch everything I can about what a phenomenon Swift has become. Bloomberg has a great data visual thing about her new billionaire status, The Wall Street Journal has a good look at what you might call the Taylor Swift Industrial Complex, The Eras Tour is still one of the biggest movies on the planet, TikTok is melting down trying to figure out who all the new songs are about, and honestly, no one on the planet makes a lyric video better than Taylor Swift.

You may not like her music (though you should), but there’s no denying that Taylor Swift is on pretty rare celebrity and artist ground here. And also, I mean, any excuse to listen to “Blank Space” 4,500 more times is fine by me.

See you next week!

samedi 28 octobre 2023

Google researchers use off-the-shelf headphones to measure heart rate

Illustration: The Verge

Typical heart rate monitoring in wearable tech, like smart watches or wireless earbuds, relies at least partially on photoplethysmography (PPG), which uses light pulses to measure blood activity. It works generally well, but it has its limitations. Google scientists wrote in a new research blog spotted by 9to5Google yesterday that they had tried a different approach, called audioplethysmography (APG), that uses ultrasound to measure heart rate. And they did it with off-the-shelf active noise-canceling (ANC) earbuds and a software update.

The trick works by bouncing a low-intensity ultrasound signal off the inside of the ear canal and using the tiny microphone that helps make ANC work to detect skin surface perturbations as blood pumps through it. According to the blog, the technique was “resilient” even given a bad ear seal, differing ear canal size, or darker skin tones. That last one is notable since heart rate accuracy with darker natural skin tones or tattoos has been an ongoing problem with smartwatches and other wearables until now.

A cross-section of the ear canal showing soundwaves bouncing off the ear canal and back to an earbud, and a graph showing pulse data. — *A gif illustrating audioplethysmography.*

Google’s researchers also found the ultrasound approach worked fine when music was playing, but said that it had issues in noisy environments and that “the APG signal can sometimes be very noisy and could be heavily disturbed by body motion.” However, they found they could overcome the motion problem by using multiple frequencies and teasing out the most accurate signal among them.

In addition to commercially available earbuds, the researchers also used purpose-built prototypes to test the effect of microphone placement. The field study was performed with 153 participants. The researchers said the median error rate for heart rate and heart rate variability was 3.21 percent and 2.70 percent, respectively.

Heart rate monitoring headphones have been around for a while, but they use the PPG approach and can be very sensitive to intense movement or a bad fit.

Bear in mind that this is only a study and doesn’t mean Google is about to release headphones that do this (or update your Pixel Buds Pro to do it). Still, it hints at the company’s ideas as it makes more of an effort in the wearables space. If you want to read Google’s study in detail, you can download it on Google’s site.

Blue-Collar Workers Are the New Social Media Stars

Blue-Collar Workers Are the New Social Media Stars A new crop of popular social media personalities includes a trucker, a shepherd and a commercial fisherman. What they earn from sponsorships is just a nice bonus.

vendredi 27 octobre 2023

The jury finally hears from Sam Bankman-Fried

Photo Illustration of Sam Bankman-Fried in front of a graphic background of inverted money. — The royal “we.” | Photo Illustration by Cath Virginia / The Verge | Photo by Michael M. Santiago, Getty Images

He was an introvert!

It is honestly kind of incredible to watch a man torpedo his own credibility on direct testimony. We’re not even at the cross yet, and the judge has already instructed him to answer the question he’s being asked by his own lawyer.

The jury is watching all of this intently.

The main thing that’s been clear so far from Bankman-Fried’s testimony is that the man really loves the sound of his own voice. So far, the count of “Objection, narrative!” to Bankman-Fried’s answers, followed by “Sustained” is at three.

Also, sometimes when Bankman-Fried says “we,” he only means himself.

Yesterday, during an evidentiary hearing, Bankman-Fried was repeatedly scolded by Judge Lewis Kaplan for not answering prosecutor Danielle Sassoon’s questions on cross-examination. Today, Bankman-Fried was scolded by Kaplan for not answering his counsel Mark Cohen’s questions on direct examination. Bankman-Fried has also occasionally interrupted Cohen with “yes” and “yup.”

