Accidental Tech Podcast
Accidental Tech Podcast

640: Put It in a Potato

* Pre-show: 🗣️ New ATP Member Special: ATP Top Four: Game Consoles 🗣️ * Follow-up: * Home Networking Corner 🤝🏻 Vision Pro Corner * How I use my AVP to retrofit Unifi Access Points in finished homes * SSDs and long-term storage was covered on Security Now #1023 (via Paul Chernoff) * SpinRite (Intel-o

Featured Speakers

Marco Arment, Casey Liss, John Siracusa Host

Topics Discussed

Episode Summary

Executive Summary: The episode moves through a mix of tech news and personal anecdotes: a clever Vision Pro home-networking demo, Fortnite’s return to the U.S. App Store, a Bloomberg report reinforcing Apple’s slow AI awakening, speculation about Apple’s upcoming AI APIs, and a sprawling discussion of OpenAI buying Jony Ive’s hardware startup. The show closes with Ask ATP on school trips, the bootleg cover art, photography gear tradeoffs, and CarPlay Ultra’s debut in an Aston Martin.

Main Topics: Vision Pro + home networking workaround: A Reddit video showed using Vision Pro USDZ models to map walls, ceilings, and access points in finished homes, sparking a discussion about spatial accuracy and the usefulness of Mac Virtual Display. Fortnite returns to the U.S. App Store: The hosts unpack the legal back-and-forth between Epic and Apple, explaining why Apple likely relented before a hearing and what this means for App Store enforcement and developer leverage. Apple Intelligence and internal Apple AI failures: A Bloomberg report and prior reporting are used to argue that Apple was slow to prioritize AI, with Tim Cook, Craig Federighi, and JG Giannandrea all portrayed as constrained by Apple’s culture and execution style. Apple’s rumored AI SDK for developers: The hosts react positively to rumors that Apple will expose its AI models via APIs at WWDC, stressing that broad, useful developer access could materially improve the platform if Apple avoids heavy restrictions. OpenAI acquires Jony Ive’s hardware company: The acquisition of IO sparks debate about whether a post-phone AI device could succeed, with speculation that the product may be a watch-like wearable and concern that today’s AI still falls short of the promise. Ask ATP: travel photography and school-trip memories: Listener questions prompt reflective stories about favorite school trips, plus practical advice on choosing lighter lenses, pancake primes, smaller-sensor cameras, and when to rent gear. CarPlay Ultra launches in an Aston Martin: The hosts discuss Apple’s next-gen CarPlay implementation, noting that it now ships in at least one car, but remains niche, highly integrated, and possibly less appealing than the original vision.

Key Arguments: Apple’s App Store fight with Epic appears to have been resolved pragmatically because Apple risked worse consequences if the judge got involved again. The Vision Pro demo shows how spatial computing can enable genuinely useful professional workflows, even if the underlying tech is slightly approximate. Apple’s AI delay is framed less as ignorance and more as a cultural mismatch: the company prefers knowing the end product before investing, which clashes with how LLM technology evolves. If Apple wants its platform to stay compelling, it should expose model capabilities to developers as APIs rather than keeping them private to first-party apps. A server-backed Apple AI API could be valuable, but only if priced and constrained carefully; otherwise Apple could create bad incentives or degrade the developer experience. OpenAI and Jony Ive may succeed where Humane and Rabbit failed because OpenAI controls the models and has stronger product gravity, but the core challenge is still making AI reliable enough for a screenless or low-screen device. CarPlay Ultra is technically impressive, but likely to remain a boutique feature because automakers must surrender a lot of UI real estate and design control. For travel photography, the best way to reduce weight is often to compromise on sensor size or focal flexibility rather than expecting full-frame glass to become magically light.

Data Points: Fortnite return timing: Returned to the U.S. App Store after Apple/Epic filed a joint notice resolving issues - The legal dispute reached a practical resolution after a court nudge and impending hearing Epic rewards bonus: 20% back - Epic offered 20% Epic Rewards for using its payment systems in Fortnite, Rocket League, Fall Guys, iOS, Android, PC, and web IO acquisition price: $6.5 billion - OpenAI’s all-stock purchase of Jony Ive’s hardware startup IO Humane acquisition comparison: $116 million - Used as a contrast to show how much larger IO’s valuation/exiting price was than Humane’s Reported team size: ~55 people - The hosts estimate IO as a small team relative to its acquisition price Member special count: somewhere in the 30s - The hosts estimate the number of ATP member specials available to supporters Apple AI milestone: June 9 - Bloomberg reports Apple expects to unveil the AI SDK at WWDC AWS/AI access model: free up to a limit, then paid - The hosts compare proposed Apple AI APIs to existing Apple services like CloudKit and Weather CarPlay Ultra shipping partner: Aston Martin - The first car maker discussed as shipping Apple’s next-gen CarPlay experience Typical lens advice: 35mm prime / pancake lens - Suggested as a lighter travel option for photographers wanting to reduce size and weight School trip duration: 4 weeks - Harrison’s residential camp mentioned at the start of Ask ATP

Pivotal Quotes: "If you want a really useful thing from AI, you have to actually make it do something useful and reliable." — Casey: Discussing the OpenAI/Jony Ive hardware gamble and the failure modes of screenless AI devices "It is incredibly clever. I am so mad that this never occurred to me." — Casey: Reacting to the Vision Pro home-networking video using USDZ objects to map interior spaces "If it turns out we win on appeal, we’ll kick them out again, but why don’t we just let them back in?" — Marco: Explaining Apple’s likely motivation for allowing Fortnite back into the App Store

Implications: Apple is under pressure to turn AI from a private feature into a platform capability, while OpenAI/Jony Ive are trying to invent the first truly useful AI-native device. Both efforts hinge on reliability, integration, and execution, not hype.

🔓 Sign Up for Unlimited Episode Search

About Accidental Tech Podcast

View all episodes from Accidental Tech Podcast