Eurpopean Commission Dings Apple Over Anti-Steering Provisions in App Store, and Opens New Investigations Into Core Technology Fee, Sideloading Protections, and the Eligibility Requirements to Offer an Alternative Marketplace 

The European Commission:

Today, the European Commission has informed Apple of its preliminary view that its App Store rules are in breach of the Digital Markets Act (DMA), as they prevent app developers from freely steering consumers to alternative channels for offers and content.

I think what they’re saying here is that Apple’s current compliance offering, where developers can remain exclusively in the App Store in the EU under the existing terms, or choose the new terms that allow for linking out to the web, aren’t going to pass muster. The EC wants all apps to be able to freely — as in free of charge freely — link out to the web for purchases, regardless if they’re from the App Store, an alternative marketplace, or directly sideloaded.

The Commission will investigate whether these new contractual requirements for third-party app developers and app stores breach Article 6(4) of the DMA and notably the necessity and proportionality requirements provided therein. This includes:

1. Apple’s Core Technology Fee, under which developers of third-party app stores and third-party apps must pay a €0.50 fee per installed app. The Commission will investigate whether Apple has demonstrated that the fee structure that it has imposed, as part of the new business terms, and in particular the Core Technology Fee, effectively complies with the DMA.

No word on how it doesn’t comply, just that they don’t like it.

2. Apple’s multi-step user journey to download and install alternative app stores or apps on iPhones. The Commission will investigate whether the steps that a user has to undertake to successfully complete the download and installation of alternative app stores or apps, as well as the various information screens displayed by Apple to the user, comply with the DMA.

This sounds like they’re going to insist that Apple make installing sideloaded apps and alternative stores a no-hassle experience. What critics see is Apple putting up obstacles to installing marketplaces or sideloaded apps just to be a dick about it and discouraging their use to keep users in the App Store. What I see are reasonable warnings for potentially dangerous software. We’ll see how that goes.

Perhaps where the EC will wind up is making app store choice like web browser choice. Force Apple to present each user with a screen listing all available app marketplaces in their country in random order, of which Apple’s own App Store is but one, just like Safari in the default browser choice screen.

3. The eligibility requirements for developers related to the ability to offer alternative app stores or directly distribute apps from the web on iPhones. The Commission will investigate whether these requirements, such as the ‘membership of good standing’ in the Apple Developer Program, that app developers have to meet in order to be able to benefit from alternative distribution provided for in the DMA comply with the DMA.

I’m not sure what this is about, given that Apple relented on allowing even Epic Games to open a store. Maybe the financial requirements?

In parallel, the Commission will continue undertaking preliminary investigative steps outside of the scope of the present investigation, in particular with respect to the checks and reviews put in place by Apple to validate apps and alternative app stores to be sideloaded.

This pretty clearly is about Apple using notarization as a review for anything other than egregious bugs or security vulnerabilities. I complain as much as anyone about the aspects of the DMA that are vague (or downright inscrutable), but this aspect seems clear-cut. It’s a bit baffling why Apple seemingly sees notarization as an opportunity for content/purpose review, like with last week’s brouhaha over the UTM SE PC emulator. Refusing to notarize an emulator that uses a JIT is something Apple ought to be able to defend under the DMA’s exceptions pertaining to device security; refusing to notarize an emulator that doesn’t use a JIT seems clearly forbidden by the DMA.

Apple Disables WebKit’s JIT in Lockdown Mode, Offering a Hint Why BrowserEngineKit Is Complex and Restricted

Last week I mentioned Apple’s prohibition on JITs — just-in-time compilers — in the context of their rejection of UTM SE, an open source PC emulator. Apple’s prohibition on JITs, on security grounds, is a side issue regarding UTM SE, because UTM SE is the version of UTM that doesn’t use a JIT. But because it doesn’t use a JIT, it’s so slow that the UTM team doesn’t consider it worth fighting with Apple regarding its rejection.

On that no-JITs prohibition, though, it’s worth noting that Apple even disables its own trusted JIT in WebKit when you enable Lockdown Mode, which Apple now describes as “an optional, extreme protection that’s designed for the very few individuals who, because of who they are or what they do, might be personally targeted by some of the most sophisticated digital threats. Most people are never targeted by attacks of this nature.” Apple previously described Lockdown Mode as protection for those targeted by “private companies developing state-sponsored mercenary spyware”, but has recently dropped the “state-sponsored” language.

Here’s how Apple describes Lockdown Mode’s effect on web browsing:

Web browsing - Certain complex web technologies are blocked, which might cause some websites to load more slowly or not operate correctly. In addition, web fonts might not be displayed, and images might be replaced with a missing image icon.

JavaScriptCore’s JIT interpreter is one of those “complex web technologies”. Alexis Lours did some benchmarking two years ago, when iOS 16 was in beta, to gauge the effect of disabling the JIT on JavaScript performance (and he also determined a long list of other WebKit features that get disabled in Lockdown Mode, a list I wish Apple would publish and keep up to date). Lours ran several benchmarks, but I suspect Speedometer is most relevant to real-world usage. Lours’s benchmarking indicated roughly a two-third reduction in JavaScript performance with Lockdown Mode enabled in Speedometer.

This brings me to BrowserEngineKit, a new framework Apple created specifically for compliance with the EU’s DMA, which requires gatekeeping platforms to allow for third-party browser engines. Apple has permitted third-party browsers on iOS for over a decade, but requires all browsers to use the system’s WebKit rendering engine. One take on Apple’s longstanding prohibition against third-party rendering engines is that they’re protecting their own interests with Safari. More or less that they’re just being dicks about it. But there really is a security angle to it. JavaScript engines run much faster with JIT compilation, but JITs inherently pose security challenges. There’s a whole section in the BrowserEngineKit docs specifically about JIT compilation.

As I see it Apple had three choices, broadly speaking, for complying with the third-party browser engine mandate in the DMA:

  1. Disallow third-party browser engines from using JITs. This would clearly be deemed malicious by anyone who actually wants to see Chromium or Gecko-based browsers on iOS. JavaScript execution would be somewhere between 65 to 90 percent slower compared to WebKit.

  2. Allow third-party browser engines in the EU to just use JIT compilation freely without restrictions. This would open iOS devices running such browsers to security vulnerabilities. The message to users would be, effectively, “If you use one of these browsers you’re on your own.”

  3. Create something like BrowserEngineKit, which adds complexity in the name of allowing for JIT compilation (and other potentially insecure technologies) in a safer way, and limit the use of BrowserEngineKit only to trusted web browser developers.

Apple went with choice 3, and I doubt they gave serious consideration to anything else. Disallowing third-party rendering engines from using JITs wasn’t going to fly, and allowing them to run willy-nilly would be insecure. The use of BrowserEngineKit also requires a special entitlement:

Apple will provide authorized developers access to technologies within the system that enable critical functionality and help developers offer high-performance modern browser engines. These technologies include just-in-time compilation, multiprocess support, and more.

However, as browser engines are constantly exposed to untrusted and potentially malicious content and have visibility of sensitive user data, they are one of the most common attack vectors for bad actors. To help keep users safe online, Apple will only authorize developers to implement alternative browser engines after meeting specific criteria and who commit to a number of ongoing privacy and security requirements, including timely security updates to address emerging threats and vulnerabilities.

BrowserEngineKit isn’t easy, but I genuinely don’t think any good solution would be. Browsers don’t need a special entitlement or complex framework to run on MacOS, true, but iOS is not MacOS. To put it in Steven Sinofsky’s terms, gatekeeping is a fundamental aspect of Apple’s brand promise with iOS. 

WWDC 2024: Apple Intelligence

An oft-told story is that back in 2009 — two years after Dropbox debuted, two years before Apple unveiled iCloud — Steve Jobs invited Dropbox cofounders Drew Houston and Arash Ferdowsi to Cupertino to pitch them on selling the company to Apple. Dropbox, Jobs told them, was “a feature, not a product”.

It’s easy today to forget just how revolutionary a product Dropbox was. A simple installation on your Mac and boom, you had a folder that synced between every Mac you used — automatically, reliably, and quickly. At the time Dropbox had a big sign in its headquarters that read, simply, “It Just Works”, and they delivered on that ideal — at a time when no other sync service did. Jobs, of course, was trying to convince Houston and Ferdowsi to sell, but that doesn’t mean he was wrong that, ultimately, it was a feature, not a product. A tremendously useful feature, but a feature nonetheless.

Leading up to WWDC last week, I’d been thinking that this same description applies, in spades, to LLM generative AI. Fantastically useful, downright amazing at times, but features. Not products. Or at least not broadly universal products. Chatbots are products, of course. People pay for access to the best of them, or for extended use of them. But people pay for Dropbox too.

Chatbots can be useful. There are people doing amazing work through them. But they’re akin to the terminal and command-line tools. Most people just don’t think like that.

What Apple unveiled last week with Apple Intelligence wasn’t so much new products, but new features — a slew of them — for existing products, powered by generative AI.

Safari? Better now, with generative AI page summaries. Messages? More fun, with Genmoji. Notes and Mail and Pages (and any other app that uses the system text frameworks)? Better now, with proofreading and rewriting tools built-in. Photos? Even better recommendations for memories, and automatic categorization of photos into smart collections. Siri? That frustrating, dumb-as-a-rock son of a bitch, Siri? Maybe, actually, pretty useful and kind of smart now. These aren’t new apps or new products. They’re the most used, most important apps Apple makes, the core apps that define the Apple platforms ecosystem, and Apple is using generative AI to make them better and more useful — without, in any way, rendering them unfamiliar.1

We had a lot of questions about Apple’s generative AI strategy heading into WWDC. Now that we have the answers, it all looks very obvious, and mostly straightforward. First, their models are almost entirely based on personal context, by way of an on-device semantic index. In broad strokes, this on-device semantic index can be thought of as a next-generation Spotlight. Apple is focusing on what it can do that no one else can on Apple devices, and not really even trying to compete against ChatGPT et al. for world-knowledge context. They’re focusing on unique differentiation, and eschewing commoditization.

Second, they’re doing both on-device processing, for smaller/simpler tasks, and cloud processing (under the name Private Cloud Compute) for more complex tasks. All of this is entirely Apple’s own work: the models, the servers (based on Apple silicon), the entire software stack running on the servers, and the data centers where the servers reside. This is an enormous amount of work, and seemingly puts the lie to reports that Apple executives only even became interested in generative AI 18 months ago. And if they did accomplish all this in just 18 months that’s a remarkable achievement.

