Eshu Marneedi

OpenAI on Monday announced a slew of new additions to ChatGPT, its artificial intelligence chatbot, in a “Spring Update” event streamed in front of a live audience of employees in its San Francisco office. Mira Murati, the company’s chief technology officer, led the announcements alongside some engineers who worked on their development while Sam Altman, OpenAI’s chief executive, live-posted from the audience on the social media website X. I highly recommend watching the entire presentation, as it is truly one of the most mind-blowing demonstrations one will ever see. It is just 26 minutes long and is available for free on OpenAI’s website. But here is the rundown of the main announcements:

1. A new large language model, called GPT-4o, with “O” standing for “omni.” It is significantly speedier at producing responses than GPT-4 while being as intelligent as the older version of the generative pre-trained transformer.
2. A new, improved voice mode that integrates a live camera so ChatGPT can see and speak concurrently. Users can interrupt the robot while it speaks, and the model acts more expressively, tuning its responses to the user’s emotions.
3. A native ChatGPT application for macOS with which users can ask the chatbot questions with a keyboard shortcut, share their screen for questions, and ask ChatGPT about clipboard contents.

Again, the video presentation is compulsory viewing, even for the less technically inclined. No written summary will be able to describe the emotional rush felt while watching a robot act like a human being. The most compelling portion of the demonstration was when the two engineers spoke to the chatbot on an iPhone, through the app, and watched it rattle off eloquent, human-like responses to questions asked naturally. It really is something to behold.

However, something stuck out to me throughout the banter between the humans and the chatbot: the expressiveness. Virtual assistants, no matter how good their text-to-speech capabilities may be, still speak like inanimate non-player characters, in a way. Their responses are tailored specifically to questions posed by the users, but they still sound pre-written and artificial due to the way they speak. Humans use filler words, like “um,” “uh,” and “like” frequently; they take long pauses to finish thoughts before speaking them aloud; and they read and speak expressively, with each word sounding different each time. Emphasis might be placed on different parts of the word, it might be said at different speeds — the point is, humans do not speak perfectly. They speak like humans.

The new voice mode version of ChatGPT, ChatGPT 4o, speaks just like a real person would. It laughs, it takes pauses, it places emphasis on different parts of words and sentences, and it speaks loosely. It acts more like a compassionate friend than a professional assistant — it does not aim to be formal in any way, but it also tries to maintain some degree of clarity. For example, it won’t meander like a person may, but it does sound like it may meander. For example, when the chatbot viewed a piece of paper with the words “I ♥ ChatGPT,” it responded oddly carefully: “Oh, stop it, you’re making me blush!” Aside from the fact that robots cannot blush, the way it said “oh” and the space that came after it had the same expression and emotion that it would carry if a human had said it. The chatbot sounded surprised, befuddled, and flustered, even though it had prepared that response after solving essentially what was just a tough algebra problem.

Other instances, however, seemed pretty awkward: ChatGPT seemed very talkative in the demonstration, such as when the presenters made mistakes or asked the robot to wait a second. Instead of simply replying “Sure” or just firing back with an “mhmm” as a person would, it gave an annoyingly verbose answer: “Sure, I’d love to see it whenever you’re ready!” No person would speak like that unless they were trying to be extra flattering or appear overly attentive. It could be that ChatGPT’s makers programmed the robot to perform this way for the presentation just so that the audience could hear more of the Scarlett Johansson-esque voice straight from the movie “Her,” but the constant talkativeness broke the immersion and made me want to frankly tell it to quiet down a bit.

The robot also seemed oddly witty, as if it carried some sass in its responses. It wasn’t rude, of course, but it sounded like a very confident salesperson when it should’ve been more subdued. It liked to use words like “Whoops!” and added some small humor to its responses — again, signs of wordiness. I assume the reason for this is to make the robot sound more humanlike because awkward silences are unpleasant and may lead users to think ChatGPT is processing information or not ready to receive a request. In fact, while in voice mode, it’s always processing information and ready to receive requests. It can be interrupted with no qualms, it can be asked different questions, and it can wait on standby for more information. Because GPT-4o is so quick at generating responses, there is zero latency between questions, which is jarring to adjust to but also mimics personal interactions.

Because ChatGPT has no facial expressions, it has to rely on sometimes annoying audio cues to keep the conversation flowing. That doesn’t mean ChatGPT can’t sense users’ emotions or feelings, though — the “O” in GPT-4o enables it to understand tacit intricacies in speech. It can also use the camera to detect facial expressions, but the more interesting use was what it could do with its virtual apparatus. Not only can users speak to ChatGPT while it is looking at something by way of its “omni-modal” capabilities, but users can share their computer screens and make selections on the fly to receive guidance from ChatGPT as if it were a friend looking over their shoulder. An intriguing demonstration was when the robot was able to guide a user through solving a math equation, identifying mistakes as they were made on the paper without any additional input. That was seriously impressive. Another example was with writing code: ChatGPT could look at some code in a document and describe what it did, then make modifications to it.

ChatGPT 4o’s underlying technology is still OpenAI’s flagship GPT-4 LLM, which is still available for paying customers — though I wouldn’t know why one would use it as it’s worse and has lower usage limits. But the new LLM is now trained on audio and visual data in addition to text. Previously, as Murati described during the event, ChatGPT would have to perform a dance of transcribing speech, describing images, processing the information like a normal LLM text query, and then finally running the answer through a text-to-speech model. GPT-4o performs all of those steps inherently as part of its processing pipeline. It natively supports multimodal input and processes it naturally without performing any modifications. It knows what objects are in real life, it knows how people speak, and it knows how to speak like them. It is truly advanced technology, and I can’t wait to use it when it launches “in the coming weeks.”

While the concept of a truly humanlike chatbot is still unsettling to me, I feel like we’ll all become accustomed to assistants such as the one OpenAI announced on Monday. And I also believe they’ll be more intertwined with our daily lives due to their deep integration with our current technology like iPhones and Macs, unlike AI-focused devices (grifts) like the ones from Humane and Rabbit. (The new Mac app is awesome.) It’s an exciting, amazing time for technology.

Tim Nudd, reporting for Ad Age:

Apple apologized Thursday for a new iPad Pro commercial that was met with fierce criticism from creatives for depicting an array of creative tools and objects—from a piano, to a camera, to cans of paint—being destroyed by an industrial crusher.

The tech giant no longer plans to run the commercial on TV…

But many viewers had a more chilling interpretation, seeing the spot as a grim representation of technology crushing the history of human creativity—something the creative industry is already existentially worried about with the rise of AI.

In an exclusive statement obtained by Ad Age, Apple apologized for the “Crush” spot and said it didn’t mean to cause offense among its creative audience.

“Creativity is in our DNA at Apple, and it’s incredibly important to us to design products that empower creatives all over the world,” said Tor Myhren, the company’s VP of marketing communications. “Our goal is to always celebrate the myriad of ways users express themselves and bring their ideas to life through iPad. We missed the mark with this video, and we’re sorry.”

The spot rolled out on Apple’s YouTube and CEO Tim Cook’s X account on Tuesday, but had not received any paid media. Plans for a TV run have now been scrapped.

This is the video in question. Two things:

1. This is the first time I have seen Apple pull an advertisement from the airwaves in recent memory. The backlash was fierce this time around, with many feeling frustrated and upset at the (terrible) visual of these beautiful pieces of technology and instruments being crushed by what looked like a hydraulic press. I understand what Apple was aiming for here — that the new iPad Pro is powerful enough to replace all of these tools while being remarkably thin — and in a way, the imagery fits the theme. But in practice, looking at the commercial is just sad. I understand why so many professionals — the target market for the advertisement, too — were disturbed by this video, and I think Apple made the right decision here. I appreciate how the company has handled this situation; it takes courage to remove the main commercial for a star product just a day after it was announced.

2. When I first viewed the advertisement during Apple’s Tuesday event, I wasn’t very perturbed by it, but that was mostly because I wasn’t paying much attention. But after Cook posted the video on the social media website X, I watched it again after reading some posts from filmmakers and other creators about how it made them feel, and I was suddenly uneasy. This commercial comes at a time when much of the creative industry is alarmed by the advent of generative artificial intelligence. For their precious tools, like guitars, pianos, and paints, to be destroyed and replaced by a slender tablet marketed as an “AI-focused” device is cruel. I think Apple could’ve instead offered a brighter picture of how the new iPad Pro could be used, featuring creators in their working spaces using the iPad to enhance their workflows. Nobody is seriously going to throw out their drum kit to replace it with the AI-powered drummer in the new version of Logic Pro announced Tuesday, so why advertise the device like that?

Apple, in the words of Myhren, the company spokesperson, truly did “miss the mark.” It’s unusual coming from Cupertino, which typically makes the very best awe-inspiring advertisements. For example, I thought the digital campaign that followed the event comparing the new iPad Pro to a teal iPod nano was great — it is peak Apple; just as Steve Jobs would’ve intended. I know Apple values and loves physical, antique objects, even if they’re from another era — just look at how much the company celebrates its history in so many of its advertisements. I don’t know why the team tasked with producing this commercial chose to portray the new iPad Pro this way in a stunning deviation of decorum.

The thinnest, most powerful iPads take center stage

Let loose. Image: Apple.

Apple on Tuesday announced updates to its iPad lineup, including refreshed iPad Air and iPad Pro models, adding a new, larger size to the iPads Air and new screen technology and processors to both new iPads Pro. The company also announced new accessories, such as a new Apple Pencil Pro and Magic Keyboard for the iPads, as well as software updates to its Pro apps on iPadOS. The new announcements come at a time when Apple’s iPad lineup has remained stagnant — the company has not announced new tablets since October 2022, when the iPad Pro was last updated with the M2 chip. On Tuesday, Apple gave the iPad Air the M2 — an upgrade from the previous M1 from when it was last updated in 2022 — and the iPad Pro the M4, a new processor with more cores, a custom Display Engine, and enhanced Neural Engine for artificial intelligence tasks.

Most iPad announcements as of late aren’t particularly groundbreaking — more often than not, iPad refreshes typically feature marginal improvements to battery life and processors, and Apple usually resorts to rehashing old iPadOS feature announcements during its keynotes to fill the time. Tuesday’s event, however, was a notable exception: Apple packed the 38-minute virtual address chock full of feature enhancements to the high-end iPads, with Tim Cook, the company’s chief executive, calling Tuesday “the biggest day for iPad since its introduction” at the very beginning of the event. I tend to agree with that statement: The iPad Pro, for the first time ever, debuted with a new Mac Apple silicon processor before the Mac itself; it now features a “tandem” organic-LED display with two panels squished together to appear brighter; and it’s now thinner and lighter than ever before. These are not minor changes.