We spent a lot of the morning explaining vocabulary. I will spare you the full list, though I will say that explaining “Amazon Web Services” and “database” was a bit too detail-oriented. Then Bankman-Fried tried to define “market manipulation.” After Bankman-Fried gave his definition, Kaplan told the jury that he was the final authority on that, thanks.

In fairness to Bankman-Fried, he has been clearer and much easier to understand than he was at the evidentiary hearing yesterday. There was a minimum of word salad today. I don’t know if he was more relaxed, or he’d just been more rehearsed, but I will certainly be watching to see if he suddenly becomes much less coherent when Sassoon gets him for the cross.

Here is the story of FTX, from his point of view.

Bankman-Fried, who informed us he’s “somewhat introverted, naturally,” gave us a rather prolonged tour of his pre-Alameda Research life, which I will skip. In 2017, during a crypto bull run, he started his cryptocurrency trading firm. He knew “basically nothing” about cryptocurrency at the time, he explained, but he wanted to do arbitrage on it anyway.

Alameda Research was named for Alameda County in California, which was where its first office was set up. As for its name, here’s what Bankman-Fried said on the stand:

Effectively, we wanted to be under the radar at that point in time. I didn’t want to call it Sam’s Crypto Trading Firm or anything like that. We — there are a lot of competitors and people who we didn’t particularly want to know what we were building out because they would race to do it. “Research” was a sort of generic word, which filled out the company name. And that was — it was far better than the internal name that we had at that point, which was Wireless Mouse.

I would find this much more believable if I hadn’t already watched a video of Bankman-Fried explaining on a podcast that the name made it easier to get a bank account. That happened during the first day of Gary Wang’s testimony. Bankman-Fried was there, too. You know who else was there? The jury.

Anyway, Bankman-Fried went on a hiring spree for Alameda. He rounded up his merry gang of alleged co-conspirators. First, Wang, to program the computers. Then, Nishad Singh, about a month after founding Alameda Research. Finally, Caroline Ellison.

Though Bankman-Fried was the CEO, and also the majority owner, he wanted to be clear: he did not supervise Wang’s direct work. Anyway, after a bunch of wildly successful arbitrage — 50 percent to 100 percent annualized returns, per his testimony — he decided to found a cryptocurrency exchange, FTX. He figured he’d fail; that there was only a 20 percent chance of success. Bankman-Fried did not define what he expected the time period to be on this estimate, but arguably 20 percent was a much higher chance of success than FTX would enjoy once Alameda dipped into the customer deposits.

By the way, because he was such a good guy, Bankman-Fried made a point of “periodically” handling support tickets himself. “I worried if I didn’t, I would lose touch with the actual concerns of the customers,” he testified. What he did not do was create a risk team, which he is now characterizing as a “big mistake.”

Risk is an inherent part of a futures exchange, which is even more like a casino than regular cryptocurrency. Not having a risk team, when you are any kind of financial anything, is certainly a choice. It is especially a choice when you go around telling everyone your crypto exchange is very good and safe.

FTX’s big selling point was its “risk engine,” which was supposed to prevent big losses that would then be spread around all the rest of the customers. But Bankman-Fried testified that in 2020, the “risk engine was effectively sagging under the weight” of the exchange’s rapid growth. So its time to liquidation went up — it took minutes to determine which accounts needed to be liquidated. As a result, at one point the risk engine got stuck in a catastrophic feedback loop that would have created losses in the “trillions of dollars,” Bankman-Fried testified. As part of that feedback loop, Alameda teetered on the brink of liquidation, which “would have disastrous consequences” for FTX.

Because of that experience, Bankman-Fried suggested an “alert” or “delay” that would keep Alameda from being liquidated by a bug. This is the supposed origin story of “allow_negative,” which Bankman-Fried says was the eventual result of that conversation, and that he says he didn’t know about until very recently.