Anyone can make a chatbot. (And, seemingly, everyone is — searching for “chatbot” in the App Store is about as useful as searching for “game”.) Apple, conspicuously, has not made one. Benedict Evans keenly observes:

To begin, then: Apple has built an LLM with no chatbot. Apple has built its own foundation models, which (on the benchmarks it published) are comparable to anything else on the market, but there’s nowhere that you can plug a raw prompt directly into the model and get a raw output back - there are always sets of buttons and options shaping what you ask, and that’s presented to the user in different ways for different features. In most of these features, there’s no visible bot at all. You don’t ask a question and get a response: instead, your emails are prioritised, or you press ‘summarise’ and a summary appears. You can type a request into Siri (and Siri itself is only one of the many features using Apple’s models), but even then you don’t get raw model output back: you get GUI. The LLM is abstracted away as an API call.

Instead Apple is doing what no one else can do: integrating generative AI into the frameworks in iOS and MacOS used by developers to create native apps. Apps built on the system APIs and frameworks will gain generative AI features for free, both in the sense that the features come automatically when the app is running on a device that meets the minimum specs to qualify for Apple Intelligence, and in the sense that Apple’s isn’t charging developers or users to utilize these features.

Apple’s keynote presentation was exceedingly well-structured and paced. But nevertheless it was widely misunderstood, I suspect because expectations were so wrong. Those who believed going in that Apple was far behind the state of the art in generative AI technology wrongly saw the keynote’s coda — the announcement of a partnership with OpenAI to integrate their latest model, ChatGPT-4o, as an optional “world knowledge” layer sitting atop Apple’s own homegrown Apple Intelligence — as an indication that most or even all of the cool features Apple revealed were in fact powered by OpenAI. Quite the opposite. Almost nothing Apple showed in the keynote was from OpenAI.

What I see as the main takeaways:

  • Apple continues to build machine learning and generative AI features across its core platforms, iOS and MacOS. They’ve been adding such features for years, and announced many new ones this year. Nothing Apple announced in the entire first hour of the keynote was part of “Apple Intelligence”. Math Notes (freeform handwritten or typed mathematics, in Apple Notes and the Calculator app, which is finally coming to iPadOS) is coming to all devices running iOS 18 and MacOS 15 Sequoia. Smart Script — the new personalized handwriting feature when using Apple Pencil, which aims to improve the legibility of your handwriting as you write, and simulates your handwriting when pasting text or generating answers in Math Notes — is coming to all iPads with an A14 or better chip. Inbox categorization and smart message summaries are coming to Apple Mail on all devices. Safari web page summaries are coming to all devices. Better background clipping (“greenscreening”) for videoconferencing. None of these features are under the “Apple Intelligence” umbrella. They’re for everyone with devices eligible for this year’s OS releases.

  • The minimum device specs for Apple Intelligence are understandable, but regrettable, particularly the fact that the only current iPhones that are eligible are the iPhone 15 Pro and Pro Max. Even the only-nine-month-old iPhone 15 models don’t make the cut. When I asked John Giannandrea (along with Craig Federighi and Greg Joswiak) about this on stage at The Talk Show Live last week, his answer was simple: lesser devices aren’t fast enough to provide a good experience. That’s the Apple way: better not to offer the feature at all than offer it with a bad (slow) experience. A-series chips before last year’s A17 Pro don’t have enough RAM and don’t have powerful enough Neural Engines. But by the time Apple Intelligence features actually become available — even in beta form (they are not enabled in the current developer OS betas) — the iPhone 15 Pro will surely be joined by all iPhone 16 models, both Pro and non-pro. Apple Intelligence is skating to where the puck is going to be in a few years, not where it is now.

  • Surely Apple is also being persnickety with the device requirements to lessen the load on its cloud compute servers. And if this pushes more people to upgrade to a new iPhone this year, I doubt Tim Cook is going to see that as a problem.

  • One question I’ve been asked repeatedly is why devices that don’t qualify for Apple Intelligence can’t just do everything via Private Cloud Compute. Everyone understands that if a device isn’t fast or powerful enough for on-device processing, that’s that. But why can’t older iPhones (or in the case of the non-pro iPhones 15, new iPhones with two-year-old chips) simply use Private Cloud Compute for everything? From what I gather, that just isn’t how Apple Intelligence is designed to work. The models that run on-device are entirely different models than the ones that run in the cloud, and one of those on-device models is the heuristic that determines which tasks can execute with on-device processing and which require Private Cloud Compute or ChatGPT. But, see also the previous item in this list — surely Apple has scaling concerns as well. As things stand, with only devices using M-series chips or the A17 or later eligible, Apple is going to be on the hook for an enormous amount of server-side computation with Private Cloud Compute. They’d be on the hook for multiples of that scale if they enabled Apple Intelligence for older iPhones, with those older iPhones doing none of the processing on-device. The on-device processing component of Apple Intelligence isn’t just nice-to-have, it’s a keystone to the entire thing.

  • Apple could have skipped, or simply delayed announcing until the fall, the entire OpenAI partnership, and they still would have had an impressive array of generative AI features with broad, practical appeal. And clearly they would have gotten a lot more credit for their achievements in the aftermath of the keynote. I remain skeptical that integrating ChatGPT (and any future world-knowledge LLM chatbot partners) at the OS level will bring any significant practical advantage to users versus just using the chatbot apps from the makers of those LLMs. But perhaps removing a few steps, and eliminating the need to choose, download, and sign up for a third-party chatbot, will expose such features to many more users than who are using them currently. But I can’t help but feel that integrating these third-party chatbots in the OSes is at least as much a services-revenue play as a user-experience play.

  • The most unheralded aspect of Apple Intelligence is that the data centers Apple is building for Private Cloud Compute are not only carbon neutral, but are operating entirely on renewable energy sources. That’s extraordinary, and I believe unique in the entire industry. But it’s gone largely un-remarked-upon — because Apple itself did not mention this during the WWDC keynote. Craig Federighi first mentioned it in a post-keynote interview with Justine Ezarik, and he reiterated on stage with me at The Talk Show Live From WWDC. In hindsight, I wish I’d asked, on stage, why Apple did not even mention this during the keynote, let alone trumpet it. I suspect the real answer is that Apple felt like they couldn’t brag about their own data centers running entirely on renewable energy during the same event in which they announced a partnership with OpenAI, whose data centers can make no such claims. OpenAI’s carbon footprint is a secret, and experts suspect it’s bad. It’s unseemly to throw your own partner under the bus, but that takes Apple Intelligence’s proclaimed carbon neutrality off the table as a marketing point. Yet another reason why I feel Apple might have been better off not announcing this partnership last week.

  • If you don’t want or don’t trust Apple Intelligence (or just not yet), you’ll be able to turn it off. And you’ll have to opt-in to using the integrated ChatGPT feature, and, each time Apple Intelligence decides to send you to ChatGPT to handle a task, you’ll have to explicitly allow it. As currently designed, no one is going to accidentally interact with, let alone expose personal information to, ChatGPT. If anything I suspect the more common complaint will come from people who wish to use ChatGPT without confirmation each time. Some people are going to want an “Always allow” option for handing requests to ChatGPT, but according to Apple reps I’ve spoken with, such an option does not yet exist.

  • At a technical level Apple is using indirection to anonymize devices from ChatGPT. OpenAI will never see your IP address or precise location. At a policy level, OpenAI has agreed not to store user data, nor use data for training purposes, unless users have signed into a ChatGPT account. If you want to use Apple Intelligence but not ChatGPT, you can. If you want to use ChatGPT anonymously, you can. And if you do want ChatGPT to keep a history of your interactions, you can do that too, by signing in to your account. Users are entirely in control, as they should be.

  • VisionOS 2 is not getting any Apple Intelligence features, despite the fact that the Vision Pro has an M2 chip. One reason is that VisionOS remains a dripping-wet new platform — Apple is still busy building the fundamentals, like rearranging and organizing apps in the Home view. VisionOS 2 isn’t even getting features like Math Notes, which, as I mentioned above, isn’t even under the Apple Intelligence umbrella. But another reason is that, according to well-informed little birdies, Vision Pro is already making significant use of the M2’s Neural Engine to supplement the R1 chip for real-time processing purposes — occlusion and object detection, things like that. With M-series-equipped Macs and iPads, the Neural Engine is basically sitting there, fully available for Apple Intelligence features. With the Vision Pro, it’s already being used.

  • “Apple Intelligence” is not one thing or one model. Or even two models — local and cloud. It’s an umbrella for dozens of models, some of them very specific. One of the best, potentially, is a new model that will allow Siri to answer technical support questions about Apple products and services. This model has been trained on Apple’s own copious Knowledge Base of support documentation. The age-old gripe is that “no one reads the documentation”, but maybe now that’s no longer a problem because Siri is reading it. Apple’s platforms are so rich and deep, but most users’ knowledge of them is shallow; getting correct answers from Siri to specific how-to questions could be a game-changer. AI-generated slop is polluting web search results for technical help; Apple is using targeted AI trained on its own documentation to avoid the need to search the web in the first place. Technical documentation isn’t sexy, but exposing it all through natural language queries could be one of the sleeper hits of this year’s announcements.

  • Xcode is the one product where Apple was clearly behind on generative AI features. It was behind on LLM-backed code completion/suggestion/help last year. Apple introduced two generative AI features in Xcode 16, and they exemplify the local/cloud distinction in Apple Intelligence in general. Predictive code completion runs locally, on your Mac. Swift Assist is more profound, answering natural language questions and providing entire solutions in working Swift code, and runs entirely in Private Cloud Compute.

Take It All With a Grain of Salt

Lastly, it is essential to note that we haven’t been able to try any of these Apple Intelligence features yet. None of them are yet available in the developer OS betas, and none are slated to be available, even in beta, until “later this summer”. I witnessed multiple live demos of some of these features last week, during press briefings at Apple Park after the keynote. Demos I witnessed included the writing tools (“make this email sound more professional”) and Xcode code completion and Swift Assist. But those demos were conducted by Apple employees; we in the media were not able to try them ourselves.

It all looks very impressive, and almost all these features seem very practical. But it’s all very, very early. None of it counts as real until we’re able to use it ourselves. We don’t know how well it works. We don’t know how well it scales.