But, as I’ve said many times before, I think the biggest limitation to the iPad’s success is not the lack of killer hardware, but the lack of professional software that allows people to create and invent with the iPad. While Apple’s “magical sheet of glass” is now “impossibly thin” and more powerful than Cupertino’s lowest-end $1,600 MacBook Pro announced just last October, its software, iPadOS, continues to be worthless for anything more than basic computing tasks, like checking email or browsing the web. And while the new accessories, like the new Magic Keyboard made out of aluminum featuring a function row, are more professional and sturdy, they still don’t do anything to make the device more capable for professional users. Add to that the $200 price increase — the base-model 11-inch iPad Pro now starts at $1,000, while the larger 13-inch model starts at $1,300 — and the new high-end iPads feel disappointing. I don’t think the new iPads Pro are bad — they’re hardly so — or even a bad value, knowing how magical the iPad feels, but I wish they did more software-wise.

Here are my takeaways from Tuesday’s “Let loose” Apple event.

iPads Air

The easiest-to-cover announcement was the new iPads Air — plural. Before Tuesday, the iPad Air — Apple’s mid-range tablet — only came in one size: 10.9 inches. Now, the device comes in two sizes: the same 11-inch smaller version, and a new 13-inch form factor. Aside from the size, the two models are identical in their specifications. Both models feature M2 chips, their cameras have been relocated to the horizontal edge to make framing easier due to how most users hold iPads, and new storage options have been added now up to 1 terabyte. The smaller model’s prices also remain the same, starting at $600, and the 13-inch version sells for $750. Starting storage has also been increased to 128 gigabytes, and there is now a 512-GB variant.

The new iPads Air, otherwise, are identical to the last-generation model, with the same camera and screen resolutions and mostly identical accessories support. The first-generation Magic Keyboard from 2020 remains compatible, but the second-generation Apple Pencil from 2018 that worked with the previous model is not. (More on this later.) They both come in four colors — Space Gray, Blue, Purple, and Starlight — and ship May 15, with pre-orders open on Tuesday.

I am perplexed by the iPads Air, particularly the smaller version, which is often more expensive than a refurbished last-generation iPad Pro of the same size. Choosing to buy the latter would be more cost-effective, and the iPads Pro also have Face ID and a 120-hertz ProMotion display. Add to that the better camera system and identical processor, and I truly don’t see a reason to purchase a new (smaller) iPad Air. The larger model is a bit of a different case, since buying a larger refurbished iPad Pro would presumably be more expensive, so I can understand if buyers might want to buy the newer 13-inch iPad Air for its larger screen, but the low-end model continues to be a fantastically bad value.

The M4

Rather than use October’s M3 processor in the new iPads Pro, Apple revealed a new system-on-a-chip to power the new high-end tablet: the M4. Exactly as predicted by Mark Gurman, a reporter at Bloomberg with an astonishing track record for Apple leaks¹, the new M4 is built on Taiwan Semiconductor Manufacturing Company’s enhanced second-generation 3-nanometer fabrication process called N3E. The new process will presumably provide efficiency and speed enhancements, but I think they will be negligible due to iPadOS’ limited feature set and software bottlenecks. The processor, by default, is binned² to a nine-core central processing unit — with three performance cores and six efficiency cores — and a 10-core graphics processor, but users who buy the 1- or 2-TB models will receive a non-binned 10-core CPU with four performance cores. The low-end storage tiers also only have 8 GB of memory, whereas the high-end versions have 16 GB, though both versions still have the same memory bandwidth at 120 gigabytes per second.

John Ternus, Apple’s senior vice president of hardware engineering, repeatedly mentioned during the event that the new iPad Pro would not be “possible” without the M4 chip, but I struggle to see how that is true. The new processor has what Apple calls a “Display Engine,” which Apple only made a passing reference to, presumably because it is not very impressive. As far as I know, the M3’s “Display Engine,” so to speak — which is already present in MacBooks Pro with the M3 — powers two external displays, so I’m having a hard time understanding what is so special about the OLED display found in the new iPads that warrants the upgraded, dedicated Display Engine. (It isn’t even listed on Apple’s “tech specs” page for the iPads Pro, for what it’s worth.)

Whatever the Display Engine’s purpose may be, Apple claims the M4 is “1.5 times faster” in CPU performance than the M2, though, once again, I don’t see a reason for the performance improvements because iPadOS is so neutered compared to macOS. I have never had a performance issue with my M2 iPad Pro, and I don’t think I will notice any difference when I use the M4 model. Other than for the cynical reason of trying to shift more iPad sales during Apple’s next fiscal quarter, I don’t see a reason for the M4’s existence at all. I’m unsurprised by its announcement, but also awfully confused. Expect to see this processor in refreshed Mac laptops in the fall, too.

iPads Pro

The star of the show, per se, was the new iPad Pro lineup, both the 11-inch and 13-inch models. (There is no longer a “12.9-inch” model, which I am grateful for.) Both models have been “completely redesigned” and feature new displays, cases, processors, and accessories. The update is the largest since the complete redesign and nixing of the Home Button and Lightning port in 2018, but it isn’t as monumental as that year’s revamp. From afar, the new models look identical to 2022’s versions, aside from the redesigned camera arrangement, which is now color-matched to the device’s aluminum body à la iPhones, whereas it was previously just made out of black glass. The displays are now “tandem OLED” panels, which use a special technology to fuse two OLED panels for maximum brightness and earn the display a new name of “Ultra Retina XDR.” (The iPhone’s non-tandem OLED display is called “Super Retina XDR,” and the previous generation’s 12.9-inch model’s mini-LED display was called the “Liquid Retina XDR” display.) And just like the iPads Air, the iPads Pro’s front-facing camera has been relocated to the horizontal position.

Most impressively of all, Apple managed to thin the iPads down significantly from their previous girth. Apple, in a Jony Ive-like move, called the new 13-inch model the “thinnest device” it has “ever made” — even thinner than the iPod nano, which held the title previously. Ternus, the Apple executive, also assured that the device didn’t compromise on build quality or durability, though I would imagine the new model is easier to bend and break than before. (Tough feat.) I do not understand the obsession over thinness here, but the new model is also lighter than ever before due to the more compact OLED display. The new iPads Pro are so thin that the Apple Pencil hangs off the edge when magnetically attached to the side, which may be inconvenient when the iPad is set on a table; Thunderbolt cables plugged into the iPad also protrude upward from the body, a consequence of the sheer thinness. One thing is for certain, however: The new iPads Pro do look slick, especially in the new Space Black finish.

The thinness is a byproduct — or consequence, rather — of the new beautiful OLED display found on both models, replacing the LED “Liquid Retina” display of the last-generation 11-inch model and mini-LED display of the 12.9-inch version. While the mini-LED display was able to reproduce high-dynamic-range content with high brightness levels down to a specific “zone” of the panel, it also suffered from a phenomenon called “blooming,” where bright objects on a dark background would display a glowing halo just outside of the object. OLED displays feature individually lit pixels, allowing for precise control over the image, alleviating this issue. The panel’s specifications are impressive on their own: 1,000 nits of peak brightness when displaying standard-dynamic-range content, 1,600 nits of peak localized brightness when content is in HDR, a two-million-to-one contrast ratio, and a ProMotion refresh rate from 10 hertz to 120 hertz. The new display, as Apple says, truly is “the most advanced display in a device” of the iPad’s kind. I would argue it’s one of the most advanced displays in a consumer electronics device, period, aside from probably Apple’s own Vision Pro. It truly is a marvel of technological prowess, and Apple should be proud of its work.

Apple allows buyers who purchase a 1- or 2-TB model the option to coat the display in a nano-texture finish for a $100 premium, which will virtually eliminate glare and provide for a smoother writing and drawing experience when using the Apple Pencil. The finish is the same as found on the Pro Display XDR and Studio Display, and while I don’t think it is for me, I appreciate the option. (I do wonder how wiping away fingerprints would work, though, since this is the first time Apple has applied the coating to a touch device.) One quirk of the nano-texture coating, however, is that it cannot cover the Face ID sensors, located at the side of the iPad Pro, so the finish stops at the edge of the screen itself, displaying a glossy bezel around the display. I think it looks strange, but this problem couldn’t possibly be alleviated without redesigning Face ID entirely.

Apple has made some noteworthy omissions to the product, however. Most distinctly of all, it has removed the ultra-wide lens at the back of the iPad, a lens it added in the product’s 2020 refresh. Personally, I have never once touched the ultra-wide camera, and I don’t know of anyone who did, but it might be missed by some. To compensate, Apple has added a new AI-powered shadow remover to the document scanner in iPadOS, powered by the M4’s improved Neural Engine and a new ambient light sensor, which takes a prominent space in the iPad’s new camera arrangement. I’m unsure about how I feel about its physical attractiveness — there are only so many ways to design a camera on a tablet computer before it gets boring — but I think the swap is worth the trade-off. (The ultra-wide camera at the front added in 2021, which powers Center Stage, has not been removed.) The SIM card slot has also been removed from cellular-equipped models, mirroring its omission from 2022’s iPhone 14 Pro, and the 5G millimeter-wave antenna located at the side has also been axed reportedly due to a lack of usage.

The new models have both received price increases of $200, with the 11-inch model starting at $1,000, and the 13-inch at $1,300. I think those prices are fair; I expected increases to be more substantial due to the cost of OLED panels. Base storage amounts have also been subsequently bumped; the new models begin with 256 GB of storage and are configurable up to 2 TB. They ship May 15, just like the iPads Air, and are available for pre-order beginning Tuesday.

Hardware-wise, the new iPads Pro are truly some of the most impressive pieces of hardware Apple has manufactured yet, and I’m very excited to own one. But I can’t help but ask a simple question about these new products: why? Apple has clearly dedicated immense time, energy, and money to these new iPads, and it’s very apparent from the specifications and advertising. Yet when I unbox my new iPad come next week, I’ll probably use it the same, just as I have always used my iPad. It won’t be any better at computing than my M2 iPad Pro I’ve owned for the last year and a half. The Worldwide Developers Conference in June is where the big-ticket software announcements come, but just as Parker Ortolani, a product manager at Vox Media, said on Threads, we have collectively been waiting for “the next big iPadOS update” since the first iPad Pro was launched in 2015 — before iPadOS even existed. iPadOS is a reskinned version of iOS, and Apple must change that this year at WWDC. Until then, the new iPads, while spectacular from every imaginable hardware angle, lack a purpose.

