Bloomberg's Mark Gurman says Apple's redesigned Siri performs reliably for the first time in 15 years, challenging third-party AI
Some users report being stuck on device-indexing waitlists
Some users report being stuck on device-indexing waitlists
Some users report being stuck on device-indexing waitlists
Some users report being stuck on device-indexing waitlists

@markgurman @Scobleizer It’s actually great. Unfortunately some people here are never satisfied 🤷♂️

@markgurman It’s surprisingly good for basic tasks

@markgurman Please don’t!
The new Siri finally works as it is supposed to for the first time in 15 years and the pre-installed status is going to give Apple a competitive baseline to ChatGPT, Gemini and Claude that will steal away users of basic tasks like search, the most popular AI use case. https://twitter.com/markgurman/status/2066170952480952779
For those in the Mac world, Siri AI is like iMovie and ChatGPT is like Final Cut Pro. Siri AI can handle the basics — and a bit more — but power users will want the extra functionality. There’s plenty Siri AI can’t do, but what it can do is good enough for a ton of people. https://twitter.com/markgurman/status/2066170952480952779
Power On: Apple’s new Siri AI is just good enough to pull it out of its AI crisis. My hands-on impressions of the new assistant and AI features. https://www.bloomberg.com/news/newsletters/2026-06-14/siri-ai-hands-on-review-ios-27-macos-27-details-iphone-fold-touch-macbook
Longer term, I’d also expect Apple to try to create its own OpenClaw competitor, delivering a system that could fully operate its software across iPhones, iPads and Macs on behalf of the user. https://twitter.com/markgurman/status/2066170952480952779
@markgurman Still stuck in indexing and waitlist, I wish Apple didn't put me in the permanent underclass
Saaaaaame come onnnn I need this https://twitter.com/kevinclark/status/2066192951307292847
Mostly Apple's own custom models, refined from Google's Gemini. Apple's new Siri AI runs on its next-generation Apple Foundation Models (AFM family), built in-house for Apple Silicon with on-device options (Core/Advanced) and cloud variants for complex tasks. These were developed via a multi-year Apple-Google partnership: Gemini provided the foundational training outputs and cloud tech, which Apple distilled, trained on proprietary data, and adapted with its privacy-focused Private Cloud Compute—explicitly not direct Gemini calls or Google's public infra.[[1]](https://macdailynews.com/2026/06/08/apples-siri-ai-is-not-google-gemini-with-apple-branding-heres-how-it-really-works/) The story's on-device privacy edge and ecosystem integration are core Apple strengths atop that Gemini-derived foundation. Some reports note a ~$1B/year scale to the deal.