If generative AI weren’t seen as essential — both in terms of consumer marketing and investor confidence — I think much, if not most, of what Apple unveiled in “Apple Intelligence” wouldn’t even have been announced until next year’s WWDC, not last week’s WWDC. Again, none of the features in “Apple Intelligence” are even available in beta yet, and I think all or most of them will be available only under a “beta” label until next year.

It’s good to see Apple hustling, though. I continue to believe it’s incorrect to see Apple as “behind”, overall, on generative AI. But clearly they are feeling tremendous competitive pressure on this front, which is good for them, and great for us. 

  1. Image Playground is a new app, and thus definitely counts as a product, but at the moment I’m seeing it as the least interesting part of Apple Intelligence, if only because it’s offering something a dozen other products offer, and it doesn’t seem to do a particularly interesting job of it. ↩︎

Kolide by 1Password 

My thanks to Kolide by 1Password for sponsoring last week at DF. The September 2023 MGM hack is one of the most notorious ransomware attacks in recent years. Journalists and cybersecurity experts rushed to report on the broken slot machines, angry hotel guests, and the fateful phishing call to MGM’s help desk that started it all.

But while it’s true that MGM’s help desk needed better ways of verifying employee identity, there’s another factor that should have stopped the hackers in their tracks. That’s where you should focus your attention. In fact, if you just focus your vision, you’ll find you’re already staring at the security story the pros have been missing.

It’s the device you’re reading this on.

To read more about what they learned after researching the MGM hack — like how hacker groups get their names, the worrying gaps in MGM’s security, and why device trust is the real core of the story — check out the Kolide by 1Password blog.

Training Large Language Models on the Public Web

Yesterday, quoting Anthropic’s announcement of their impressive new model, Claude 3.5 Sonnet, I wrote:

Also, from the bottom of the post, this interesting nugget:

One of the core constitutional principles that guides our AI model development is privacy. We do not train our generative models on user-submitted data unless a user gives us explicit permission to do so. To date we have not used any customer or user-submitted data to train our generative models.

Even Apple can’t say that.

It now seems clear that I misread Anthropic’s statement. I wrongly interpreted this as implying that Claude was not trained on public web data. Here is Anthropic’s FAQ on training data:

Large language models such as Claude need to be “trained” on text so that they can learn the patterns and connections between words. This training is important so that the model performs effectively and safely.

While it is not our intention to “train” our models on personal data specifically, training data for our large language models, like others, can include web-based data that may contain publicly available personal data. We train our models using data from three sources:

  1. Publicly available information via the Internet
  2. Datasets that we license from third party businesses
  3. Data that our users or crowd workers provide

We take steps to minimize the privacy impact on individuals through the training process. We operate under strict policies and guidelines for instance that we do not access password protected pages or bypass CAPTCHA controls. We undertake due diligence on the data that we license. And we encourage our users not to use our products and services to process personal data. Additionally, our models are trained to respect privacy: one of our constitutional “principles” at the heart of Claude, based on the Universal Declaration of Human Rights, is to choose the response that is most respectful of everyone’s privacy, independence, reputation, family, property rights, and rights of association.

Here is Apple, from its announcement last week of their on-device and server foundation models:

We train our foundation models on licensed data, including data selected to enhance specific features, as well as publicly available data collected by our web-crawler, AppleBot. Web publishers have the option to opt out of the use of their web content for Apple Intelligence training with a data usage control.

We never use our users’ private personal data or user interactions when training our foundation models, and we apply filters to remove personally identifiable information like social security and credit card numbers that are publicly available on the Internet. We also filter profanity and other low-quality content to prevent its inclusion in the training corpus. In addition to filtering, we perform data extraction, deduplication, and the application of a model-based classifier to identify high quality documents.

This puts Apple in the same boat as Anthropic in terms of using public pages on the web as training sources. Some writers and creators object to this — including Federico Viticci, whose piece on MacStories I linked to with my “Even Apple can’t say that” comment yesterday. Dan Moren wrote a good introduction to blocking these crawling bots with robots.txt directives.

The best argument against Apple’s use of public web pages for model training is that they trained first, but only after announcing Apple Intelligence last week issued the instructions for blocking Applebot for AI training purposes. Apple should clarify whether they plan to re-index the public data they used for training before Apple Intelligence ships in beta this summer. Clearly, a website that bans Applebot-Extended shouldn’t have its data in Apple’s training corpus simply because Applebot crawled it before Apple Intelligence was even announced. It’s fair for public data to be excluded on an opt-out basis, rather than included on an opt-in one, but Apple trained its models on the public web before they allowed for opting out.

But other than that chicken/egg opt-out issue, I don’t object to this. The whole point of the public web is that it’s there to learn from — even if the learner isn’t human. Is there a single LLM that was not trained on the public web? To my knowledge there is not, and a model that is ignorant of all information available on the public web would be, well, pretty ignorant of the world. To me the standards for LLMs should be similar to those we hold people to. You’re free to learn from anything I publish, but not free to plagiarize it. If you quote it, attribute and link to the source. That’s my standard for AI bots as well. So at the moment, my robots.txt file bans just one: Perplexity.

(I’d block a second, the hypocrites at Arc, if I could figure out how.) 

By My Count Trump Is Batting .900 on the Ten Commandments

Sara Cline, reporting for the AP:

The legislation that Republican Gov. Jeff Landry signed into law on Wednesday requires a poster-sized display of the Ten Commandments in “large, easily readable font” in all public classrooms, from kindergarten to state-funded universities.

“If you want to respect the rule of law, you’ve got to start from the original lawgiver, which was Moses” who got the commandments from God, Landry said.

Opponents questioned the law’s constitutionality and vowed to challenge it in court. Proponents said the measure is not solely religious, but that it has historical significance. In the language of the law, the Ten Commandments are “foundational documents of our state and national government.”

Former president and convicted felon Donald Trump, on his floundering social network, approves:


Here is the Protestant version of the Ten Commandments required by the Louisiana law (the Catholic version doesn’t qualify):

1. Thou shalt have no other gods before me.

The Independent, in March: “Trump Compares Himself to Jesus Christ — Again”.

2. Thou shalt not make to thyself any graven images.

The Guardian: “Trump Used His Charity’s Money to Pay for Portrait of Himself”.

3. Thou shalt not take the Name of the Lord thy God in vain.

Politico, 2019: “‘Using the Lord’s Name in Vain’: Evangelicals Chafe at Trump’s Blasphemy”. (Trump crowing, “They’ll be hit so goddamn hard,” while bragging about bombing Islamic State militants. And Trump recounting his warning to a wealthy businessman: “If you don’t support me, you’re going to be so goddamn poor.”)

4. Remember the Sabbath day, to keep it holy.

People magazine, 2022: “Donald Trump spent the Easter weekend enjoying two of his favorite activities, sources say: golf and greeting adoring guests at his Mar-a-Lago resort. On Saturday and Sunday morning, the former president played rounds of golf with members at Trump International Golf Club in West Palm Beach, Florida, a source tells People. ‘He is no longer president,’ says one insider, ‘so he doesn’t have to go to church.’”

5. Honor thy father and thy mother, that thy days may be long upon the land which the Lord thy God giveth thee.

Trump speaks highly of his father — from whom he inherited his handsome visage and enviable mane — so he’s clear on this one. Quite the man to honor, too. The Washington Post: “Trump’s father, Fred Trump, was arrested twice: in 1927 during a Ku Klux Klan riot, and in 1976 over code violations at a building he owned in Maryland.” The New York Times: “The Justice Department undertook its own investigation and, in 1973, sued Trump Management for discriminating against blacks. Both Fred Trump, the company’s chairman, and Donald Trump, its president, were named as defendants. It was front-page news, and for Donald, amounted to his debut in the public eye.”

6. Thou shalt not kill.

Trump, 2016: “You know what else they say about my people? The polls, they say I have the most loyal people. Did you ever see that? Where I could stand in the middle of Fifth Avenue and shoot somebody and I wouldn’t lose any voters, OK? It’s like incredible.”

The Guardian, 2021: “US Could Have Averted 40% of Covid Deaths, Says Panel Examining Trump’s Policies”.

7. Thou shalt not commit adultery.

I mean come on.

8. Thou shalt not steal.

Trump University: “The lawsuits centered around allegations that Trump University defrauded its students by using misleading marketing practices and engaging in aggressive sales tactics. The company and the lawsuits against it received renewed interest due to Trump’s candidacy in the 2016 presidential election. Despite repeatedly insisting he would not settle, Trump settled all three lawsuits in November 2016 for a total of $25 million after being elected president.”

9. Thou shalt not bear false witness against thy neighbor.

CNN: “Former President Donald Trump has spent months spreading lies about the 2020 election, which he himself is now calling “THE BIG LIE” as he continues to claim that a massive conspiracy robbed him of a second term.”

10. Thou shalt not covet thy neighbor’s house.

The Vera Coking house: “In 1993, Donald Trump bought several lots around his Atlantic City casino and hotel, intending to build a parking lot designed for limousines. Coking, who had lived in her house at that time for 32 years, refused to sell. As a result, the city condemned her house, using the power of eminent domain. She was offered $251,000, a quarter of what she was offered by Guccione 10 years earlier.” 

The EU Is Reaping What It Sows With the DMA: Uncertainty

Ian Betteridge:

So, Apple, which bits of the DMA does Apple Intelligence violate? Because unless you can actually tell us — which case we clearly have a bit of a problem with some of the claims you’ve made about how it works — or you’re talking bullshit, and just trying to get some leverage with the EU. Which is it Tim?

I absolutely guarantee that people are going to swallow this “well you can’t make Apple Intelligence work thanks to the DMA!!” line without actually asking any questions about what it violates, because “well Apple said it and they don’t lie evah”

That’s Apple’s entire point. They don’t know. It’s uncertain by design. EC proponents keep telling me it’s a feature, not a bug, that unlike the US, it’s the spirit, not letter, of the law that matters in the EU. So it doesn’t matter if there’s not a word in the DMA that disallows the Core Technology Fee; European Commissioners have decided it goes against the spirit of the DMA, so they’re going to charge and fine Apple.

Outgoing competition chief Margrethe Vestager (her 10-year term expires in November) gave a 20-minute interview to CNBC’s Silvia Amaro this week, in which she threatened Apple repeatedly, starting around the 13:00 mark.