Apple Pencil Pro and Magic Keyboard

Apple announced updates to its two most popular accessories for the iPad Air and iPad Pro: the Apple Pencil, and the Magic Keyboard. The second-generation Apple Pencil, first announced in 2018, has remained unchanged since its first debut and has been compatible with all high-end iPads since 2020, and the Magic Keyboard, first announced in 2020, has also been kept untouched. Both products on Tuesday received major overhauls: Apple debuted the Apple Pencil Pro, a new product with haptic feedback and a touch sensor for more controls, and the new Magic Keyboard, which is now finished in a sturdier aluminum, has a function key row, and features a redesigned hinge. Both products are strictly only compatible with Tuesday’s iPads; subsequently, prior versions of the Apple Pencil and Magic Keyboard cannot be used with the new iPads Pro or iPads Air, aside from the USB Type-C Apple Pencil released in October, which remains as a more affordable option.

The Magic Keyboard’s redesign, Apple says, makes it a more versatile option for “professional” work on iPadOS. The keys, using the same scissor-switch mechanism as the previous generation, now have a more tactile feel due to the hefty aluminum build, which also adds rigidity for use on a lap — the lack thereof was a pitfall of the older Magic Keyboard. The trackpad is now larger and features haptic feedback, just like Mac laptops, and the hinge is more pronounced, making an audible click sound when shut. The Magic Keyboard also adds a small function row at the very top of the deck, adding an Escape key for anyone bullish enough to code on an iPad. (This would’ve been a great time to put Terminal on iPadOS.) While the new additions will undoubtedly add weight to the whole package, I think the trade-off is worth it because it makes the iPad feel more like a Mac. The new Magic Keyboard retails for the same price: $300 for the 11-inch version, and $350 for the 13-inch one. It, again, ships May 15, with pre-orders available Tuesday.

The Apple Pencil Pro, while not as visually striking of an upgrade as the Magic Keyboard, does build on the foundation of the second-generation Apple Pencil well. That stylus, which Apple still sells for older iPads, features a double-tap gesture, which allows quick switching between drawing tools, such as the pen and eraser. The new stylus builds on the double-tap feature, adding a touch sensor to the bottom portion of the stalk which can be squeezed and tapped for more options. Instead of only double-tapping the pencil, users are now able to squeeze it to display a palette of writing tools — not just the eraser. This integration works in apps that support the new PencilKit features in iPadOS; for those that don’t, the double-tap gesture works just as it did before. To select a tool, it can simply be tapped on the screen like normal with the pencil.

The Apple Pencil Pro also supports a feature called “barrel roll,” which allows users to move their fingers in a circle around the pencil to finely control its angle on the virtual page, just like someone would do with a real pencil. And when squeezing, double-tapping, or using the barrel roll gesture, a new Haptic Engine added to the pencil will provide tactile feedback for selections. Apple also added Find My functionality to the pencil, though it is unclear if it included Precision Finding, the feature that utilizes the ultra-wideband chip in recent iPhones to locate items down to the inch. (I don’t think it did since the iPad doesn’t have a U2 chip.)

The Apple Pencil Pro retails for $130 — the same price as the second-generation Apple Pencil — and is available for pre-order starting Tuesday, with orders arriving May 15. The more comedic aspect of this launch, however, is the new Apple Pencil Compare page on Apple’s website, which looks genuinely heinous. Apple now produces and sells four different Apple Pencils, all with separate feature sets and a hodgepodge of compatibility checks. To review:

• Apple Pencil Pro: The latest version is compatible with the M2 iPads Air and M4 iPads Pro announced Tuesday. It retails for $130.
• Second-generation Apple Pencil: The older version of the Apple Pencil is compatible with iPads Pro from 2018 and newer and the fourth- and fifth-generation iPads Air from 2020 and 2022. It is not compatible with any of the new iPads announced Tuesday. It also sells for $130.
• USB-C Apple Pencil: The new USB-C Apple Pencil from October, which does not have double-tap or pressure sensitivity, is compatible with every iPad with a USB-C port, including the latest models. It is available for $70.
• First-generation Apple Pencil: This pencil is for compatibility with older, legacy iPads, as well as the now-discontinued ninth-generation iPad. It costs $100.

No reasonable person will choose to remember that information, so Apple has assembled an Apple Pencil compatibility page, which is absolutely abhorrent. There is even a Contact Us link on the page for those who need assistance to figure out the chaos. “Who wants a stylus?”

Conclusion

As I have stated many times throughout this article, I think the new hardware announced Tuesday is spectacular. The new iPads Air fit in well with the lineup, the 10th-generation iPad has received a price reduction of $50, replacing the archaic ninth-generation model which had a Home Button and Lightning port, and the new iPads Pro are marvels of engineering. I think all models are well-priced, I like the new design of the Magic Keyboard, and I’m thankful the Apple Pencil has been updated.

But none of the above overshadows how disappointed I am in the iPad’s software, iPadOS. As good as the new hardware may be, I don’t think I will use it any differently as I do my current iPad now. That’s a shame — for how much work was put into Tuesday’s announcements, the bespoke software for the iPad should do better. Until then, the iPad will continue to remain a product in Apple’s lineup — nothing more, and nothing less.

A correction was made on May 5, 2024, at midnight: An earlier version of this article stated that the new M2 iPad Air supports the second-generation Apple Pencil. That is not true; it only supports the USB-C Apple Pencil and the new Apple Pencil Pro. I regret the error.

A correction was made on May 14, 2024, at 2:11 a.m.: An earlier version of this article stated that the USB-C Apple Pencil was released in March. It was actually released in October of last year. I regret the error.

Ben Smith for Semafor interviewed Joe Kahn, the executive editor of The New York Times. Here is what Kahn had to say in response to Smith’s question about The Times’ role in saving democracy:

It’s our job to cover the full range of issues that people have. At the moment, democracy is one of them. But it’s not the top one — immigration happens to be the top [of polls], and the economy and inflation is the second. Should we stop covering those things because they’re favorable to Trump and minimize them? I don’t even know how it’s supposed to work in the view of Dan Pfeiffer or the White House. We become an instrument of the Biden campaign? We turn ourselves into Xinhua News Agency or Pravda and put out a stream of stuff that’s very, very favorable to them and only write negative stories about the other side? And that would accomplish — what?

I think The New York Times has completely misunderstood what “independent journalism” is. Kahn and other Times journalists, whose work I read regularly, think of us — those accusing The Times of journalistic malpractice — as wanting them to favor the Biden administration or to be against former President Donald Trump somehow. That couldn’t be farther from the truth — it is my firm belief that news shouldn’t be biased toward a political candidate.

News, however, should be biased toward the truth, and The Times warps the truth however it wants to fit the public’s narrative. That’s exactly what Kahn is doing here by using the polls as a determinant for what to cover and how to cover it. I understand the core message: that America’s most respected newspaper should cover America’s problems. But, oftentimes, America’s problems and the way it interprets them are disconnected from reality. It is the job of the country’s newspapers of record to influence public opinion, not report on only what Americans seem to care about.

It’s the job of the news media to report the facts without subjectivity, and Kahn clearly knows this and restates it multiple times throughout the interview. But, Kahn also released this piece of truth: “I think the general public actually believes that he’s responsible for these wars, which is ridiculous, based on the facts that we’ve reported,” referring to President Biden. If the public, by Kahn’s own admission, is so foolish to believe Biden started the wars in Europe and the Middle East, why should The Times’ newsroom cover reality through the public’s (incorrect) lens, as Kahn says The Times is doing?

The Times’ job is to cover reality, regardless of whether it favors the incumbent or his predecessor. Currently, it’s not doing that. It’s warping the news to please its audience, which is not news-making. Once again, my request is not for The Times to be a knight defending democracy by praising Biden’s every move. I want it to be objective in its reporting. Currently, it isn’t — and I feel like that is on purpose.

Apple announced its earnings for the second quarter on Thursday, and Tim Cook, the company’s chief executive, interviewed with CNBC. CNBC wrote the following:

Cook also said Apple has “big plans to announce” from an “AI point of view” during its iPad event next week as well as at the company’s annual developer conference in June.

I don’t even understand why this was reported on, because artificial intelligence is the new craze both in Silicon Valley and Wall Street. Of course the chief executive of the world’s second-largest technology company — which reported revenue down 4 percent this quarter — would try to pump his stock price, and of course he would do that by saying there will be an AI-related announcement at next week’s hotly anticipated Apple event. It makes logical sense from a business perspective: If Cook can motion investors to hold off on dumping Apple stock this week, he can launch new iPads next week, point to the sales numbers, and watch the stock hike again. That is his job.

Later, CNBC retracted its original quote, but gave the full context to Zac Hall, editor at large at 9to5Mac, somehow:

We’re getting into a period of time here where we’re extremely excited like I’m in the edge of my seat literally because next week, we’ve got a product event that we’re excited about. And then just a few weeks thereafter, we’ve got the… Worldwide Developers Conference coming up and we’ve got some big plans to announce in both of these events. From an AI point of view…

Cook is not saying there will be AI-related announcements at these events, he is just saying (a) that there are “big plans” and (b) there will be announcements some time between now and the end of eternity “from an AI point of view.” Those are mutually exclusive statements — it is foolish to assume otherwise because Cook is well-trained before he sits in front of the media. Apple never reveals what it will announce before an event, even when it would be in the interest of the stock price.

So, that all begs the question: Will there be AI at next week’s event or not? It’s impossible to say conclusively, but I think there will certainly be mentions of AI during the presentation. However, I do not believe Apple will announce AI software of its own just a month before WWDC, where software is usually debuted. I imagine the AI references will be limited to passing mentions of how the new iPads Pro are “great for AI computing” and how you can run AI models with apps on the App Store, just like the “Scary fast” Apple event from October, where the company announced the M3 MacBooks Pro. The mentions will exist to please investors and to hold them off just a bit longer for WWDC, where the big-ticket AI features will be introduced via iOS 18.

Thursday’s keynote will not be a preview of AI features — or at least, so I think. Instead, it looks like it’ll serve as a filler event to build anticipation for the true announcements coming in the summer, while also finally refreshing the iPads, which is long overdue. This scenario also takes into account Mark Gurman’s report for Bloomberg on Sunday that said Apple will ship the M4 in the new iPads Pro: M4 or not, this event is slated to be hardware-focused, and I think the only AI references next week will exist to appease Wall Street. My final take: No AI at next week’s event.

