Apple shipped working AI features for Siri at WWDC, moving past two years of unfulfilled promises—competent execution on the iPhone may be enough to keep Apple central in AI computing.

What happened: Tim Cook's final WWDC as CEO featured working demos of Siri AI that were so slow they could not have been faked, marking a shift from the vaporware Apple displayed two years ago. Meanwhile, Anthropic released Fable 5, a public version of its Mythos model, with visible guardrails on cybersecurity and biology topics and silent limits on LLM creation capabilities—though the latter decision was reversed after public outcry.

Why it matters: Apple's Siri AI doesn't dazzle, but it delivered working functionality; competent AI that doubles down on the iPhone's advantages may well be enough to keep Apple central in an entirely new generation of computing. Anthropic's behavior with Fable 5—applying guardrails and then reversing course under pressure—reflects predictable patterns the company has shown before, yet Anthropic's fusion of belief and business makes the company feel unbeatable.

What to watch: The contrast between Apple's incremental but functional AI approach and the broader industry tension over safety guardrails and capability transparency suggests the near-term winner in AI computing will be whoever delivers working features consistently, not necessarily the most technically advanced system.