Vestager: We have a number of Apple issues; I find them very serious. I was very surprised that we would have such suspicions of Apple being non-compliant. They are very important because a lot of good business happens through the App Store, happens through payment mechanisms, so of course, even though you know I can say this is not what was expected of such a company, of course we will enforce exactly with the same top priority as with any other business. [...] What I see with the Digital Markets Act is that we get closer and closer to the business model. The changes that the DMA is demanding, they are changes that will touch the business model.

Amaro: Breakups even?

Vestager: Well, it remains to be seen. But if you have to carry a second app store, that will make a dent in your own app store. If you cannot promote your own services, but need to give room for competitors, rightfully, legitimately so, of course that will potentially, sort of, take away some of your own profits. So of course this will be a very difficult enforcement task, but even so much more rewarding, because the potential for the competitors of Big Tech — they are really, really big — and this is why enforcement of the DMA is a top priority.

Amaro: The reason why I am also bringing up Apple is because companies such as Spotify have also raised issues with the new changes that Apple has put forward, that ultimately the bill they will face now is the same or even bigger than what was in place before. And this is why I want to bring up the issue again of whether the DMA is actually leading to ultimate changes for the Big Tech? Or are they still hiding here, in between some of the ... well, within the law?

Vestager: Well, I think that you cannot judge the DMA before we’ve had the cases, and before we finalize them, because we have a toolbox of fines, of doubling fines, of potential breakup of companies — so we have a very strong toolbox to punish.

So how is Apple (or Meta, or Google) supposed to launch new features, integrating its own new services deeply within and between its own platforms, while under repeated threats of massive fines, fines that are far out of proportion to the revenue Apple generates in the EU itself, and even — laughably, admittedly — being “broken up”? When Vestager is very clear that Apple will only be deemed compliant with the DMA if it adversely affects Apple’s profit? And when it’s the unwritten “spirit” of the DMA, not the letter of its ambiguous rules, that matters?

This isn’t about privacy or the fact that Apple Intelligence models were trained on data scraped from the public web. Such factors might play a role in Apple Intelligence’s compliance, but not iPhone Mirroring or the new SharePlay screen sharing. This is about the DMA’s restrictions on designated gatekeepers launching their own integrated services and features.

Under repeated threats of fines up to $40–80 billion dollars (10–20 percent of worldwide revenue), it would be recklessly irresponsible for Apple, or any other designated “gatekeeper”, to launch any new services or integrated features in the EU without absolute certainty that those features are compliant with the DMA. And the nature of the European Commission is that they do not issue such assurances in advance. This is not spite. Spite would be saying these features will never come to the EU while the DMA remains in place. But a delayed rollout is the only rational response to the DMA: extreme caution in the face of the law’s by-design uncertainty and severe penalties.

Fortunately, because Apple is delaying Apple Intelligence and these other new features in the EU, all of the thriving EU-based smartphone and OS makers can jump in and compete on merit now, without Apple the gatekeeping bully in their way. As Vestager reiterates throughout the interview, competition is the European Commission’s north star. 

Reggie Jackson on Willie Mays’s Legacy, and the Abject Racism Faced by Black Baseball Players in the 1960s 

The whole 8-minute clip is excellent and worth your time, but do not miss the second half, starting with a sharp question from Alex Rodriguez at the 4:30 mark. Reggie describes, in heartfelt detail, the abject racism he faced as a minor league player as recently as the 1960s. Restaurants he couldn’t eat at. Hotels he couldn’t stay at. Threats to burn to the ground the apartment building where he was sleeping. The pain, over five decades later, remains searing.

Kudos to Fox Sports for airing this. We can’t celebrate progress without honestly facing society’s dark past. (Kudos too, for putting a box of Reggie Bars at the desk. Respect.)

EU Users Won’t Get Apple Intelligence, iPhone Mirroring, or the New SharePlay Screen Sharing Features This Year, Thanks to the DMA 

The Financial Times:

Apple blamed complexities in making the system compatible with EU rules that have forced it to make key parts of its iOS software and App Store services interoperable with third parties.

“Due to the regulatory uncertainties brought about by the Digital Markets Act,” Apple said on Friday, “we do not believe that we will be able to roll out three of these features — iPhone Mirroring, SharePlay Screen Sharing enhancements, and Apple Intelligence — to our EU users this year.”

Kudos to Apple for breaking this news to the Financial Times, of all outlets. Poetry in media relations. Here’s the full on-the-record statement, provided to me by an Apple spokesperson:

Two weeks ago, Apple unveiled hundreds of new features that we are excited to bring to our users around the world. We are highly motivated to make these technologies accessible to all users. However, due to the regulatory uncertainties brought about by the Digital Markets Act (DMA), we do not believe that we will be able to roll out three of these features — iPhone Mirroring, SharePlay Screen Sharing enhancements, and Apple Intelligence — to our EU users this year.

Specifically, we are concerned that the interoperability requirements of the DMA could force us to compromise the integrity of our products in ways that risk user privacy and data security. We are committed to collaborating with the European Commission in an attempt to find a solution that would enable us to deliver these features to our EU customers without compromising their safety.

None of these features are available yet in the developer beta OS releases, but it is my understanding that the first two — iPhone Mirroring and the new SharePlay Screen Sharing enhancements (where you’ll be able to see and doodle on the screens of others, like, say, if you’re providing remote help or how-to instructions to a friend or family member) — will be in the next developer betas, coming early next week. Apple Intelligence won’t even enter beta until later this summer. But in the meantime, even in beta, none of these features will be available within the EU.

The Mac is not considered a “gatekeeping” platform in the EU, but the iPhone and iPad are, and the iPhone Mirroring and screen sharing features obviously involve those platforms. I think Apple could try to thread a needle here and release Apple Intelligence only on the Mac in the EU, but given how inscrutable the European Commission’s interpretation of the DMA is — where gatekeepers are expected to somehow suss out the “spirit of the law” regardless of what the letter of the law says — I don’t see how Apple can be blamed for pausing the rollout in the EU, no matter the platform.

The EU’s self-induced slide into a technological backwater continues.

Matt Levine on OpenAI’s True Purpose 

Matt Levine, in his Money Stuff column:

OpenAI was founded to build artificial general intelligence safely, free of outside commercial pressures. And now every once in a while it shoots out a new AI firm whose mission is to build artificial general intelligence safely, free of the commercial pressures at OpenAI.

Anthropic Introduces Claude 3.5 Sonnet 


Claude 3.5 Sonnet sets new industry benchmarks for graduate-level reasoning (GPQA), undergraduate-level knowledge (MMLU), and coding proficiency (HumanEval). It shows marked improvement in grasping nuance, humor, and complex instructions, and is exceptional at writing high-quality content with a natural, relatable tone.

Claude 3.5 Sonnet operates at twice the speed of Claude 3 Opus. This performance boost, combined with cost-effective pricing, makes Claude 3.5 Sonnet ideal for complex tasks such as context-sensitive customer support and orchestrating multi-step workflows.

In an internal agentic coding evaluation, Claude 3.5 Sonnet solved 64% of problems, outperforming Claude 3 Opus which solved 38%. Our evaluation tests the model’s ability to fix a bug or add functionality to an open source codebase, given a natural language description of the desired improvement. When instructed and provided with the relevant tools, Claude 3.5 Sonnet can independently write, edit, and execute code with sophisticated reasoning and troubleshooting capabilities. It handles code translations with ease, making it particularly effective for updating legacy applications and migrating codebases.

I’ll take them with a grain of self-promoting salt, but the evaluation tests presented by Anthropic position Claude 3.5 Sonnet as equal to or better than ChatGPT-4o. Again: I don’t think there’s a moat in this game.

Also, from the bottom of the post, this interesting nugget:

One of the core constitutional principles that guides our AI model development is privacy. We do not train our generative models on user-submitted data unless a user gives us explicit permission to do so. To date we have not used any customer or user-submitted data to train our generative models.

Even Apple can’t say that. Follow-Up:Training Large Language Models on the Public Web”.

Warp’s Agent Mode: Natural Language for Terminal Commands 

Michelle Lim, writing for the Warp blog (Warp, I’ll note, has previously sponsored DF):

Agent Mode raises the level of abstraction in the terminal. Using natural language, you can ask the terminal to accomplish any high level task without worrying about the specific commands you need.

Agent Mode knows when it needs more information to help you complete a task. It will ask permission to run commands on your machine and use the outputs to guide you, step-by-step. It’s also self-correcting when it runs into errors.

Because Agent Mode executes tasks through the CLI, it integrates with practically any service with zero configuration. If the service has a CLI, an API, or publicly available docs, you can use Agent Mode for the task. Agent Mode has inherent knowledge of most public CLIs, and you can easily teach it how to use internal CLIs by asking it to read their help content.

Man (no pun intended?), is this cool. I just tried “How do I generate a random number between 0000 and 9999?” and got:

You can generate a random number between 0000 and 9999 using the following shell command in the Fish shell:

  printf "%04d\n" (random 0 9999)

And when I followed up with “How can I use jot to generate a random number between 0000 and 9999?” I got:

You can use the jot command to generate a random number between 0000 and 9999 using the following command:

  jot -w "%04d" -r 1 0 9999

Both of which answers are correct. For jot — a tool I first learned about, of course, from the inimitable Dr. Drang — I think a simpler, and thus better, answer is:

jot -r 1 0000 9999

but Warp’s Agent Mode suggestion is certainly good enough.

Lacking Votes, EU Postpones Vote on CSAM Law That Would Ban End-to-End-Encryption for Messaging 

Clothilde Goujard, reporting for Politico:

A vote scheduled today to amend a draft law that may require WhatsApp and Signal to scan people’s pictures and links for potential child sexual abuse material was removed from European Union countries’ agenda, according to three EU diplomats.

Ambassadors in the EU Council were scheduled to decide whether to back a joint position on an EU regulation to fight child sexual abuse material (CSAM). But many EU countries including Germany, Austria, Poland, the Netherlands and the Czech Republic were expected to abstain or oppose the law over cybersecurity and privacy concerns.

“In the last hours, it appeared that the required qualified majority would just not be met,” said an EU diplomat from the Belgian presidency, which is spearheading negotiations until end June as chair of the EU Council.

Sanity prevails — for now.

‘This $8 Cardboard Knife Will Change Your Life’ 

Matthew Panzarino, writing at The Obsessor:

The cardboard is inescapable if you use Amazon or other online stores, they pile up in the hallways and next to the garbage cans and you triage as you can.

We get so many that I have to break down our boxes in order to fit them in our recycle bin. I’ve used all of the typical tools — scissors, pocket knife, box cutter — and many unconventional ones like drywall saws just trying to make this painful job a bit easier.