Mishaal Rahman, reporting for Android Authority on Tuesday:

If everything an AI gadget like the Rabbit R1 can do can be replicated by an Android app, then why aren’t these companies simply releasing an app instead of hardware that costs hundreds of dollars, requires a separate mobile data plan to be useful, and has terrible battery life? It turns out that’s exactly what Rabbit has done… sort of.

See, it turns out that the Rabbit R1 seems to run Android under the hood and the entire interface users interact with is powered by a single Android app. A tipster shared the Rabbit R1’s launcher APK with us, and with a bit of tinkering, we managed to install it on an Android phone, specifically a Pixel 6a.

Once installed, we were able to set up our Android phone as if it were a Rabbit R1. The volume up key on our phone corresponds to the Rabbit R1’s hardware key, allowing us to proceed through the setup wizard, create a “rabbithole” account, and start talking to the AI assistant. Since the Rabbit R1 has a significantly smaller and lower resolution display than the Pixel 6a, the home screen interface only took up a tiny portion of the phone’s display. Still, we were able to fire off a question to the AI assistant as if we were using actual Rabbit R1 hardware, as you can see in the video embedded below.

The Rabbit R1, just like the Humane Ai Pin, is nothing more than a shiny object designed to attract hungry venture capitalists. The entire device is an Android app, a low-end MediaTek processor, and a ChatGPT voice interface wrapped up in a fancy orange trench coat — in other words, nothing more than a grift that retails for $200. I’ve said this time and time again: These artificial intelligence-powered “gadgets” are VC money funnels whose entire job is to turn profits then disappear six months later when Apple and Google add more broad AI functionality to their mobile operating systems. In the bustle of the post-October 2022 AI sphere, Rabbit raised a few million dollars in Los Angeles, built together an Android app with a rabbit animation, bulk bought some off-the-shelf cheap electronics from China, engineered a bright orange case, put the parts together, made its founder dress up like an off-brand Steve Jobs, and poof, orders started flooding in by the thousands. Ridiculous.

The Rabbit R1, in many ways, is more insulting than the Humane Ai Pin, which I’ve already bashed enough. It is significantly more affordable, priced at $200 with no subscription — unlike Humane’s $700, $24-a-month product — but it is quite literally worse than the Ai Pin from Rabbit’s chief rival VC funnel in every metric. The entire device, as Marques Brownlee, a YouTuber better known as MKBHD, demonstrated in his excellent review of the device, is a ChatGPT wrapper with an ultra-low-end camera and a knob — or wheel, rather — used in favor of a touch screen presumably to make it seem less like a smartphone. In practice, it is a bad, low-end smartphone that does one thing — and only one task — extraordinarily poorly, consistently flubbing answers and taking seconds to respond. It is a smartphone that does everything poorly aside from looking great. (Teenage Engineering designed the Rabbit R1; I’ll give the product design props.) I am astonished that we are living in a world where this $200 low-end Android smartphone is receiving so much media attention.

Rahman contacted Jesse Lyu, Rabbit’s chief executive and co-founder, for comment on his article, and Lyu, grifter-in-chief at Rabbit, naturally denied the accusations in the stupidest way possible. I don’t even understand how this made it to publication; it’s genuinely laughable. Lyu’s justification for the device is that Rabbit sends data and queries to servers — presumably its own servers — for processing. Here is a non-comprehensive list of iOS apps with large language models built in that send data to the web for processing: OpenAI’s ChatGPT, Microsoft Copilot, Anthropic Claude, and Perplexity — also known as every single AI processing app made by a large corporation because it is all but impossible to run LLMs on even the most sophisticated, powerful smartphone processors, let alone any random inexpensive MediaTek chip, such as found on the R1. The Rabbit R1 is an Android app that exchanges data with the internet with a cellular radio and some network calls. Any 15-year-old could engineer this in two weeks from the comfort of their bedroom.

I aggressively smeared the Humane Ai Pin not because I thought it was a grift, but because I thought it had no reason to exist. I thought and still think that Humane built an attractive piece of hardware and that the company still has conviction in creating a product akin to the smartphone in the hopes of eventually eclipsing it. (I think this entire idea is flawed, and that Humane will eventually go bankrupt, but at least Humane’s founders are set on their ambition.) Rabbit as an entire company, by stark contrast, is built on a throne of lies and scams: It came out of the woodwork randomly during the Consumer Electronics Show in January after raising $10 million the month prior from over-zealous VC firms, threw a launch party in New York with influencers and press alike, then shipped an Android app to consumers for $200. It’s a cheap smear of hard-working, dedicated hardware markers; it makes a mockery of true innovators in our very complicated technology climate in 2024. These “smartphone replacement” VC attractions ought to be bankrupt by, if not right after, June.

Ridiculous, but quite possible. Mark Gurman, reporting for Bloomberg in his Power On newsletter:

Earlier this month, I broke the news that Apple is accelerating its computer processor upgrades and plans to release the M4 chip later this year alongside new iMacs, MacBook Pros, and Mac minis. The big change with the M4: A new neural engine will pave the way for fresh AI capabilities. Now here’s another development. This year’s Macs may not be the only AI-driven devices with M4 chips.

I’m hearing there is a strong possibility that the chip in the new iPad Pro will be the M4, not the M3. Better yet, I believe Apple will position the tablet as its first truly AI-powered device — and that it will tout each new product from then on as an AI device. This, of course, is all in response to the AI craze that has swept the tech industry over the last couple years.

By introducing the new iPad Pro ahead of its Worldwide Developers Conference in June, Apple could lay out its AI chip strategy without distraction. Then, at WWDC, it could focus on how the M4 chip and new iPad Pros will take advantage of the AI software and services coming as part of iPadOS 18 later this year. I fully expect Apple to position the A18 chip in the iPhone 16 line as built around AI as well.

To be fair, though, these new products aren’t engineered and developed entirely around AI. This is partly about marketing. Hardware with even more impressive capabilities is further out. As I’ve reported, Apple is working on a table-top iPad connected to a robotic arm, as well as a home robot.

For context, the M3 line of processors debuted in late October last year, so it has only been roughly six months since the latest generation of Apple’s high-end processors came into the market. Every single bone in my body disagrees viscerally with every aspect of this rumor — it does not make sense from any logical perspective whatsoever because there is no way Apple would sell an iPad Pro that is faster and more capable than the MacBook Air and base-model MacBook Pro. It would be genuinely embarrassing for it to sell a device that runs iPadOS — a moderately enhanced version of iOS — with a more powerful chip than the $1,600 MacBook Pro.

That bit of illogical thinking is, however, small compared to the timeframes we’re working with here: Apple has never produced two full generations of Apple silicon only six months apart. With Taiwan Semiconductor Manufacturing Company’s factories booked trying to meet M3 3-nanometer process node demand, I have a tough time believing TSMC can fabricate more 3-nm chips to meet the demand for iPads Pro and Mac laptops. Also, the M3 processor is based on the architecture of the A17 Pro, which first debuted in September in iPhones 15 Pro, so the M4 would have to be based on the eventual A18 Pro (or A18, whatever it may be called), which has not even been announced yet — it will be announced this September. Historically, Apple has always based the Mac’s Apple silicon chip on that year’s iPhone chip.

And about the “artificial intelligence” focus: I genuinely can’t see Apple marketing the new iPads Pro — which are slated to debut at Apple’s virtual May 7 event — as “AI-focused” without first announcing AI features as part of iPadOS in June, at the Worldwide Developers Conference. What would potential buyers do with the AI-focused Neural Engine in iPadOS? This entire rumor is bending my mind because it seems so impossible and brazen. The M4, historically speaking, has no business being in these new iPads Pro — period — because no publicly available software can take advantage of the new Neural Engine on iPadOS, the M4 would stretch TSMC’s fabrication facilities to their max, and the M4’s debut in May would not line up with Apple’s product timelines. It’s completely nonsensical.

I trust Gurman — I truly do. There is not a time in recent memory when he has been wrong. But Gurman has been seesawing non-stop on this Apple event, which he earlier said wouldn’t even be an event in the first place. He also said that the iPads would be announced in March or April, but the event is taking place in May. Though he corrected some of these rumors later on, I heavily doubt his reporting here. If Apple truly does announce the M4 chip in the new iPads Pro in May, I won’t be shocked, because Gurman said it — but for now, I’m choosing to take this rumor with a grain of salt.

Also, one final note from Gurman’s newsletter this week, serving as some follow-up to my writing on Saturday about Apple Vision Pro:

Vision Pro demand has dropped considerably at many Apple stores. One retail employee says they haven’t seen one Vision Pro purchase in weeks and that the number of returns equaled the device’s sales in the first month that it was available.

Yikes.

Apple emailed me on Friday asking me how I’ve been enjoying my Apple Vision Pro. Here is what I wrote to the company.

Apple: Tell us why you’re not satisfied with your Apple Vision Pro.

Me: My main problem with Apple Vision Pro is its lack of content. Plain and simple, there is not much to do with it. I bought it because part of my job is to write about technology, but I probably wouldn’t have if I didn’t have that motive. Apple Vision Pro, a few months later, suffers from a lack of a use case. Everyone knows what to do with their iPhone, Mac, or iPad — but Apple Vision Pro? You might watch a movie, play a game or two, or fiddle around with the operating system, visionOS. But it’s restricted computing-wise in the same way the iPad is, which makes it impossible to use as a Mac replacement; it’s not sharable in the same way a TV is; and it’s not even nearly as easy to use as the iPad or iPhone, both of which can be used just by picking them up and tapping the screen.

Apple Vision Pro, from the get-go, is complicated to use. You have to ensure it has enough charge to use it since it doesn’t have very good standby battery life, then connect the battery, adjust the strap, place it on your head, adjust your hair, adjust the strap again, ensure it fits well, then unlock it and begin using it. For all of that to be worth it, there needs to be a seriously compelling reason to put it on. If something can be done 90 percent as enjoyably with an iPad or iPhone, most people — including myself — will just use that over Apple Vision Pro. Each time I use it, it’s a game of calculus: Is it worth it to do all this for the 10 percent of joy I’ll get?

What Apple wants people to think is that Apple Vision Pro does more than a traditional computer — not that Apple Vision Pro does what a traditional computer does but better. In practice, Apple Vision Pro does what a normal computer or mobile device does — but exponentially worse. Not only is it a hassle to use most of the time, but its software — based on iPadOS — functions in the same crippled ways iPadOS does. And with the lack of enthusiasm from third-party developers, the product is even more lackluster.