There is a problem with this story. “Allow_negative” was coded and switched on in 2019. I saw the code in court, and so did Bankman-Fried, who was also there for the testimony. Perhaps you are wondering, was the jury also there? Reader, it was.

Bankman-Fried denied he knew about the effectively infinite line of credit Alameda Research received from FTX. This argument was peculiar; essentially my take-away was that the CEO of a financial company simply didn’t pay attention to finances.

FTX couldn’t get bank accounts right away. Bankman-Fried anticipated it would take a year or two. Rather than wait, he decided to use Alameda as the “payment provider” for bank transfers. “My understanding at the time was that there were teams managing the process,” he said. “At the time, I wasn’t entirely sure what was happening.”

Well, sure, understandable! He’s an introvert!

In 2021, FTX was growing to millions of users, with $1 billion of revenue. Bankman-Fried said he worked 12 to 22 hours a day, and took one day off every couple of months. Because FTX had grown so much, he could no longer run both companies, he said. Bankman-Fried handed the company off to Caroline Ellison and Sam Trabucco, who immediately after being named co-CEO promptly drifted away to early retirement. (Quiet quitting king!) Bankman-Fried did remain involved in hedging and risk at Alameda, though.

About that $1 billion of revenue in 2021: Bankman-Fried definitely did not know that Singh, his employee, had backdated interest payments to get FTX “over the line” to $1 billion. See, he’d just asked his employees to check and see if there was any source of funds that was missing to get to $1 billion. This testimony was especially rambling.

Oh, also that MobileCoin loss? The one Wang said Alameda took to keep off FTX’s balance sheet? Yeah, so it was a totally innocent thing where what actually happened was that Bankman-Fried thought it was appropriate that Alameda take the position as a backstop liquidity provider, that’s all.

In June 2022, Bankman-Fried heard about the account called “fiat@ftx” tracking how much money Alameda owed to FTX, he testified. He did not know what it was and did not bother to find out. He was busy! That was when Bankman-Fried directed Ellison to repay Alameda’s lenders, because he thought Alameda was good for it. He also gave BlockFi and Voyager, two crypto lenders, some capital infusions for good measure.

Remember that testimony Adam Yedidia gave about a conversation with Bankman-Fried in August 2022 about the enormous amount of money Alameda owed FTX? Well, Bankman-Fried remembers it differently. See, Yedidia was just asking about Alameda’s risk profile, and Bankman-Fried wasn’t talking about insolvency at all.

Also, when Singh and Bankman-Fried had the dramatic balcony conversation at their penthouse, it was just that Singh thought Alameda’s liabilities had gotten too high, and FTX was spending too much money on marketing. But Bankman-Fried still thought that Alameda had more assets than liabilities, so it was all fine, and besides, if Singh thought he was going to be better at marketing, he could take it over. It didn’t have anything to do with the money Alameda owed FTX at all.

Of course not! Bankman-Fried didn’t learn about the $8 billion liability associated with Alameda until October 2022, he said. And he learned it all by himself, by looking at a computer database. When he found it, he was “very surprised!”

Besides the two obvious lies Bankman-Fried told on the stand — about Alameda Research’s name and about “allow_negative” — I have been struck by how little he seems to know about his own companies. Apparently, Singh, Wang, and Ellison were out there just doing whatever their little hearts desired. Because Bankman-Fried was a CEO, but definitely not the kind that pays any attention to money at his crypto trading firm and futures exchange.

We had to stop for the day, but I am very excited to hear on Monday about what new surprises Bankman-Fried will have in November 2022, when FTX falls.

Humane’s Ai Pin could cost $1,000 — and require a subscription

A photo from Imran Chaudhri’s TED talk about the Humane Ai Pin. The device projects details of a phone call onto his hand. — Image: TED

The Ai Pin, the new gadget / wearable device / projector / thing from the secretive startup Humane, might cost as much as $1,000 and may require a monthly subscription for data, according to The Information.