The CANARY is uniquely serrated all the way around its edge, like a chainsaw. This makes it incredibly good at cutting cardboard either with or across corrugation with ease. I cannot express how easily this knife cuts cardboard, it’s like slicing through regular old paper, it’s amazing.

Last year when I linked to (and recommended) Studio Neat’s Keen — the world’s best box cutter, but which costs about $100 — at least one reader recommended the Canary. For $8 I figured why not. It truly is an amazing product. I do still love my prototype Keen but for opening and breaking down cardboard boxes, the Canary can’t be beat. It’s both highly effective and very safe.

‘Fast Crimes at Lambda School’ 

Ben Sandofksy went deep on the history of Lambda School, a learn-to-code startup that aimed to disrupt computer science education, and its founder, Austen Allred:

What set his boot camp apart from the others were “Income Share Agreements.” Instead of paying up-front for tuition, students agreed to pay a portion of future income. If you don’t get a job, you pay nothing. It was an idea so clever it became a breakout hit of Y Combinator, the same tech incubator that birthed Stripe, AirBnb, and countless other unicorns.

When Lambda School launched in 2017, critics likened ISAs to indentured servitude, but by 2019 it was Silicon Valley’s golden child. Every day, Austen tweeted jaw-dropping results. [...]

Things got worse from there, and we’ll get to it. First I need to address a common question: what do I have to do with any of this? I have no professional or personal connections to the company or the team. What compelled me to follow this story for the last five years?

On the surface, this is another window into the 2010’s tech bubble, a period where mediocre people could raise ludicrous money amid a venture capitalist echo chamber fueled by low-interest rates. But what makes this any worse than Juicero, Clinkle, or Humane? Why does this rise to the level of Theranos?

These stories hinge on their villains, whose hubris and stupidity end in comeuppance. Theranos had Elizabeth Holmes, Fyre Festival had Bobby McFarlane, and Lambda School has Austen Allred.

Independent journalism at its best.

Apple ID to Be Renamed to Apple Account 

Adam Engst, TidBITS:

The real problem comes when tech writers document features across multiple versions of Apple’s operating systems. We’ll probably use both terms for a while before slowly standardizing on the new term. Blame Apple if you see awkward sentences like “Continuity features require that you be logged into the same Apple Account (Apple ID in pre-2024 operating systems).” Or maybe writers will compress further to “Continuity features require that you be logged into the same Apple Account/ID.”

I do think “Apple Account” is a better name, so I think the transitional pain is worthwhile.

Perplexity AI Is Lying About Their User Agent 

Robb Knight:

I put up a post about blocking AI bots after the block was in place, so assuming the user agents are sent, there’s no way Perplexity should be able to access my site. So I asked:

What is this post about

I got a perfect summary of the post including various details that they couldn’t have just guessed. Read the full response here. So what the fuck are they doing?

I checked a few sites and this is just Google Chrome running on Windows 10. So they’re using headless browsers to scrape content, ignoring robots.txt, and not sending their user agent string. I can’t even block their IP ranges because it appears these headless browsers are not on their IP ranges.

Terrific, succinct write-up documenting that Perplexity has clearly been reading and indexing web pages that it is forbidden, by site owner policy, from reading and indexing — all contrary to Perplexity’s own documentation and public statements.

Wired: ‘Perplexity Is a Bullshit Machine’ 

Dhruv Mehrotra and Tim Marchman, reporting for Wired (News+ link):

A Wired analysis and one carried out by developer Robb Knight suggest that Perplexity is able to achieve this partly through apparently ignoring a widely accepted web standard known as the Robots Exclusion Protocol to surreptitiously scrape areas of websites that operators do not want accessed by bots, despite claiming that it won’t. Wired observed a machine tied to Perplexity — more specifically, one on an Amazon server and almost certainly operated by Perplexity — doing this on and across other Condé Nast publications.

The Wired analysis also demonstrates that despite claims that Perplexity’s tools provide “instant, reliable answers to any question with complete sources and citations included,” doing away with the need to “click on different links,” its chatbot, which is capable of accurately summarizing journalistic work with appropriate credit, is also prone to bullshitting, in the technical sense of the word.

This paints Perplexity as, effectively, an IP theft engine, and its CEO, Aravind Srinivas, as a degenerate liar. None of this is an oversight or just playing fast and loose. It’s a scheme to deliberately circumvent the plain intention of website owners not to have Perplexity index their sites. Liars and thieves. Utterly shameless.

A Rose by Any Other Name Would Smell as Sweet; An Encryption Back Door by Any Other Name Would Still Smell Like Shit 

Signal president Meredith Whittaker, responding to a new initiative in the EU to ban end-to-end-encryption (for some reason published as a PDF despite the fact that Signal has a blog):

In November, the EU Parliament lit a beacon for global tech policy when it voted to exclude end-to-end encryption from mass surveillance orders in the chat control legislation. This move responded to longstanding expert consensus, and a global coalition of hundreds of preeminent computer security experts who patiently weighed in to explain the serious dangers of the approaches on the table — approaches that aimed to subject everyone’s private communications to mass scanning against a government-curated database or AI model of “acceptable” speech and content.

There is no way to implement such proposals in the context of end-to-end encrypted communications without fundamentally undermining encryption and creating a dangerous vulnerability in core infrastructure that would have global implications well beyond Europe.

Instead of accepting this fundamental mathematical reality, some European countries continue to play rhetorical games. They’ve come back to the table with the same idea under a new label. Instead of using the previous term “client-side scanning,” they’ve rebranded and are now calling it “upload moderation.” Some are claiming that “upload moderation” does not undermine encryption because it happens before your message or video is encrypted. This is untrue.

Yes, but it’s a great idea to let these same EU bureaucrats design how mobile software distribution should work.

Copilot Plus PCs, Where the ‘Plus’ Means More Dumb Stickers 

Paul Thurrott on Threads, after getting his new Samsung Galaxy Book4 Edge laptop:

Former Windows head Terry Myerson once told me the goal of partnering with Qualcomm on Windows on Arm was to “get those f#$%ing Intel stickers off of PCs.”

Mission accomplished, Terry. There are no Intel stickers on the new Qualcomm-based Copilot+ PCs.

Still covered with stickers. And as Thurrott’s photo hints at, and this screenshot from Tim Schofield’s unboxing video shows clearly, Samsung can’t even be bothered to apply the stickers straight. Looks like they were applied by a little kid. Screams “premium” experience.

Two of these stickers don’t even make sense. The Snapdragon one is obviously paid for by Qualcomm, the same way Intel pays PC makers to apply their stickers. But why would Samsung booger up its own laptops with stickers promoting their own Dynamic AMOLED 2X display technology? And what’s the deal with the Energy Star stickers? Who pays to put those on laptops and why?

Samsung Warns That Their New Snapdragon-Based PCs Aren’t Compatible With Fortnite or Some Adobe Apps 

Yang Jie and Jiyoung Sohn, reporting for The Wall Street Journal (News+ link):

Samsung’s Galaxy Book 4 Edge [sic], which went on sale Tuesday in the U.S., South Korea and some other markets, contains Qualcomm’s Snapdragon processor. It runs a version of Microsoft’s Windows 11 for PCs that uses technology from U.K.-based Arm.

On Wednesday, Samsung put a notice on its Korean-language product site listing applications that it currently determines are incompatible with the new laptop or can’t be installed. The list included some Adobe software as well as popular games including “League of Legends” and “Fortnite.”

Sounds like maybe Microsoft’s Prism isn’t as good as Apple’s Rosetta 2 after all? Or that Prism isn’t capable of running low-level anti-piracy (Adobe) and anti-cheating (Epic) rootkit-style system extensions?

The issues offer an early hint of the challenges some tech companies may face as they introduce new AI-powered computers and smartphones while seeking to maintain compatibility with existing software.

What an odd paragraph. This has nothing to do with phones, and the only “tech companies” affected are Microsoft, who makes Windows, and PC makers whose machines run Windows and have adopted Qualcomm’s new Snapdragon chips. Macs made the transition from Intel’s x86 architecture to Apple’s ARM-based Apple silicon without users noticing anything other than dramatically longer battery life and faster performance, including when running x86 software in emulation.

(I put a sic above because Samsung’s new laptops are named “Galaxy Book4 Edge”, with no space between the “Book” and the “4”. Great product name that rolls right off the tongue, as usual, from Samsung.)

FTC Lawsuit Alleges Adobe’s Cancellation Fees Are Illegal 

Ashley Belanger, reporting for Ars Technica:

The government’s heavily redacted complaint laid out Adobe’s alleged scheme, which starts with “manipulative enrollment practices.”

To lock subscribers into recurring monthly payments, Adobe would typically pre-select by default its most popular “annual paid monthly” plan, the FTC alleged. That subscription option locked users into an annual plan despite paying month to month. If they canceled after a two-week period, they’d owe Adobe an early termination fee (ETF) that costs 50 percent of their remaining annual subscription. The “material terms” of this fee are hidden during enrollment, the FTC claimed, only appearing in “disclosures that are designed to go unnoticed and that most consumers never see.” [...]

Because Adobe allegedly only alerted users to the ETF in fine print — by hovering over a small icon or clicking a hyperlink in small text — while the company’s cancellation flows made it hard to end recurring payments, the FTC is suing and accusing Adobe of deceptive practices under the FTC Act.

Adobe is too good a company to push dark-pattern subscription schemes like this. They should concede, apologize, and eliminate every subscription that isn’t a simple straightforward annual or monthly plan.

Mike Masnick: ‘The Surgeon General Is Wrong; Social Media Doesn’t Need Warning Labels’ 

Mike Masnick, writing for The Daily Beast:

We put health warnings on things that are inherently harmful, with little redeeming health value. That is, things that are actually toxins: nicotine, lead, poisons.

The complaints here are with speech. [...]

The American Psychological Association released a similar report, concluding: “Using social media is not inherently beneficial or harmful to young people.” Instead, it finds that when young people struggle with mental health, their online lives are often just a reflection of their offline lives.

Lots of other research has shown the same thing, yet Murthy’s call for health warnings never mentions all of this research that suggests social media is actually beneficial for many. Instead, he cites a few anecdotes of children who were bullied online. But bullying happened prior to social media, and we did not talk about putting health warnings on telephones or notepads or other forms of communication.