None of this is to say Apple Vision Pro is a lackluster product — it clearly isn’t. Every time I use it, I generally enjoy my time fiddling with new applications and experiences. But Apple sells many computers in various form factors, and most of those devices do the job of Apple Vision Pro just good enough that it’s not worth the effort to wear the headset most of the time. This is a solvable problem: Just make more content. Apple needs to incentivize third-party developers to make more experiences, produce more content itself, and improve the software experience to make computing better. For example, multitasking is impossible on visionOS, even though the inherent nature of Apple Vision Pro could make it a more capable Mac with an infinite amount of screen space. Why is using windows such a hassle on visionOS when they could be spectacular on this revolutionary spatial computer?

I particularly enjoyed the Major League Soccer highlight reel published in March and some of the other Apple-made immersive videos available through the TV app on visionOS. There should be much, much more of that kind of content available for paying subscribers. I know Apple has enthusiasm for this product, but looking at visionOS does not make that apparent.

Apple: What types of video content are you most interested in watching on Apple Vision Pro?

Me: Immersive video content, such as the MLS soccer highlight reel, available to all Apple Vision Pro users. Flat, 2D content isn’t as appealing because other devices can view it just fine, but immersive content is absolutely outstanding.

Apple: What one thing, if anything, would you add to or change about Apple Vision Pro?

Me: Make it lighter. The weight adds so much discomfort to using the product. It’s hard to use while lying down, uncomfortable while perfectly upright, and moderately uncomfortable while in a slightly reclined position — which is currently the most advisable way to use it. It rests on the cheeks and forehead evenly, but also terribly. It’s fatiguing to use for long periods of time. It needs to be lighter.

Sapna Maheshwari, David McCabe, and Cecilia Kang, reporting for The New York Times:

Just over a year ago, lawmakers displayed a rare show of bipartisanship when they grilled Shou Chew, TikTok’s chief executive, about the video app’s ties to China. Their harsh questioning suggested that Washington was gearing up to force the company to sever ties with its Chinese owner — or even ban the app.

Then came mostly silence. Little emerged from the House committee that held the hearing, and a proposal to enable the administration to force a sale or ban TikTok fizzled in the Senate.

But behind the scenes, a tiny group of lawmakers began plotting a secretive effort that culminated on Wednesday, when President Biden signed a bill that forces TikTok to be sold by its Chinese owner, ByteDance, or risk getting banned. The measure, which the Senate passed late Tuesday, upends the future of an app that claims 170 million users in the United States and that touches virtually every aspect of American life.

For nearly a year, lawmakers and some of their aides worked to write a version of the bill, concealing their efforts to avoid setting off TikTok’s lobbying might. To bulletproof the bill from expected legal challenges and persuade uncertain lawmakers, the group worked with the Justice Department and White House.

And the last stage — a race to the president’s desk that led some aides to nickname the bill the “Thunder Run” — played out in seven weeks from when it was publicly introduced, remarkably fast for Washington.

“Thunder Run” is a McCarthyist First Amendment violation straight from the Second Red Scare. This law was only able to pass because it was attached to a much-needed foreign aid appropriations bill, funding Ukraine and Israel and providing billions of dollars of humanitarian aid to vulnerable populations. It was included to ensure broad support within the Republican Party — a compromise House Speaker Mike Johnson of Louisiana had to make to ensure members of his party would support the aid package. Republicans aren’t known for being the smartest of people, but it’s wrong to solely place the blame on their antics this time. Moderate Democrats played a hand in pushing the bill over the finish line, effectively stripping half the country of their First Amendment rights.

The government has yet to provide concrete evidence of a national security threat, which is strange, because the only sound legal argument for this law is national security. ByteDance is owned and effectively controlled by the Chinese Communist Party, and there is potential for the Chinese to compromise the security of the United States with access to hundreds of millions of Americans’ phones. Yet, there is zero evidence of this happening in actuality — I’m not saying that it isn’t happening, but there is no evidence for the public to see. When this law is challenged in court — and it absolutely will be — this will be the primary aspect of the case, as silencing speech because “terrorist content is promoted” is easily one of the most illegal things Congress can ever do. From the First Amendment of the Constitution:

Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the Government for a redress of grievances.

That is the entire amendment, and there is no need for more because it’s extremely descriptive. Congress cannot make a law that prohibits the “free exercise of speech” — period, no matter what that speech is — unless it’s in the interests of the government’s job of protecting the American people. If there is no attempt to prove that this law is in support of that job, this bill should get tossed straight into a fire. There is a reason that yelling “I’m going to bomb this airplane!” in an airport is not considered “free exercise of speech,” so the same logic applies here: The government must provide tangible proof that TikTok’s continued Chinese ownership is an equivalent danger to someone yelling “I’m going to bomb this airplane!”

When the government makes that argument, it cannot use the speech on TikTok as proof, because no matter the speech, speech alone cannot impede national security. For example, 4chan is filled to the brim with unfavorable, illegal speech, but the government can’t directly punish the platform owner for “threatening national security” due to Section 230 of the Communications Acts of 1934 and 1996 which give platform owners immunity from what people say on them. Forcing the divestiture of a company because of speech on a platform — illegal or not; supporting terrorists or communists is not illegal — is a blatant violation of the First Amendment and has been proven case after case before the Supreme Court.

That leaves just one more question: Why won’t ByteDance divest TikTok? It’s a good question that I’ve also been pondering because I’m not one to favor Chinese control over one of the largest social media platforms in the United States. I want a TikTok divestiture, but I don’t want it to be forced by the government. I’ve come to this conclusion: ByteDance won’t ever divest TikTok not because it doesn’t want to, but because it legally cannot due to its Chinese ownership and control. The CCP is truly an authoritarian government, and TikTok is its best way of manipulating the public’s image of it, so it’s willing to let a Chinese company suffer financially if it means displaying a facade of strength in front of the United States. The CCP doesn’t care about the money ByteDance makes — it’s communist — but it does care about the data ByteDance generates. It wants power, and the best way for it to truly change the public’s perception of it is by threatening the public’s favorite social media platform.

Naturally, if TikTok vanishes in a year — a prospect that I think is still thoroughly unlikely — Americans will solely place the blame on their government, not on TikTok or China. And that point of contention between Americans and their government is exactly the reason why China doesn’t want to divest TikTok. The Chinese government wants power and strength; it wants to change the way Americans perceive it across the Pacific. This bill just gave China a brand new, effective strategy. Nice work, Washington — you’ve been outsmarted by Beijing again.

Viola Zhou on Tuesday reported an extremely thorough, quite lengthy piece about the cultural differences and headaches in Taiwan Semiconductor Manufacturing Company’s new semiconductor fabrication plant in Arizona, typically called a “fab.” This was my favorite part that truly exemplifies the differences between work culture in the East and the West:

At Fab 18, nearly all communication took place in Taiwanese and Mandarin Chinese, the two most widely spoken languages in Taiwan. The Americans found it difficult to understand meetings, production guidelines, and chatter among local engineers. In theory, every American was supposed to have a Taiwanese buddy — a future Arizona worker who would help them navigate the workplace. But the Americans said their buddies were often too busy to help with translations, or else not familiar enough with the technical processes because they were freshly transferred from other production lines.

Many trainees, including Bruce, relied on Google Translate to get through the day, with mixed results. Technical terms and images were hard to decipher. One American engineer said that because staff were not allowed to upload work materials to Google, he tried to translate documents by copying Chinese text into a handwriting recognition program. It didn’t work very well…

TSMC’s work culture is notoriously rigorous, even by Taiwanese standards. Former executives have hailed the Confucian culture, which promotes diligence and respect for authority, as well as Taiwan’s strict work ethic as key to the company’s success. Chang, speaking last year about Taiwan’s competitiveness compared to the U.S., said that “if [a machine] breaks down at one in the morning, in the U.S. it will be fixed in the next morning. But in Taiwan, it will be fixed at 2 a.m.” And, he added, the wife of a Taiwanese engineer would “go back to sleep without saying another word.”

During their visit, the Americans got a taste of the company’s intense work culture. To avoid intellectual property leaks, staff were banned from using personal devices inside the factory. Instead, they were given company phones, dubbed “T phones,” that couldn’t be connected to most messaging apps or social media. In one department, managers sometimes applied what they called “stress tests” by announcing assignments due the same day or week, to make sure the Americans were able to meet tight deadlines and sacrifice personal time like Taiwanese workers, two engineers told Rest of World. Managers shamed American workers in front of their peers, sometimes by suggesting they quit engineering, one employee said.

This story reported a challenge with chip manufacturing in the United States that I hadn’t considered until now: cultural differences. Semiconductor manufacturing — like any manufacturing — is a very male-dominated industry, and also one that happens to be low-paying in the East with frankly atrocious working conditions, so bringing that industry to the West, where workers expect shorter work hours and more humane treatment, is difficult. Later in the story, Zhou reports how female co-workers were mistreated, how employees weren’t permitted to bring their phones to work, and how there was a disconnect between the Taiwanese managers and American workers. I suggest you read the piece in its entirety.

The obvious solution to this problem is for the managers themselves to be American, but that’s unfeasible as of now because someone needs to train those managers for the job. Unlike other multinational corporations that have operated in the United States for decades, TSMC needs to first train the Americans to a level of seniority before they can take over the plants entirely. This creates a major bottleneck in the form of a chicken-and-egg problem: TSMC needs American workers to function as supervisors, but Americans want to leave their jobs at TSMC because of the Taiwanese managers. It’s a tough problem to solve, but one that I think can be ironed out with some changes to C-suite leadership.

Unlike the rank-and-file managers who presume authority over the day-to-day operations of the Arizona plant, C-suite executives in Taiwan should be able to rectify this issue by training the Taiwanese managers better. The onus shouldn’t be on the Americans to change — the Taiwanese need to better adapt to the Americans’ way of work. The United States will never be Taiwan or China, and I think TSMC management understands that. The U.S. government is providing the funding, the clients are providing the orders — now, it’s time for TSMC to change how it manages its employees to better its recruiting strategy.

These managers need a change of attitude because that’s inherently the main job of management — to adapt to workers’ preferences. If TSMC doesn’t change the way it controls its workers, the projects will fall apart fast. Funding comes once projects get off the ground, especially if President Biden wins re-election this fall, but for the projects to succeed, workers need to be satisfied. TSMC’s Glassdoor ratings are not desirable for the No. 1 semiconductor manufacturer in the world. Americans care a lot about the culture of the company for which they work, and TSMC needs to understand that and better adapt to American work culture.