The mysterious device has been in development for years, but we got our first good look at it during co-founder Imran Chaudhri’s presentation at TED this year. In the presentation, he used then unnamed device to accept a phone call, get information about where to buy a gift, translate a sentence that is then spoken in an AI-made version of his voice in French, and even get an opinion on whether he can eat a chocolate bar. It was an impressive demo, though we had a lot of questions about how it all worked.

As the November 9th launch of the device draws near, however, the shape of the Ai Pin is becoming clearer — and The Information’s report has some new details. (Including that the company moved to a November 9th announcement from a planned launch timed with the October 14th eclipse after reports that Sam Altman, Humane’s biggest shareholder, was in discussions with Jony Ive about some kind of AI gadget.)

The Ai Pin is “a small, screenless device about the size of a saltine cracker,” The Information reports, that will have “a camera, a microphone and speaker, a variety of sensors, and a laser projector.” You’ll attach it to your clothes magnetically (as seen in the device’s appearance at Paris Fashion Week). It will have a Snapdragon chip from Qualcomm that will give it “smartphone-level speed, connectivity, camera capabilities and security,” and Humane plans to be an MVNO so that it can sell cellular data that customers can use with the Ai Pin. (Humane co-founder Bethany Bongiorno described the Ai Pin this month as a standalone device that is a “phone, contextual computer, and software platform.”)

a stand-alone device.
a phone, contextual computer, and software platform.
built from the ground up for Ai. https://t.co/LVZz9YB8F8
— Bethany Bongiorno (@bella_bongiorno) October 5, 2023

The AI features in the Ai Pin will be powered by a proprietary large-language model, The Information says — which we kind of already knew from a short Time writeup that said the device uses “a mix of proprietary software and OpenAI’s GPT.” (A previous version of that article said the Ai Pin used GPT-4, but it that “4” is not present in the live version.) The Time piece also described a “Trust Light” privacy indicator that activates when the camera, microphone, or “input sensors” are on.

Our Ai Pin made the “@TIME Best Inventions” list. Thank you to our team, partners, and investors across the planet. See you on November 9th! https://t.co/S7xZsDFD32
— Humane (@Humane) October 24, 2023

There’s still a lot we don’t know about the device, or Humane’s future once it launches. But according to The Information’s sources, Humane has a grander vision than just the Ai Pin; “they suggested that Humane is hoping to create a total paradigm shift in consumer computing.” I’m a little skeptical about that one, but I guess we’ll see if Humane hints at that future when it shares more about the Ai Pin on November 9th.

Did A.I. Write Product Reviews? Gannett Says No.

Did A.I. Write Product Reviews? Gannett Says No. Writers and editors at Reviewed, a product recommendation website, say that the company used artificial intelligence to create reviews last week.

jeudi 26 octobre 2023

Ford hits the brakes on $12 billion in EV spending because EVs are too expensive

Ford F-150 Lightning at a dealership — Photo by Scott Olson/Getty Images

Ford is postponing $12 billion in EV factory building, including a planned battery factory in Kentucky. The reasons given were an unwillingness by customers to pay extra for its electric vehicles. You see, they’re too expensive, and now Ford’s massive transformation into an EV company is now going to take a lot longer than before.

Ford’s EV business continues to lose money, around $1.3 billion this past quarter in adjusted earnings. So far this year, Ford has lost $3.1 billion on its EV spending and has said it’s going to lose a total of $4 billion for the year.

The Kentucky plant, a “mega campus” that builds lithium ion batteries for electric cars, will be put on hold. But its Blue Oval City project in Tennessee was still moving forward.

Ford’s not alone in all this, of course. General Motors is pushing back production of its new slate of electric trucks and SUVs. Tesla CEO Elon Musk spent a large chuck of his last earnings call moaning about interest rates. It’s rough out there right now.