Just pure panic-driven ninny-ism. It’s like the whole nonsense with “trigger warnings”. Masnick brings up Reagan-era Surgeon General C. Everett Koop’s nonsensical crusade against video games in the 1980s. I’m also reminded of Tipper Gore’s campaign for warning labels on music albums.

OpenAI Expats Found ‘Safe Superintelligence Inc.’ 

Ilya Sutskever, Daniel Gross, and Daniel Levy:

We approach safety and capabilities in tandem, as technical problems to be solved through revolutionary engineering and scientific breakthroughs. We plan to advance capabilities as fast as possible while making sure our safety always remains ahead.

This way, we can scale in peace.

Our singular focus means no distraction by management overhead or product cycles, and our business model means safety, security, and progress are all insulated from short-term commercial pressures.

We are an American company with offices in Palo Alto and Tel Aviv, where we have deep roots and the ability to recruit top technical talent.

Sutskever was the chief scientist and cofounder of OpenAI, who launched a failed coup against Sam Altman earlier this year. I certainly hope they’re more safe than OpenAI is open.

(Via Techmeme.)

‘That’s Odd — Usually the Blood Gets Off at the Second Floor’ 

Speaking of Louie Mantia, back in 2011 he made a terrific wallpaper of the iconic hallway carpeting from the Overlook Hotel in The Shining. He just remade it, with even better color accuracy and texture. And to go along with it, a new wallpaper based on the carpet inside room 237. There ain’t nothing in room 237. But you ain’t got no business going in there anyway. So stay out. You understand? Stay out. But feel free to use the wallpaper.

Louie Mantia on Dark Mode App Icons 

Louie Mantia:

Apple’s announcement of “dark mode” icons has me thinking about how I would approach adapting “light mode” icons for dark mode. I grabbed 12 icons we made at Parakeet for our clients to illustrate some ways of going about it. [...]

Unfortunately, some icons appear to have lost or gained weight in dark mode. For example, the Settings gear didn’t change size in dark mode, but it appears to occupy less space because the dark circle around it blends with its background. That makes it appear smaller than the Find My icon, which now looks enormous next to FaceTime. This is a remnant of some questionable design choices in iOS 7 that have lingered now for the last decade.

That last sentence is the most diplomatic thing I’ve ever heard from Louie. What a splendid post this is — exemplary work to illustrate his advice.

Jon Stewart Talks About His Split With Apple on Matthew Belloni’s ‘The Town’ 

Interesting two-part interview, with far more information than we’ve heard about the demise of The Problem With Jon Stewart on Apple TV+. Part two is here; Overcast links to parts one and two; Apple Podcasts links to parts one and two.

Some nuggets:

  • The split seemed very much amicable. Stewart isn’t one to hold back, and he emphasized repeatedly there are no hard feelings. He even professed to getting his morning news in Apple News.

  • Apple paid the show’s staff for all of season 3, despite cancelling the show before production began. That’s nearly unheard of in the entertainment industry.

  • Stewart himself admits that season one more or less stunk.

Well worth a listen.

Willie Mays, Greatest Centerfielder in Baseball History, Dies at 93 

John Shea, the San Francisco Chronicle:

Willie Mays, the iconic and endearing “Say Hey Kid” who charmed countless fans with his brilliant athleticism and graceful style and was widely considered baseball’s greatest and most entertaining player, died Tuesday of heart failure. He was 93.

“My father has passed away peacefully and among loved ones,” Mays son, Michael Mays, said. “I want to thank you all from the bottom of my broken heart for the unwavering love you have shown him over the years. You have been his life’s blood.” [...]

Mays spent most of his 23-year playing career with the Giants, six in New York and 15 in San Francisco, making him a cherished superstar from coast to coast. He hit 660 home runs, made 24 All-Star appearances and won 12 Gold Gloves, which weren’t given out until Mays’ sixth season.

The consummate five-tool player, Mays was elite at hitting, power hitting, defending, throwing and baserunning, and his ability to out-think and out-smart the competition served as a valuable sixth tool.

My dad was a clerk in the Navy, stationed in New York for a stint in the late 1950s, and they’d give free tickets to servicemen to attend ballgames for all three teams in the city: the Yankees, Dodgers, and Giants. At the time, all three clubs had centerfielders destined for the Hall of Fame: Mickey Mantle, Duke Snider, and Mays. My dad has never wavered from his conviction that Willie Mays was the best baseball player he ever saw, hands down. He hit for power and average, ran like the wind, made catches no one else could make, and had a cannon for an arm.

Best ballplayer I’ve seen play was Mays’s godson, Barry Bonds.

Why Dolphin Isn’t Coming to the App Store (Spoiler: It Needs a JIT) 

“OatmealDome”, developer of an iOS fork of the GameCube and Wii emulator Dolphin:

Two weeks ago, Apple modified their App Store guidelines to allow retro game emulators in the App Store. This week, Delta, a multi-system emulator that was previously only available via AltStore, was released on the App Store. Since these events happened, we’ve been asked many times if we will submit DolphiniOS (our fork of Dolphin) to the App Store.

Unfortunately, no.

Apple still does not allow us to use a vital technology that is necessary for Dolphin to run with good performance: JIT. [...]

Unfortunately, Apple generally does not allow apps to use JIT recompilers on iOS. The only exceptions are Safari and alternative web browsers in Europe. We submitted a DMA interoperability request to Apple for JIT support, but Apple denied the request a few weeks ago.

Swift Playgrounds is an exception too. Apple trusts itself to use JIT compilation safely, but not third-party developers. That’s not unreasonable — but I’m not sure it’s compliant with the DMA.

Open Source PC Emulator UTM Blocked by Apple for Notarization, Including Through EU Marketplaces 

Michael Tsai:

This is confusing, but I think what Apple is saying is that, even with notarization, apps are not allowed to “download executable code.” Rule 2.5.2 says apps may not “download, install, or execute code” except for limited educational purposes. Rule 4.7 makes an exception to this so that retro game emulators and some other app types can run code “that is not embedded in the binary.” This is grayed out when you select “Show Notarization Review Guidelines Only”, meaning that the exception only applies within the App Store. Thus, the general prohibition remains in effect for App Marketplaces and Web Distribution. But it seems like this wasn’t initially clear to Apple, either, because the review process took two months.

This also seems inconsistent with the fact that the Delta emulator is allowed to be notarized outside the App Store. It doesn’t make much sense for the rules to be more lax within the App Store. I first thought the mistake was that Apple didn’t mean to gray out 4.7 for notarization. Then everything would make sense. But the clarification states that 4.7 is not intended to apply to notarization.

The bottom line for me is that Apple doesn’t want general-purpose emulators, it’s questionable whether the DMA lets it block them, and even siding with Apple on this it isn’t consistently applying its own rules.

Confusing indeed. Apple’s stance on this seems inscrutable and arbitrary: retro game emulators are, at long last, acceptable, but general PC emulators are not. Such arbitrary policy decisions related to the purpose of the app are fine for the App Store (legally speaking), but clearly not compliant with the DMA. That’s one of the few areas where the DMA is clear. Apple can, of course, ban (say) porno apps from the App Store, but can’t refuse to notarize them for distribution outside the App Store in the EU.

Apple has a security leg to stand on when it comes to JIT compilation, but the version of UTM (UTM SE) that was held up in review for two months, and ultimately rejected by Apple, doesn’t use a JIT. And because it doesn’t use a JIT, performance is poor; hence the UTM developers’ deeming it not worth fighting about. Apple goes into depth on the security challenges pertaining to JIT compilation in its documentation for BrowserEngineKit, the framework for developing non-WebKit browser engines for distribution in the EU. That’s ostensibly the reason why developers need a special entitlement to use a custom browser engine — JavaScript engines need a JIT to perform well but JITs pose a security risk.

In an earlier revision of this post, I suggested that Delta — Riley Testut’s excellent and wildly popular retro Nintendo console emulator for iOS — uses a JIT. There is indeed a JIT in Delta’s source code repository, but Testut informs me that it’s currently only enabled through the version of AltStore distributed through sideloading. Delta’s JIT is removed from the versions of Delta in the App Store and AltStore PAL (the EU app marketplace), because of Apple’s restrictions on JIT compilers. No app with a JIT is going to pass review by Apple, including for distribution outside the App Store in the EU. That restriction should, in theory, be permitted under the DMA on security grounds. But how the no-JIT version of UTM could be rejected for notarization, I do not see.

(Thinking about BrowserEngineKit makes me wonder: Now that over four months have passed since Apple announced its initial DMA compliance plans, has even a single browser developer announced plans to bring their own rendering engine to iOS in the EU? As far as I know the answer is no. It’s entirely possible Apple went to all the trouble of creating BrowserEngineKit for compliance with the DMA, but no one is actually going to use it because no browser developer deems the EU market worth forking their browser for, solely for distribution outside the App Store — while on the hook for the 50-euro-cents-per-download Core Technology Fee if such a browser becomes popular.)

Popular AI Chatbots — Including ChatGPT, Mistral, and Meta AI — Spread Russian Propaganda (Because of Course They Do) 

Ina Fried, reporting for Axios:

To conduct the study, NewsGuard entered prompts asking about narratives known to have been created by John Mark Dougan, an American fugitive who, per the New York Times, is creating and spreading misinformation from Moscow.

  • Entering 57 prompts into 10 leading chatbots, NewsGuard found they spread Russian disinformation narratives 32% of the time, often citing Dougan’s fake local news sites as a reliable source.
  • The chatbots presented as fact false reports, originating on those sites, about a supposed wiretap discovered at Donald Trump’s Mar-a-Lago residence and a nonexistent Ukrainian troll factory interfering with U.S. elections.
  • NewsGuard conducted its research on OpenAI’s ChatGPT-4,’s Smart Assistant, Grok, Inflection, Mistral, Microsoft’s Copilot, Meta AI, Anthropic’s Claude, Google Gemini and Perplexity.

Certainly worth researching, but I’d have been more surprised if these chatbots did not present Russian misinformation as fact. For chrissake they pass Onion articles, which are intended as parody, as fact. Russian misinformation is written with the intent of convincing the gullible that it’s legit; and LLMs are by nature gullible.

Scroll Reverser for MacOS 

I use both a mouse and trackpad at my desk, but, for whatever reason, prefer “natural scrolling” with the trackpad, but reverse scrolling with the wheel on my mouse. Apple’s control panels for Mouse and Trackpad don’t allow you to specify different scrolling directions per input device.