Alex Heath, reporting a politically heated story for The Verge on Wednesday:

Google fired 28 employees in connection with sit-in protests at two of its offices this week, according to an internal memo obtained by The Verge. The firings come after 9 employees were suspended and then arrested in New York and California on Tuesday.

The fired employees were involved in protesting Google’s involvement in Project Nimbus, a $1.2 billion Israeli government cloud contract that also includes Amazon. Some of them occupied the office of Google Cloud CEO Thomas Kurian until they were forcibly removed by law enforcement. Last month, Google fired another employee for protesting the contract during a company presentation in Israel.

In a memo sent to all employees on Wednesday, Chris Rackow, Google’s head of global security, said that “behavior like this has no place in our workplace and we will not tolerate it…”

He also warned that the company would take more action if needed: “The overwhelming majority of our employees do the right thing. If you’re one of the few who are tempted to think we’re going to overlook conduct that violates our policies, think again. The company takes this extremely seriously, and we will continue to apply our longstanding policies to take action against disruptive behavior — up to and including termination.”

I say this as a liberal, somewhat progressive person politically: Google had every single right to fire each last one of these “protestors.” Gaby Del Valle for The Verge also reported earlier on Wednesday that the demonstrators occupied Google’s offices illegally, even after they were asked to leave by management, which led the company to call the police to arrest nine of them. A man who works for Google said to the protestors, as quoted by The Verge: “We’re asking you to leave again for the last time.” Then, when they stayed, a police officer offered the demonstrators a plea deal of sorts: “Listen, we’ll let you walk out the door right now — it’s a non-issue if you’re willing to go. If not, you’re going to be arrested for trespassing.”

Every one of the 28 employees who was fired Wednesday evening was given multiple chances to leave a secured, locked building that they were not permitted to use for demonstrations, but they flagrantly violated orders given by a representative for the owner of the building, who also happens to be their employer. If that isn’t a reason for termination of employment, I do not know what is. This type of lunatic behavior would get any employee fired because the protestors engaged in illegal activity. It’s trespassing — illegally occupying a building when the owner gives repeated instructions to leave. When that owner is your employer, termination is a fair punishment.

Those complaining about “free speech,” like many right-wingers, do not actually understand what free speech means in this context. Employees are permitted to protest, especially against their employer, for a variety of reasons. That is protected speech under the Constitution as long as the protestors don’t cause a disturbance, also known as “disorderly conduct” in criminal law. Trespassing goes beyond disorderly conduct; it’s a felony offense to occupy a building when disallowed. From what is known currently, it doesn’t seem that any protestors were charged with crimes — they were simply fired for staging a rogue protest against their employer. Trespassing and causing a disturbance is enough of a reason to fire an employee according to the law. If an employee randomly stood up from their desk and began shouting, they would be reprimanded.

Online activists are calling this protest “peaceful,” when breaking the law is exactly what makes a protest the opposite of “peaceful.” No matter what the protest was for — Black Lives Matter, LGBTQ rights, or Israel’s military campaign in Gaza — a protest in a building where demonstrators are unauthorized to be is illegal, and therefore, punishable. Punishing an employee for failing to obey orders is a right given by the Constitution’s First Amendment to Google, and a corporation disciplining an employee or contractor for psychotic behavior is protected legally under free speech rights in the United States. Nobody has a right to be upset about how the protestors were treated in this case — while their cause’s importance can be debated until the end of time, their actions are undoubtedly flawed.

I truly cannot believe people who managed to land a job at one of the world’s largest technology firms are so stupid that they occupied the private office of their chief executive as if they were rioters on January 6, when a mob of Republican supporters stormed the Capitol in Washington to stop the certification of the 2020 election. The blatant lawlessness exhibited in this protest is appalling and should be condemned in the strongest terms. A functioning democracy necessitates the right to protest, but this wasn’t any ordinary protest — it was a stunning spectacle of incompetence, mindlessness, and arrogance unlike one displayed in Silicon Valley before. “Big Tech” employees have protested via many walk-outs, sit-ins, and other protests to reject their employers' policies, but they have always done so peacefully and respectfully, inspiring change for everyone in a dignified manner. This was the complete opposite.

Condemning the protestors isn’t an endorsement of Israel’s actions in Gaza, Google’s deal with the Israeli government, or the U.S. government’s foreign policy with respect to Jerusalem. Anyone who appreciates dignity and the right to protest in the workplace should be ashamed of Wednesday’s events because they demonstrate the rogue, nonsensical mentality of pro-Palestinian mobsters who are taking rights away from peaceful protestors with their illegal actions. In addition to breaking trespassing laws, they chanted “From the river to the sea, Palestine will be free!” a slogan deemed antisemitic by many, presumably including Jewish people at Google, and pinned banners to the wall with antisemitic language. Google has the right to maintain company policy and remove employees who disrupt the workplace with hateful messages, regardless of what political ideologies those messages are linked to. Google is a company made of people, and if they feel disrespected, they have the right to take action.

For the sacred right to peacefully protest in the United States to remain intact, protestors need to remain respectful and mindful of their neighbors. If they aren’t, the country risks another January 6 — but this time, much, much worse. Political violence and lawlessness are never acceptable.

Gaby Del Valle, reporting for The Verge:

Sens. Wyden and Lummis introduce an amendment limiting FISA’s warrantless wiretapping powers. The amendment would reverse a provision included in the recent House bill reauthorizing Section 702 of FISA that expands the definition of “electronic communications service provider,” which critics say would force Americans to essentially spy for the government.

“Forcing ordinary Americans and small businesses to conduct secret, warrantless spying is what authoritarian countries do, not democracies,” Wyden said in a statement.

The House of Representatives recently re-authorized Section 702 of the Foreign Intelligence Surveillance Act, which, ironically, allows the Federal Bureau of Investigation to perform clandestine searches of American citizens on American soil without a warrant or their consent. The government justifies this by saying FISA is a critical national security tool to prevent foreign attacks on the United States, but the act is mostly used to surveil Americans, not foreigners. If FISA is not re-authorized by Saturday, the government can no longer spy on its citizens — foreign and abroad — however it pleases without even a warrant to justify its actions.

Senator Ron Wyden, Democrat of Oregon, and many progressive Democrats in the House who unsuccessfully voted against re-authorizing FISA have expressed concerns about this warrantless searching, but nevertheless, it seems increasingly likely that the original version of Section 702 will be re-authorized by the weekend. But the new version of FISA, which extends the program for two years, also includes an amendment courtesy of House Republicans to expand the definition of “electronic communications providers,” i.e., companies that are required to provide data when the government requests it. This definition, until now, has only included smartphone makers and other telecommunications companies, like Apple and Google, but the latest House amendment that passed also requires cloud computing firms to provide data.

This effectively means that the government, from now on, will be permitted to ask Amazon Web Services for all the user data associated with one account, which could include practically an entire person’s business life because most people and small businesses use one cloud provider to host their website or other business tools. Google Cloud, AWS, and even Apple’s iCloud servers are all susceptible to this unprecedented, warrantless searching, which, as Wyden and Senator Cynthia Lummis, Republican of Wyoming, say, is precisely what authoritarian regimes do. Russia and China employ this exact kind of surveillance to silence their people, and it is absolutely astounding to me that nobody has taken this law to court over a violation of the Fourth Amendment.

If the new amendment is passed — which seems increasingly likely based on the bipartisan support the latest bill has enjoyed in both the House and Senate — it would open Americans up to a brand new front of dangerous government surveillance akin to communist regimes like China. I don’t think Congress should axe FISA entirely, unlike many progressives and far-right “Make America Great Again” Republicans, but I think there should be an amendment to prevent warrantless searches. And Wyden and Lummis’ amendment in the Senate, in an ideal world, should pass, because allowing the government to surveil data stored in the cloud is blatant government overreach.

Foo Yun Chee, reporting for Reuters:

Meta Platforms and other large online platforms should give users an option to use their services for free without targeted advertising, EU privacy watchdog the European Data Protection Board said on Wednesday.

The EDPB’s opinion came after it was asked by national privacy regulators in the Netherlands, Norway, and Germany to look into consent or pay models adopted by large online platforms such as Meta.

“If controllers do opt to charge a fee for access to the ‘equivalent alternative’, they should give significant consideration to offering an additional alternative. This free alternative should be without behavioural advertising,” EDPB said in a statement.

The board’s ruling on Wednesday gives the European Commission, the European Union’s executive body, the ability to force Meta and “other large online platforms,” like Google, to provide their services for free without any targeted advertising. In other words, this outlandish decision allows the commission to dictate how a corporation makes money, even if the “recommended” method is non-viable. The non-technical equivalent of this ruling is the government telling a baker they can’t price their bread at $5 because it thinks that is too expensive. The government doesn’t actually know that the ingredient cost per loaf is $4.75, but it also doesn’t care to find out, so it just punishes the baker for selling the bread at a 25-cent profit even though the baker needs the 25 cents to continue their operations. It is entirely unfair.

The European Commission has been waiting on this ruling since March so it can begin to force Meta to offer non-targeted advertising for free, instead of forcing consumers to subscribe to the ad-free versions of Instagram and Facebook as Meta currently does in response to the Digital Markets Act, which went into effect in early March. E.U. users can choose to pay Meta 10 euros (around $11) monthly to remove all ads — including the targeted ones — because the bloc’s DMA forces “Big Tech” to offer users the ability to disable targeted advertising somehow. The subscription is Meta’s way of sneaking around the true intention of the legislation, which is for companies to offer a simple toggle switch for users to disable targeting for free. The commission didn’t like Meta’s clever idea, so it complained to the EDPB, which, naturally, ruled in the commission’s favor.

It’s important to keep in mind that the DMA does not specifically state that this scheme — which is known as a “pay or OK” tactic inside Brussels — is illegal, or that it shouldn’t be employed. The law simply requires there to be some way for users to opt out of targeted advertising, even if that method is asking for payment since the legislation says nothing about payment entirely. But commissioners, prominently Margrethe Vestager, the commission’s antitrust chief, have decided Meta’s perfectly legal compliance with the DMA is too unacceptable, so they have begun re-interpreting the law to enforce it at their whim. It’s the same old tactic the European Union has been playing for months. As I wrote in March, the commission is “playing a one-sided, rigged game while laughing manically in the corner at everyone falling face-flat on the ground.”