Customers would probably agree. Most of the early adopters have, well, adopted, and the next tier of possible customers has enough sticker shock to keep their wallets closed. Ford has tried to address this with new releases like the F-150 Lightning Flash, a mid-priced trim of its electric truck. The company says that the customers will decide how many EVs it makes — and right now, that means tapping the brakes on big projects.

It’s not all bad news. Ford reached a tentative agreement with the United Auto Workers last night, being the first of the Big Three US automakers to get a deal. Sure, the strike cost it around $1.3 billion, and the company pulled its guidance for 2023 — meaning its not confident it can hit the targets it laid out earlier in the year.

Here is Intel’s new Bong-filled hip-hop hold music

Graphical illustration with Intel branding — Illustration by Alex Castro / The Verge

I know, I know, my headline is a lot — but it’s true!

Intel, the chipmaker, has turned its iconic five-note Intel Bong sound mark into a piece of music that now greets investors, analysts, and journalists who tune in to the company’s earnings calls. It’s got a drum kit, record scratch, and everything:

“Wait, is that Intel?” asked my wife the moment the conference call hold music began playing over my speakers. Mission accomplished, I presume! Also... it’s kind of catchy?

Intel corp comms director William Moss confirms this is the first time the company’s used the tune.

In case you’re more interested in Intel’s future than its muzak, its Q3 2023 earnings call suggested it’s doing well! While sales aren’t exactly growing and the PC sales slump isn’t over just yet, the company made a second consecutive quarterly profit, in this case $310 million, and reported better results from nearly every business compared to last quarter.

Intel CEO Pat Gelsinger says his customers already completed their inventory burn in the first half of the year (so perhaps the killer deals we’ve been seeing on PCs and components will subside?), and he spent a lot of time heralding an “AI PC” revolution. The word “AI” was uttered over 60 times on today’s call.

That’s far more often than Intel’s previous favorite phrase, “five nodes in four years,” though the latter got plenty of airtime as well. (Intel says the plan is still on track, with three customers lined up for Intel 18A as of now.)

Inside Google’s Plan to Stop Apple From Getting Serious About Search

Inside Google’s Plan to Stop Apple From Getting Serious About Search Google has worried for years that Apple would one day expand its internet search technology, and has been working on ways to prevent that from happening.

mercredi 25 octobre 2023

Valve officially releases SteamVR 2.0

A screenshot of SteamVR 2.0. — Image: Valve

Valve announced Wednesday that it has released SteamVR 2.0, launching the major update exactly a month after the surprise launch of the SteamVR 2.0 beta in September.

“In this release we’re bringing all of what’s new and exciting on the Steam platform into VR,” Valve says in a Steam post. “This is our first big step in a larger ongoing effort to better unify the Steam ecosystem for all users, providing a more consistent experience across devices. This update also allows us to add new Steam features in the future much faster and more frequently.”

SteamVR 2.0 has been a long time coming, with Valve saying in a 2019 year-in-review Steam post that it was “hard at work” on the update. Here are some of Valve’s highlights for what you’ll find in SteamVR 2.0, which match pretty closely to what Valve pointed to in September:

Most of the current features of Steam and Steam Deck are now part of SteamVR

Updated keyboard with support for dual-cursor typing, new languages, emojis, and themes

Integration of Steam Chat and Voice Chat

Improved Store that puts new and popular VR releases front and center

Easy access to Steam notifications

Interestingly, the update arrived without any sort of announcement of new hardware — Valve is rumored to be working on some kind of new VR headset, though it’s unclear what form it might take. A mystery gadget passed radio certification in South Korea in September, which could point toward a potentially imminent hardware announcement of some sort, VR headset or not. And Valve has also reportedly been developing a standalone VR headset codenamed “Deckard,” which, if released, would compete with Meta’s standalone Quest VR headsets.