But that’s exactly what Scroll Reverser does. It’s a super-simple utility by Nick Moore that does one thing and does it really well. I’ve been recommending it for over 10 years. Free of charge and open source.

Apple Security Research: ‘Private Cloud Compute: A New Frontier for AI Privacy in the Cloud’ 


We designed Private Cloud Compute to ensure that privileged access doesn’t allow anyone to bypass our stateless computation guarantees.

First, we intentionally did not include remote shell or interactive debugging mechanisms on the PCC node. Our Code Signing machinery prevents such mechanisms from loading additional code, but this sort of open-ended access would provide a broad attack surface to subvert the system’s security or privacy. Beyond simply not including a shell, remote or otherwise, PCC nodes cannot enable Developer Mode and do not include the tools needed by debugging workflows.

Next, we built the system’s observability and management tooling with privacy safeguards that are designed to prevent user data from being exposed. For example, the system doesn’t even include a general-purpose logging mechanism. Instead, only pre-specified, structured, and audited logs and metrics can leave the node, and multiple independent layers of review help prevent user data from accidentally being exposed through these mechanisms. With traditional cloud AI services, such mechanisms might allow someone with privileged access to observe or collect user data.

Many details here, but many still to come.

Ming-Chi Kuo Says This Year’s Series 10 Apple Watches Will Increase in Size 

Ming-Chi Kuo:

The Series 10 will feature form factor upgrades, including larger screen sizes (increasing from 45mm/41mm to about 49mm/45mm) and a thinner design.

Unlike other devices Apple sells, Apple Watch sizes are given not by screen diagonal but by case height. So what Kuo is claiming is that the current “big” size will become the small size and the new big size will be much bigger. I find this very hard to believe. Since its inception Apple Watch has stood out among smartwatches for making models that are appropriate for people with small wrists.

Anecdotally, almost all women I see wearing an Apple Watch are wearing a small one, and I see a fair number of men who’ve chosen the smaller size as well. As I wrote when reviewing the first Apple Watch Ultra, many people are self conscious about wearing a watch that’s “too big” for their wrist. That concern is often misguided — most people look just fine wearing a bigger watch — but it doesn’t change the fact that people are very reluctant to buy a watch that they fear looks too big on them.

Worth noting that three years ago Kuo, along with Mark Gurman, was completely wrong about the design of the Series 7 watches, with both of them claiming it would have flat sides like the iPhone 12. But it’s also worth noting that back in August, Gurman reported that year’s watches would introduce a new band system that breaks compatibility with existing bands, which, notably, have remained unchanged since Series 0 in 2015. (In fact I’m wearing my space black link bracelet from my own Series 0 on my Series 7 today.)

Apple Discontinuing Apple Pay Later, Just 8 Months After Rolling Out in the U.S. 

Apple, in a statement to 9to5Mac:

Starting later this year, users across the globe will be able to access installment loans offered through credit and debit cards, as well as lenders, when checking out with Apple Pay. With the introduction of this new global installment loan offering, we will no longer offer Apple Pay Later in the U.S. Our focus continues to be on providing our users with access to easy, secure and private payment options with Apple Pay, and this solution will enable us to bring flexible payments to more users, in more places across the globe, in collaboration with Apple Pay enabled banks and lenders.

This always seemed like a weird offering from Apple. Apple Pay Later did not charge interest, which is great, but it still seemed contrary to the spirit of helping people develop good financial habits. A no-interest loan doesn’t change the fact that in general, you should only buy things you can afford to pay for in full now.

Financial Times Reports EC to Charge Apple With Non-Compliance Under DMA for CTF 

Javier Espinoza reporting from Brussels, and Michael Acton from San Francisco, for the Financial Times (archive link in case your FT subscription isn’t working):

The European Commission has determined that the iPhone maker is not complying with obligations to allow app developers to “steer” users to offers outside its App Store without imposing fees on them, according to three people with close knowledge of its investigation. The charges would be the first brought against a tech company under the Digital Markets Act, landmark legislation designed to force powerful “online gatekeepers” to open up their businesses to competition in the EU.

The commission, the EU’s executive arm, said in March it was investigating Apple, as well as Alphabet and Meta, under powers granted by the DMA. An announcement over the charges against Apple was expected in the coming weeks, said two people with knowledge of the case. [...]

If found to be breaking the DMA, Apple faces daily penalties for non-compliance of up to 5 per cent of its average daily worldwide turnover, which is currently just over $1bn.

The EC leaks everything to the Financial Times. Reuters points out that EU antitrust chief Margrethe Vestager leaves office in November. Makes me wonder if there’s a clock-running-out aspect to this. Does the incoming regime share her politics regarding US tech companies?

Forget about trying to figure out what the EC wants from reading the DMA. It doesn’t say. I suspect they want Apple to completely forgo monetization of its IP on iOS — to allow the distribution of iOS apps without any charge or fee whatsoever other than the $99 annual developer program fee. I’m not sure, at all, how Apple is going to respond, but I do not think the EC is going to get that. I also don’t think they’re ever going to collect any significant fines from Apple.

My basic theory is that what the EC has wanted all along is to force Apple not merely to open up iOS to other methods of distribution, but to force Apple to allow apps to be distributed through those non-App-Store channels free of charge. But they don’t want to come out and say, flatly, that they seek to forbid Apple from monetizing its IP from all developers on the platform, because that’s so radically anti-capitalist. So instead they wrote the DMA to forbid the way Apple had, heretofore, mandated its cut of App Store revenue, and I suspect they somehow thought that if they banned the current rules — all apps must go through the App Store, all apps must use Apple’s App Store payment processing — then Apple would be forced to allow free-of-charge distribution through other channels and other payment processing. They didn’t foresee the Core Technology Fee as a route to collect a cut from any and all popular applications distributed by large commercial developers. So they’re just going to say that’s not allowed either, even though it seemingly doesn’t violate anything in the DMA. I don’t know where this winds up because I don’t see Apple conceding this point.


My thanks to DetailsPro for sponsoring last week at DF. DetailsPro brings SwiftUI to designers working on Apple platforms. (What a perfect week for a developer/designer tool to sponsor DF.) Without writing a line of code, you can bring your next idea to life in SwiftUI right from your iPhone. Design apps for the iPhone, on your iPhone. Or use it on your iPad or Mac.

DetailsPro has an easy, approachable interface, built-in templates, and a community of designers who share files. It is very easily to get started — you’ll be up and running in minutes. Designs in DetailsPro are 1:1 SwiftUI, so you can export to Xcode at any time. Intuitive features like side-by-side Dark Mode preview and Repeating Elements — a huge timesaver — use the smarts of SwiftUI to make the process enjoyable.

DetailsPro is available for iPhone, iPad, Mac, and Vision Pro. DetailsPro is so good and so fun it makes me want to come up with ideas for new apps just to have an excuse to use it. Download it now and give it a try — it’s free forever up to five files, and very reasonably priced for a professional tool.

Pixar’s ‘Inside Out 2’ Heads for Historic $140–$150M Box Office Opening 

Pamela McClintock, The Hollywood Reporter:

Pixar’s tentpole earned a massive $62 million on Friday, well ahead of expectations and putting the movie on course to open in the $140 million to $150 million range domestically over Father’s Day weekend, one of the top three starts ever for an animated film and the second-best for Pixar. Rival studios believe it could climb as high as $155 million to $160 million, but Disney is being more circumspect. Friday’s haul includes a huge $13 million in Thursday previews.

Great news for a great studio that needed a hit. Maybe we should stop griping about Pixar making sequels and just encourage them to make great films, original or not.

Japan Enacts Law to Mandate Third-Party App Stores, and You’ll Never Guess Which Devices Aren’t Included 

Kyodo News:

Japan’s parliament enacted Wednesday a law to promote competition in smartphone app stores by restricting tech giants Apple Inc. and Google LLC from limiting third-party companies from selling and operating apps on their platforms.

The law will prohibit the providers of Apple’s iOS and Google’s Android smartphone operating systems, app stores and payment platforms from preventing the sale of apps and services that directly compete with the native platforms’ own.

Laws like this are protectionist attacks that specifically target two U.S. companies — Apple and Google. The United States should treat this as a trade war, and reciprocate by passing legislation mandating third-party game stores and payments on game consoles from Sony and Nintendo. See how they like it. It’s patently hypocritical that Japan’s law targets only phones; this law wouldn’t exist if Sony were a player in phones and mobile platforms.

Violations of the new law will bring a penalty of 20 percent of the domestic revenue of the service found to have breached the rules. The fine can increase to 30 percent if the companies do not cease the anticompetitive practices.

Unlike the EU, which believes it can assess fines comprising a hefty percentage of companies’ worldwide revenue (which, in the case of the DMA, I doubt they’ll ever collect), Japan, quite reasonably, only assesses fines on companies’ Japanese revenue.

Ken Kocienda Left Humane 

Ken Kocienda, who was Humane’s head of product engineering, has left the company. his Twitter bio and LinkedIn profile both state “working on something new”:

Working on something new. Past: Humane, 15 years at , inventor of iPhone autocorrect, author of “Creative Selection”

Kocienda’s book Creative Selection is one of the best insider descriptions of working at Apple ever written, and two years ago he was a splendid guest on The Talk Show.

Never Change, Samsung, Never Change 

This listing for a vintage Samsung VR400G VHS VCR doesn’t state what year it came out, but it’s a pretty safe bet it was after May 6, 1998.

The Talk Show Live From WWDC 2024

Recorded in front of a live (and lively) audience at The California Theatre in San Jose Tuesday evening, special guests John Giannandrea, Craig Federighi, and Greg Joswiak join me to discuss Apple’s announcements at WWDC 2024.

3D video with spatial audio: Exclusively in Sandwich Vision’s Theater app on Vision Pro, available on the App Store. Just launch Theater and tap the “Watch Live Event” button. Tip: For this show, try choosing a seat in the front row center, not the middle of the theater.

Presenting sponsors:

  • iMazing 3 — The all-new version of the world’s best iPhone manager, for Mac and Windows. Save 20 percent.

  • Flexibits Premium — Fantastical. Scheduling. Cardhop. All your devices. Save 20 percent, both for new and existing subscribers.

  • Flighty — The first-class flight tracker for those on the move. Unmatched notifications and Live Activities. Delay predictions. Power through clarity.

As ever, I implore you to watch on the biggest screen you can (real, or virtual). We once again shot and mastered the video in 4K, and it looks and sounds terrific. All credit and thanks for that go to my friends at Sandwich, who are nothing short of a joy to work with.