Back to Wednesday: Now, with pesky legality out of the way, the commission is free to push Meta against the wall and choke its neck until it offers its services practically for free, since Meta’s primary revenue source is effective advertisement targeting. Without targeted advertisements, Meta’s average revenue per user drops significantly since advertisers want to ensure their products get placed in front of prospective buyers’ eyes — but thanks to the European Union, if Meta wants to continue operating in the bloc, it needs to take a loss on offering services to Europeans. And if Meta does the algebra and determines it’s not worth it to stay in the European Union, the commission will surely take Meta to court for leaving for it being somehow anti-consumer, even though it’s the commission’s fault in the first place for forcing the company to leave.

The only way for Meta to remain profitable in the European Union is for the company to inundate users with terrible advertisements in the hopes that it can make up for the lack of targeted advertising by simply selling more advertisements at a lower price. That move, however, would also probably draw the ire of regulators in Brussels, who clearly have nothing better to do than go after technology companies which not one ordinary European is complaining about. Europeans can already opt out of targeted advertising for free by using App Tracking Transparency, Apple’s technology for disguising a tracking identifier companies like Meta use to keep track of users across the web. But privatization does not seem to be in the commission’s interests, so it has instead opted to bully any company doing business in its sacred bloc for no reason other than politics. Is there seriously no way for Europeans to object to this madness?

Julian Barnes, reporting for The New York Times:

The threat against U.S. elections by Russia and other foreign powers is far greater today than it was in 2020, the chairman of the Senate Intelligence Committee said on Tuesday.

Senator Mark Warner, the Virginia Democrat who leads the committee, said the danger had grown for multiple reasons: Adversarial countries have become more adept at spreading disinformation, Americans are more vulnerable to propaganda, communication between the government and social media companies has become more difficult and artificial intelligence is giving foreign powers new abilities…

“With polarization in this country, and the lack of faith in institutions, people will believe anything or not believe things that come from what used to be viewed as trusted sources of information,” Mr. Warner said. “So there’s a much greater willingness to accept conspiracy theories.”

Vulnerability to influence operations, Mr. Warner said, is not confined to the United States. In Slovakia, for example, Russian information operations influenced views of Russia’s war in Ukraine.

“Slovakia was 80 percent pro-Ukraine,” he said. “Two years later, with massive amounts of Russian misinformation and disinformation, you have a pro-Russian government and 55 percent of Slovaks think America started the war in Ukraine.”

These statements from Warner aren’t shocking even in the slightest. If you look at social media now, it’s filled with disinformation from self-proclaimed Americans, both on the left and right of the political spectrum. Left-wing nuts continue to push pro-China, pro-Iran propaganda in the name of “progressivism,” while right-wingers brazenly post pro-Kremlin, anti-Ukraine rhetoric. As I’ve written before, I don’t think these are actually Americans, but rather, I surmise the ideas from Russian and Chinese bots have entirely taken over the minds of Americans, including those in Congress. Young Americans are more likely than ever to distrust institutions and their government — and rightfully so — but they are also more likely to subscribe to foreign propaganda that advances flawed ideologies.

Russia and China continue to flood social media websites with fictitious misinformation to influence the 2024 election. The two countries employ bots, and sometimes even human labor, to publish websites full of wrong information, social media posts expressing dissatisfaction with the current administration, and launch advertising campaigns to create distrust in the U.S. government. These tactics prove themselves every day to be overwhelmingly successful; a message disguised as an “America First” endorsement is much more likely to be listened to than one directly opposing American military efforts overseas. A Russian operative asking “Why are we sending our money to Ukraine when we should be securing our southern border?” has a more striking effect on right-leaning Americans than “We should stop sending money to Ukraine because it’s none of our business.”

These operations are not covert or minuscule in scale — they’re entirely widespread on the internet today, on social media websites like X and Threads. People with “radical” political views might not actually be expressing any beliefs at all — instead, they are probably just a Russian asset. My advice is not to interact with these foreign influence accounts whatsoever, and I further demand that social networks like Meta take more action to combat misinformation and perform a mass deletion of spam accounts with outrageous beliefs. This is happening in the United States, in Europe, and in many Asian countries, and internet citizens must be more vigilant in reporting it and combating the spread of dangerous propaganda that has the power to threaten our respective democracies.

Andrew Hawkins, reporting for The Verge:

Rivian is pushing a new software update that will give its customers better insight into which EV chargers to visit — and which to avoid…

“Our North Star is charging and trip planning in EVs should just work,” Wassym Bensaid, Rivian’s head of software, told The Verge. “You should not think about it.”

I had the chance to test out Rivian’s new software update during a recent road trip in an R1S SUV. Inputing a destination brought up dozens of chargers on the vehicle’s navigation, each of which displayed a letter grade. An “A” grade is a sign that the charger was in good working condition, while an “F,” well, speaks for itself.

“Surprisingly, actually, there’s multiple chargers rated F,” Bensaid said. “That was one of the ‘a ha’ moments as we went through the data.”

The new ranking system is determined by a host of data collected by Rivian’s customers, Bensaid said. Each vehicle is connected and constantly sending data back to the company’s headquarters, which then gets processed to remove “noise” that’s not essential to the decision-making algorithm.

This is extremely clever. While Tesla Superchargers are notoriously reliable and display stall information in Tesla vehicles’ infotainment systems, that functionality isn’t available for other electric vehicle manufacturers to integrate themselves, even though Rivian and other automakers have recently opened support for Tesla Superchargers, working with Tesla to develop adapters for their vehicles. (Rivian’s integration launched in March.) This new Rivian software feature, however, not only adds reliability information to the map in the vehicle itself from Tesla Superchargers, but also collects information from other brands, like Electrify America, the United States’ largest DC fast charging network outside of Tesla’s Superchargers.

Electrify America stalls are often plagued with reliability issues, and Electrify America itself doesn’t have the ability to monitor how its chargers are operating. The only way for customers to check if an Electrify America unit is functioning is to drive to one and hope for the best. Rivian’s new software system will automatically collect analytics from the chargers whenever a Rivian driver charges at one; after enough users have charged at a destination, results will appear for all other drivers through Rivian’s in-car map. While this information isn’t real-time, per se, unlike Tesla’s feature which automatically notifies drivers if a stall is broken, it is better than going to a charger without knowing anything about its reliability.

Hawkins points out that this software is limited due to how few Rivian cars traverse the roads of the United States, which is somewhat detrimental to the usefulness of the feature. Underused chargers located in rural areas are significantly less likely to ever be touched by a Rivian driver, let alone however many it requires for the software to begin calculating reliability information for it, so only busy chargers in metropolises will benefit from the software update. I still think that is better than nothing, but it also adds pressure on Electrify America and ChargePoint, another EV charger brand, to add public uptime data via an application programming interface or integrations with carmakers for them to integrate the statistics into their vehicles.

Google, writing in an email to me Monday morning:

Hi Eshu,

Thank you for being a Google One member where you enjoy extra storage, family sharing, and more. We’re writing to let you know about some updates coming to your Google One subscription starting on May 15. These changes are designed to streamline your benefits while ensuring you have a valuable subscription experience…

Phasing out two benefits: With a focus on providing the most in-demand features and benefits, we’re discontinuing free shipping for select print orders from Google Photos (in Canada, the UK, US, and EU) starting on May 15 and VPN by Google One later this year.

My subscription price is still the same, so I guess it’s time to add another product to the Google Graveyard. Speaking of Google One and the Google Graveyard, here is Abner Li, reporting for 9to5Google:

Google is now “discontinuing the VPN feature as [they] found people simply weren’t using it.” The company tells 9to5Google that the deprecation will let the team “refocus” and “support more in-demand features with Google One.”

Earlier this year, Google One hit 100 million subscribers and CEO Sundar Pichai teased it as a future growth area driven by AI. Today’s change follows this week’s news about AI editing tools in Google Photos going free in the coming months and no longer requiring a subscription save for unlimited Magic Editor usage.

Google One included a virtual private network service, which the company launched in 2020 for some odd reason, but now that’s gone too. And again, users’ subscription prices are staying the same. Anyone who subscribed to Google One, especially to a yearly plan, is now just getting prematurely cut off because Google doesn’t care about supporting products.

Again, add it to the Google Graveyard.

Richard Stengel, writing for The Atlantic:

How many times has it happened? You’re on your computer, searching for a particular article, a hard-to-find fact, or a story you vaguely remember, and just when you seem to have discovered the exact right thing, a paywall descends. “$1 for Six Months.” “Save 40% on Year 1.” “Here’s Your Premium Digital Offer.” “Already a subscriber?” Hmm, no.

Now you’re faced with that old dilemma: to pay or not to pay. (Yes, you may face this very dilemma reading this story in The Atlantic.) And it’s not even that simple. It’s a monthly or yearly subscription—“Cancel at any time.” Is this article or story or fact important enough for you to pay?

Or do you tell yourself—as the overwhelming number of people do—that you’ll just keep searching and see if you can find it somewhere else for free?

According to the Reuters Institute for the Study of Journalism, more than 75 percent of America’s leading newspapers, magazines, and journals are behind online paywalls. And how do American news consumers react to that? Almost 80 percent of Americans steer around those paywalls and seek out a free option.

Paywalls create a two-tiered system: credible, fact-based information for people who are willing to pay for it, and murkier, less-reliable information for everyone else. Simply put, paywalls get in the way of informing the public, which is the mission of journalism. And they get in the way of the public being informed, which is the foundation of democracy. It is a terrible time for the press to be failing at reaching people, during an election in which democracy is on the line. There’s a simple, temporary solution: Publications should suspend their paywalls for all 2024 election coverage and all information that is beneficial to voters. Democracy does not die in darkness—it dies behind paywalls.

I’d go one step further than Stengel did: I think paywalls entirely are mostly unnecessary for regular news websites. There should be an exception for small firms or publications that mainly focus on analytical coverage or opinion pieces, such as The New Yorker or The Atlantic, even though there is irony in an article about paywalls being hidden behind a paywall, as Stengel rightly points out. The reason I say this is because analysis isn’t information per se in the same way regular, hard news and reporting is. The people who are interested in analysis or a columnist’s views on a topic are much more likely to pay for that information — it’s just news that needs to be balanced and free.

Paywalls are a relic from a time when newspapers were bought from newsstands daily and read in coffee shops across the country. Every day, people would buy a newspaper for a quarter, read it, then leave it for the next person to read — and so on, every day. Newspaper publishers saw an opportunity to sell subscriptions to these papers to frequent readers, who perhaps could save some money by buying the newspaper at a discounted monthly or yearly rate as opposed to buying a new paper every day. When the online era took over and as people bought fewer physical newspapers, publishing houses moved to sell those subscriptions online.