Honor says its new phone lets you open apps with your eyes

Image of a woman gazing at phone screen with Uber app in a pill-shaped notch. — *The Magic Capsule technology is coming to Honor’s next flagship.* | Image: Honor

We were expecting — and we got — a lot of talk about generative AI this morning when Honor CEO George Zhao took the stage at Qualcomm’s 2023 Snapdragon Summit. But the announcement of its Honor 6 flagship came with a surprising detail: it includes a feature that lets you interact with the device using your eyes. With some notable concerns about privacy implications, it looks kinda cool.

The keynote briefly featured a rendering of what this technology will look like, showing a woman looking at her phone with a snippet of the Uber app running at the top of the screen — something like a Live Activity. By changing the direction of her gaze, she opens the app in full.

Honor calls the technology Magic Capsule, and describes it as “eye-tracking based multimodal interaction,” which is more descriptive but less fanciful than Magic Capsule.

GIF showing eye tracking cursor hovering over and opening Uber app. — *What if your phone just did your bidding based on your gaze? That would rule.*

It’s one feature on the upcoming Magic 6 that will also feature a virtual assistant that utilizes Qualcomm’s on-device AI. You can ask it to do things like gather all the videos on your device that meet a certain criteria, whittle them down by other characteristics, and have it generate a new video highlighting your clips. We’re going to see a lot more of that kind of thing in the near future, too, because this year’s Snapdragon Summit is All About AI. We were 15 minutes into the main keynote before 5G was even uttered once, for the record.

Whether — and how — Magic Capsule works is a question mark. The demo video is hardly a real-life representation, and it seems like a feature with the potential to introduce more frustration than it’s worth. The “multi-modal” descriptor seems to indicate that gaze is just one input in the system, so it could be coupled with other gestures to work reliably —perhaps like how we’ve seen PSVR 2 games use eye-tracking to highlight things before you click to confirm. Also: do you want your phone to know where you’re looking? It’s no small issue when you’re talking about a state-backed company like Honor.

All of that aside, it’s nice to see device OEMs pushing for advances in how we use our phones that don’t begin and end with an AI chatbot. Reliable eye-tracking technology would have some real accessibility benefits, and it’s not entirely out of left field. It could be handy for moments when your hands are full — Apple certainly seems to think there’s a need for new ways of controlling your devices.

Honor hasn’t said specifically when the Magic 6 will ship, but Qualcomm says that phones with its new flagship chipset will start arriving in the coming weeks.

What the U.S. Has Argued in the Google Antitrust Trial

What the U.S. Has Argued in the Google Antitrust Trial As the government wraps up its case in the landmark monopoly trial, it has built a picture of how Google became dominant in online search — and the harms that it says resulted.

mardi 24 octobre 2023

Microsoft now thirstily injects a poll when you download Google Chrome

*Microsoft Edge now pops up a poll after you press the “Download Chrome” button.* | Screenshot by Sean Hollister / The Verge

How many licks does it take to get to the center of a Google Chrome download in Microsoft’s Edge web browser? How many times will the company try to steer me away?

Let’s check!

One:

*“There’s no need to download a new web browser,” reads the top of a search for “Chrome download” in Microsoft Edge.*

Two:

*“Microsoft edge runs on the same technology as Chrome, with the added trust of Microsoft,” reads a pop-up that appears after you land on Google’s site.*

Three:

*“We love having you! Can you please take a minute to tell us why you are trying another browser?” reads a new pop-up poll.*

Four:

*“Microsoft Edge runs on the same technology as Chrome, with the added trust of Microsoft,” reads an injected ad that appeared after my download.*

Four.

*To be fair, it did appear after the download was finished.*

The pop-ups are nearly two years old, and the injected ads are from earlier this year. The poll seems to be new, though — Neowin reports it first saw the poll last weekend.

But hey, it could be worse:

I cannot believe how many stories we’ve written about the shit Microsoft has pulled to steer you away from Chrome. Even today, the company still won’t always respect your choice of default browser, though that may finally be changing in the EU.

Sadly the poll doesn’t include an “I’m boycotting Edge because you don’t respect me as a user” option.