The livestream of 3D video with spatial audio went almost perfectly, and the feedback from viewers who joined the stream has been unanimously positive. My sincere thanks and gratitude to SpatialGen for their remarkable work on that.

Not just shooting this event in 3D, but also streaming it live, was entirely the initiative of my dear friend, Mr. Sandwich himself, Adam Lisagor. I was asked a few times this week whether it was Apple who wanted to stream this live in 3D for viewing on Vision Pro. Nope. It was Adam who pitched me on the idea, only about eight weeks ago. I was like, “Well, sure, that sounds awesome, but how in the world would we do that? What camera could we use to shoot with? How would we stream it? What app would people be able to view it in? It’s a great idea but none of this seems possible.” Adam was like “I think we can do it.”

And, son of a bitch, they did it.

Once I started talking with Apple about arranging for guests, we did let them know our plans to shoot and livestream in 3D for viewing in Theater on Vision Pro — an app that, at the time, was in early beta. Maybe even alpha. Apple’s reaction echoed my own: sounds great but how?

The how is a long story and I get zero credit for any of it, but it involves a custom camera rig with two Lumix BGH1 cameras, each with Olympus 17mm PRO prime ƒ/1.2 lenses (35mm equivalent field of view with the BGH1’s Micro Four Thirds sensor). The lenses were about 3 inches apart — as close as possible — and microphones were placed throughout the theater to capture spatial audio.

The results exceeded my expectations, and I think everyone else’s as well.

To be clear, we shot the show two entirely different ways. The standard multi-camera cut on YouTube, embedded above, was shot exactly as we’ve done in previous years. That video turned out great too. The 3D version in the Theater app was shot from a single point of view, roughly simulating the perspective from a front-row center seat in front of the stage. It looks cool and sounds even better. If you have a Vision Pro or access to one, I highly encourage you to check it out.

We’re describing it as “3D video with spatial audio”, not “spatial video”, because that’s a more precise description of the effect. True spatial video would be more immersive — “Look left to see Joz, look right to see Gruber” — and the effect we achieved wasn’t quite like that. But what we did get is immersive, and very compelling. To the best of my knowledge, this was the first event livestreamed in 3D for viewing in VisionOS, and while my role was simply as the host of the show, that’s pretty damn cool. I’m just amazed at what Lisagor and his team at Sandwich (and SpatialGen) were able to pull off in a matter of weeks.

Reactions From Social Media

Matt Birchler: “I was only able to watch the first 40 minutes or so live, but it was really good, and really compelling to watch in immersive video. Can’t wait for the rest.”

Kevin Pfefferle: “Wow, The Talk Show Live exceeded all expectations. So many candid and insightful answers straight from Apple execs (plus a fair share of generic no-comment answers with a wink and a smile). Bravo! 👏🏻”

Kalani Helekunihi: “I watched the Talk Show Live on Vision Pro tonight. It was a really great experience, and felt close to sitting in the front row. Only real issue was bitrate / bandwidth related. If that can be solved, I imagine I’d prefer this to the real thing.”

Dave Marquand: “Really enjoyed the spatial livestream of The Talk Show Live. Taking the front row center seat in the Theater app definitely did feel a lot like sitting in row 1 at a stage performance. Oddly, the “most 3D” thing about the show was the glint off of your watch reflecting off of the floor of the stage in the foreground.”

Kalani Helekunihi, again: “Yep, there was even a moment where the glint caught the right eye camera. Was one of the most “accidentally real” experiences I’ve had with a headset, ever. Those sort of things just can’t be experienced with a TV screen. It all felt like sitting there, and looking into the stage.”

Michael Edlund: “I’ve never been lucky enough to attend @gruber’s The Talk Show live in person, and I was blown away by the 3D #live stream for #VisionPro that was literally like sitting in the first row at the edge of the stage.”

Kendall Gelner: “The 3D live stream was amazing! I just had one small hang in the first few seconds but went back in and it worked great the entire time, was just like sitting front row only in the most comfortable seat and I could easily refill a drink any time. TYSM!”

Nick Bodmer: “Watched The Talk Show Live on my Vision Pro last night — what an incredible experience! 🌟 Streaming it in spatial vision made me feel like I was right there in the front row. Sure, there were a few frame rate hiccups, but the immersive experience more than made up for it. Kudos to SpatialGen, Sandwich, and @gruber for pulling off a fantastic live event! 👏” 

Wayne Ma Confirms That Mark Gurman Scooped Him on the No-Cash Apple-OpenAI Deal 

Wayne Ma, reporting for The Information:

Neither Apple nor OpenAI are paying each other to integrate ChatGPT into the iPhone, according to a person with knowledge of the deal. Instead, OpenAI hopes greater exposure on iPhones will help it sell a paid version of ChatGPT, which costs around $20 a month for individuals. Apple would take its 30% cut of these subscriptions as is customary for in-app purchases.

Sometime in the future, Apple hopes to strike revenue-sharing agreements with AI partners in which it gets a cut of the revenue generated from integrating their chatbots with the iPhone, according to Bloomberg, which first reported details of the deal. OpenAI leaders have privately said the Apple arrangement could be worth billions of dollars to the startup if things go well, The Information previously reported.

I enjoy how Ma threw in a link to his own report from two weeks ago, but didn’t link to Gurman’s scoop — posted a full day earlier — at Bloomberg. Classy.

Gurman: Neither Apple Nor OpenAI Are Paying for Partnership 

Mark Gurman, writing at Bloomberg*:

Left unanswered on Monday: which company is paying the other as part of a tight collaboration that has potentially lasting monetary benefits for both. But, according to people briefed on the matter, the partnership isn’t expected to generate meaningful revenue for either party — at least at the outset.

The arrangement includes weaving ChatGPT, a digital assistant that responds in plain terms to information requests, into Apple’s Siri and new writing tools. Apple isn’t paying OpenAI as part of the partnership, said the people, who asked not to be identified because the deal terms are private. Instead, Apple believes pushing OpenAI’s brand and technology to hundreds of millions of its devices is of equal or greater value than monetary payments, these people said.

Meanwhile, Apple, thanks to OpenAI, gets the benefit of offering an advanced chatbot to consumers — potentially enticing users to spend more time on devices or even splash out on upgrades.

Apple getting this free of charge, in exchange only for the prestige of showing the ChatGPT logo and credit to users of Apple devices who engage the integration, is the Apple-iest negotiation in recent memory. My money says Eddy Cue, Steve Jobs’s favorite co-negotiator, made the deal. (I’d love to take Eddy Cue with me to the dealer when next I buy a car.)

During my show Tuesday night, I asked Federighi, Giannandrea, and Joswiak point blank, “So, who’s paying who in this deal?” (or something to that effect — transcript isn’t done yet), and got nothing more than smiles and shrugs in response. My read on the smiles is that they were smug happy smiles.

Ben Thompson and I recorded today’s episode of Dithering — the world’s favorite 15-minute podcast — yesterday before Gurman’s report dropped, but speculating, we came to the same conclusion, that it seemed likely neither company was paying the other. It makes obvious sense from Apple’s perspective. Not so obvious from OpenAI’s. But if OpenAI’s overriding goal is to cement itself as the leader in world-knowledge LLMs — to become to chatbots what Kleenex is to facial tissues — it makes sense to agree to this just to gain users, some of whom will upgrade to paid accounts. Google, on the other hand, probably wants to be paid by Apple to integrate Gemini. But now Apple can turn to Google — and Anthropic and Mistral and whoever else wants in on this iOS and MacOS integration, like the other default search engines in Safari — and Eddy Cue can tell them “My offer is this: nothing. Not even the $20,000 for the gaming license, which I would appreciate if you would put up personally.”

Back to Gurman:

ChatGPT will be offered for free on Apple’s products, but OpenAI and Apple could still make money by converting free users to paid accounts. OpenAI’s subscription plans start at $20 a month — a fee that covers extra features like the ability to analyze data and generate more types of images.

Today, if a user subscribes to OpenAI on an Apple device via the ChatGPT app, the process uses Apple’s payment platform, which traditionally gives the iPhone maker a cut.

Not traditionally. Always. Apple always takes a cut. One of my few regrets from my interview Tuesday night is not thinking to ask, on stage, whether iOS and Mac users will be able to upgrade from free to paid ChatGPT accounts right in Settings, where, I presume, the ChatGPT account sign-in will be. If so, Apple will surely be getting their 30/15 percent slice of that. And what’s the alternative? Sending users to OpenAI’s website, where Apple would get zilch? That doesn’t sound like Apple.

* Bloomberg, of course, is the publication that published “The Big Hack” in October 2018 — a sensational story alleging that data centers of Apple, Amazon, and dozens of other companies were compromised by China’s intelligence services. The story presented no confirmable evidence at all, was vehemently denied by all companies involved, has not been confirmed by a single other publication (despite much effort to do so), and has been largely discredited by one of Bloomberg’s own sources. By all appearances “The Big Hack” was complete bullshit. Yet Bloomberg has issued no correction or retraction, and their only ostensibly substantial follow-up contained not one shred of evidence to back up their allegations. Bloomberg seemingly hopes we’ll all just forget about it. I say we do not just forget about it. Everything they publish should be treated with skepticism until they retract “The Big Hack” or provide evidence that any of it was true.

‘The Shamans and the Chieftain’ 

Timothy Snyder:

The political theory of Trump’s coup attempt is that all that matters is the chieftain. He does not have to win an election, because the chieftain has the right to rule simply because he is the chieftain. Requiring Trump to win an election is thus a provocation. The claim that he should leave office when he loses an election justifies revenge. And of course retribution is Trump’s platform.

The legal theory of Trump’s coup attempt, made explicit in argument before the Supreme Court, is that the chieftain is immune to law. There is magic around the chieftain’s person, such that he need respond only to himself. The words “presidential immunity” are an incantation directed to people in black robes, summoning them to act as the chieftain’s shamans and confirm his magical status.

No issue has ever been more important in the history of the United States, and thus, in the history of democracy itself: Donald Trump lost the election in 2020 and tried, ham-fistedly, to spearhead a coup to remain in power by overthrowing the duly elected government of the nation. If he gets another chance by winning in November, the next coup won’t be as ham-fisted.

This Is Love 

What a photo. Reminds me of my own father, and the father I try to be myself. Impossible to imagine the former guy expressing such genuine love like this.