As inflation hiked, however, newspaper subscriptions became more expensive. Now, The New York Times costs $25 a month because the physical copy costs $4 per paper ($4 a day for 30 days is $120 a month, so the digital version is still cheaper, obviously). In parallel, news became more affordable due to the advent of online advertising and free journalism, causing many to drop their subscriptions and just consume free news published by digital-only websites, like The Verge or NBC News. The result of this schism is that there is a divide between free and paid journalism, as well as the quality of information that emanates from each source.

This is the core of Stengel’s piece: The online media landscape gives credence to clickbait because it spreads so quickly due to it being free. Free articles metastasize through the internet so rapidly because everyone can read them. But this abundance of free, semi-factual journalism is spreading and warping the public’s perception of the news media because good journalism just isn’t being sold and marketed properly. These online media companies often make more money than the print publishers that sell subscriptions because of how quickly the information spreads on social media. Why aren’t print publishers picking up on this?

I firmly believe more websites should drop the paywall entirely and instead opt to sell effective, non-intrusive advertising, which can often be lucrative. Newspapers themselves are already fantastic examples of how lucrative advertising can be, as they sell full-page ads in innovative formats, which advertising companies are itching to cash in on. This is part of the game of journalism in 2024 — if we want the media to be a reliable arbiter of information without relegating the ever-important job of reporting to citizen journalists on social media websites like X or Threads, the corporate overlords who control the media should get better at making it profitable.

The news industry is at a crossroads, with the hastened development of generative artificial intelligence in newsrooms and the increased popularity of foreign-owned video websites like TikTok, where more Americans are getting their information from than ever before. At a critical time like this, journalism should become more accessible, innovative, and forward-facing — and removing paywalls is a key step in that direction.

I have written about the Humane Ai Pin twice before, once when the product was first announced in November, and the other in March, when the company released a video walkthrough of the device’s features. Now, reviews are in, both on YouTube and the web, and they’re scathing. The Ai Pin is — and I cannot stress this enough — utter garbage, and that isn’t me jesting. To support this claim, here are some quotes from reviewers who have had time with the pin:

“It’s a nightmare.” — Arun Maini, Mrwhosetheboss

“It’s just frustrating.” — Michael Fisher, MrMobile

“It’s futuristic if the future sucked.” It “solves nothing and makes me feel stupid.” — Cherlynn Low, Engadget

“It just doesn’t work.” — David Pierce, The Verge

That is just a small snippet of criticism this device has received in just one day of reviews. I agree with these reviewers — this product lacks conviction, lacks a path to success, and doesn’t even do what it is advertised to do. As one commenter on YouTube put it, it feels like a late April Fools’ Day joke. It costs as much as a mid-range smartphone at $700, requires a mandatory $24 monthly subscription to function as anything more than a paperweight¹, and relies on the whims of artificial intelligence to do literally anything. And the times when it does do something, it does it wrongly, misidentifying landmarks, such as in Pierce’s review, or making up information, as shown in Fisher’s video.

AI software can be refined and tweaked over time². What can’t, however, is the very design premise of this gadget. Its primary method of interaction is a loudspeaker that is bound to annoy everyone around, and connecting to a Bluetooth headset requires interacting with a flawed, frustrating laser projector. Reviewers have described the projector as annoying, hard to use, and simply impossible to view while in daylight. Interacting with content requires learning unintuitive gestures and flailing movements of the arms. Due to this oversight, which I picked up on in November, interacting with this menace of a lapel pin causes arm strain and annoyance.

Also, the battery dies quite frequently, making Humane’s “Perpetual Power System,” i.e., extra batteries, essential to use the product for anything longer than a few hours, according to early tests. And if you do use it for extended periods, it’ll overheat, as Fisher demonstrated during a call with his mother and as Pierce encountered while using the laser projector. These occurrences show that the product is impossible to use most of the time, and is useless when it is possible to use. The onboard camera, used for first-person shots, is low-quality and lackluster, though it is neat in a pinch, but no reasonable nor sane person would say that it is worth $700 and $24 a month.

Humane’s founders, Bethany Bongiorno and Imran Chaudhuri, have marketed the Ai Pin as a smartphone companion at the very least, and at times have even come out as more confident after the flopped November launch, calling the product a smartphone replacement at a time of increased distraction. To back up this moot point — people love their phones — Bongiorno has been reposting accounts citing a book with debatable sourcing that claims smartphones are the sole cause of childhood and adolescent depression, when in fact, the rises in these numbers are correlated with the uptake in smartphone use, not caused by it. When confronted with her earlier claims that she marketed the Ai Pin as a replacement for the phone, she flat-out denied it.

Nobody should take these claims seriously because this entire project feels like a scheme for investor money. When the original plan for a more ambitious product failed, Humane pivoted to AI large language models and built this device in four months after the release of ChatGPT in October 2022. Trust the reviewers because they’re experts — and the experts say that the Humane Ai Pin is a worthless piece of garbage. So do I.

Thomas Grove, Nicholas Bariyo, Micah Maidenberg, Emma Scott, and Ian Lovett, reporting for The Wall Street Journal:

A salesman at Moscow-based online retailer shopozz.ru has supplemented his usual business of peddling vacuum cleaners and dashboard phone mounts by selling dozens of Starlink internet terminals that wound up with Russians on the front lines in Ukraine.

Although Russia has banned the use of Starlink, the satellite-internet service developed by Elon Musk’s SpaceX, middlemen have proliferated in recent months to buy the user terminals and ship them to Russian forces. That has eroded a battlefield advantage once enjoyed by Ukrainian forces, which also rely on the cutting-edge devices.

The Moscow salesman, who in an interview identified himself only as Oleg, said that most of his orders came from “the new territories”—a reference to Russian-occupied parts of Ukraine—or were “for use by the military.” He said volunteers delivered the equipment to Russian soldiers in Ukraine.

On battlefields from Ukraine to Sudan, Starlink provides immediate and largely secure access to the internet. Besides solving the age-old problem of effective communications between troops and their commanders, Starlink provides a way to control drones and other advanced technologies that have become a critical part of modern warfare.

The proliferation of the easy-to-activate hardware has thrust SpaceX into the messy geopolitics of war. The company has the ability to limit Starlink access by “geofencing,” making the service unavailable in specific countries and locations, as well as through the power to deactivate individual devices.

Russia and China don’t allow the use of Starlink technology because it could undermine state control of information, and due to general suspicions of U.S. technology. Musk has said on X that to the best of his knowledge, no terminals had been sold directly or indirectly to Russia, and that the terminals wouldn’t work inside Russia.

The Wall Street Journal tracked Starlink sales on numerous Russian online retail platforms, including some that link to U.S. sellers on eBay. It also interviewed Russian and Sudanese middlemen and resellers, and followed Russian volunteer groups that deliver SpaceX hardware to the front line.

Anyone who seriously believes Musk doesn’t have the ability to properly restrict these terminals to specific airspace via geofencing is genuinely stupid, and probably a Russian asset. Here is how Russians get access to Starlink, even though President Vladimir Putin’s propaganda would lead you to believe Starlink terminals aren’t authorized for use in Russia: First, smugglers buy Starlink hardware in Middle Eastern countries, like the United Arab Emirates, for example, then activate those terminals for use anywhere in the world — a subscription SpaceX, the company that makes Starlink, sells. Then, those smugglers market the terminals on sites like eBay so that Russians can have the hardware shipped to nearby, Kremlin-friendly countries, sold at a markup. Then, smugglers bring the terminals and receivers over the border as if they were drugs or any other illegal items. “Patriotic” Russians then wheel them over to the front lines, where idiotic Russian soldiers are so stupid that they don’t even camouflage the bright white plastic terminals.

Ukraine, which is currently fighting a brutal war with Russia, also has access to Starlink, provided due to SpaceX’s contractual obligations with the U.S. Defense Department, which requires SpaceX to deploy Starlink hardware to U.S. allies in need, like Taiwan and Ukraine. Granted, Musk neutered Ukraine’s access to these important terminals, which provide internet access in Russian-occupied areas, when needed the most, but the service still remains available there in some capacity due to U.S. contracts. Bloomberg reported Wednesday that the Defense Department pays $23 million for this deal, but the U.S. official who leaked the information to Bloomberg declined to say whether the United States would renew it with Musk or not. This type of smuggling would be concerning to the United States due to these deals, so, in March, House Democrats sent a letter to SpaceX over the illegal import of Starlink terminals by Russia, which aid Putin in controlling drones and commanding troops. Apparently, nothing came of that letter, and just a month later, The Journal reported on the continued illegal use of Starlink terminals in Russia.

Again, only a fool would believe Musk and his company have no clue about the illegal use of these products in Russia — smart people work for SpaceX and upwards of 5,000 satellites roam Earth tracking the position of the terminals in real-time, according to The Journal’s report. Musk has the tools at his disposal to halt Russia’s unlawful use of terminals to destroy civilian buildings in Ukraine and illegally occupy sovereign territory, but he never will, because he himself is a Russian asset parroting propaganda straight from Moscow on X, his social media website. However many pro-Kremlin Republican puppets there are in Congress, however, the United States should exercise its leverage and contracts with SpaceX to force Musk’s firm to comply with U.S. law and disable enemy use of Starlink satellites. The United States has given Musk a free pass on contracts for too long — if it wants to continue doing business with the world’s richest Russian propagandist, it needs to shove him down on his knees and make him beg for the money he wants.

Katie Notopoulos, reporting for Business Insider regarding Google possibly scraping Google Docs for use as training data for Gemini, Google’s artificial intelligence chatbot:

A representative for Google confirmed to Business Insider that simply changing the share settings to “anyone with the link” did not mean that a document was “public” and would be used for AI training.

To be “publicly available,” that document would need to be posted on a website or shared on social media. Basically, some kind of web crawler would need to be able to find it. That can’t happen with a file just emailed back and forth between two people — like if you send your friend a link over Gmail, for instance, Google said.

Breathe a sigh of relief.

I still think this is mildly disingenuous since there is no way to opt out of AI training, unlike on a traditional website, where disallowing AI crawlers to index the page is as simple as adding a line to the website’s robots.txt file. And, not to mention, there is no clarification in either Google Docs’ privacy policy or terms of use on how Google might use your documents to train its AI models. I hope Google makes improvements and clarifies its policy in the near future. But for now, unless you’re publicly publishing a Google Doc to the web, there is no need to worry.