Accidental Tech Podcast
Accidental Tech Podcast

581: A Different Way to Be Evil

* Pre-show: John’s two-year anniversary * Follow-up: * WWDC lottery results * John’s Affinity Designer struggles * Marco’s convert stroke → shape suggestion from the bootleg * Offset path using Contour Tool (via Vitor) * Create Compound (via Julian Kissman) * Inconsistencies noted by Christian Meyer

Featured Speakers

Marco Arment, Casey Liss, John Siracusa Host

Topics Discussed

Episode Summary

Executive Summary: The episode ranges across Apple ecosystem updates and broader tech-industry strategy: Chrome’s noisy network-change bug, vector-design workarounds in Affinity Designer, real-time OSs in cars, Microsoft’s antitrust-driven Teams split, Apple ID security ideas, iOS podcast transcription as a rare heavy cloud feature, Mac Studio/M3 Ultra rumors, and a long Vision Pro debate about content, usability, and Apple’s investment pace. The hosts repeatedly return to one theme: Apple and Microsoft both leverage platform power, but Apple may be underinvesting in Vision Pro while also inviting regulatory backlash.

Main Topics: Vision Pro content, usability, and ecosystem risk: A long discussion about Vision Pro’s current value, including the new spatial personas feature, the disappointing MLS immersive video, the lack of apps/content, and whether Apple is investing enough to justify the platform’s future. Microsoft Teams being split out of Office: The hosts analyze Microsoft’s decision to separate Teams globally after EU pressure, treating it as anti-competitive conduct that worked long enough to crush Slack before regulatory cleanup arrived. Apple podcast transcription and server-side AI: They examine Apple Podcasts’ transcription feature as a notable example of Apple doing expensive cloud-side computation at scale, possibly foreshadowing more AI-heavy services. Mac Studio / M3 Ultra rumor analysis: Discussion centers on whether the M3 Max lacks Ultra Fusion interconnects, what that would mean for future M3 Ultra and M3 Extreme chips, and whether Apple would build custom desktop-only silicon variants. Car OS architecture and real-time systems: They unpack how real-time OSs, Linux, Android Automotive, and hypervisors coexist in modern cars, and what that means for Apple CarPlay and instrument clusters. Chrome 124 beta and the 'Error: network changed' bug: A follow-up on a Chrome bug that spams network-change errors; the beta appears to reduce the issue, though the fix may be too narrow and brittle. Affinity Designer vector workflow tips: The show resolves a design-app issue with outside strokes and outlines, emphasizing Expand Stroke, Offset Path/Contour, and Create Compound as practical fixes.

Key Arguments: Apple Podcasts transcription is one of Apple’s first visibly heavy server-side compute efforts; it is more specialized than simply running Whisper and suggests a meaningful cloud-AI investment. Vision Pro has impressive moments, but Apple’s lack of apps and immersive content makes the product feel unsupported; the hosts see this as a potential strategic failure, not just a product-goods issue. Microsoft’s Teams split demonstrates a successful, if cynical, strategy: use monopoly leverage to win the market, then comply with regulation after competitors are weakened. Apple’s antitrust problems stem less from being 'mean' and more from failing to make strategically wise concessions that would have reduced long-term regulatory exposure. The MLS immersive video shows that immersive content must be edited differently from traditional video; rapid cuts and constant reframing work poorly in a head-tracked 180-degree environment. Vision Pro’s success may depend on Apple seeding the ecosystem with expensive first-party content and apps, because developers currently have little incentive to build for a tiny installed base. The M3 Max rumor is exciting because a different chip layout could imply a new desktop-focused Ultra/Extreme strategy, but the evidence is weak and based on ambiguous die-shot speculation. Real-time OSs in cars are layered with non-real-time systems; a crashed infotainment system staying alive does not prove real-time behavior, only separation of functions.

Data Points: ATP episode reference: Episode 476 - The anniversary segment references the episode where John announced he went independent. Announcement date: March 30, 2022 - Used to anchor the 'random anniversary corner' about John's independence announcement. WWDC lottery result: 3 for 3, no invitations - All three hosts report they did not receive WWDC developer lottery tickets. Estimated WWDC developer tickets: 2,250–2,500ish - The hosts estimate the lottery pool size based on the discussion. Chrome beta version: 124 beta - Used to test whether the 'Error: network changed' bug is fixed. Chrome beta run time without the bug: About 1.5 days - John reports not seeing the error during a prolonged Gmail test. Immersive video runtime: About 5 minutes - The MLS highlight reel is described as a short highlight package. Vision Pro launch date: February 2 - Referenced when discussing how little immersive content has arrived since launch. Vision Pro discussion timeframe: Just short of 2 months - Used to frame the lack of follow-on immersive content after launch. ATPs membership price: About $8/month - Mentioned in the membership ad read as the cost of the ad-free feed and bonus content. Apple TV+ launch investment example: Millions of dollars - Used as a comparison for the level of first-party investment Apple may need for Vision Pro content. iPhone / Vision Pro horizon: 3 years - Marco suggests evaluating Vision Pro’s viability on a three-year timeline.

Pivotal Quotes: "The more Apple tightens its grip, the more developers will slip through its fingers." — Marco / John (quoted in discussion): A summary line for the antitrust and ecosystem-control argument; used to frame Apple’s long-term strategic risk. "I think it is possible for this product to succeed, but I am not yet seeing any evidence that Apple will be able to do what it takes to make it succeed." — Marco: His bottom-line view on Vision Pro after discussing the lack of apps, content, and visible investment. "This is a big step, but I wouldn't necessarily extrapolate it to much else that Apple could be doing server-side in the near future for their other products." — Marco: On Apple Podcasts transcription: impressive, but not necessarily proof Apple will do similar cloud-heavy processing everywhere.

Implications: The episode suggests Apple may face real strategic risk if it fails to invest aggressively in Vision Pro content and if it keeps overreaching on platform control. Microsoft’s Teams move shows how antitrust pressure can reshape bundled products after the fact.

🔓 Sign Up for Unlimited Episode Search

About Accidental Tech Podcast

View all episodes from Accidental Tech Podcast