Apple shipped iOS 19.4 on Monday morning, and with it the update that the company's AI critics have been demanding since Apple Intelligence launched 16 months ago: an actually good version of Siri. The update, officially branded Apple Intelligence 2.0 internally though Apple has not used that numbering publicly, introduces what the company is calling Siri Reasoning Mode — a chain-of-thought processing system that lets Siri work through multi-step requests rather than pattern-matching to a single command.
The original Apple Intelligence Siri was fine at simple tasks. Set a timer. Send a message. Tell me the weather. Add this to my calendar. Anything beyond a single-intent command reliably produced either a wrong answer or a punt to a web search. The gap between what Apple promised at WWDC 2024 and what shipped was significant enough that multiple reviewers created taxonomies of failure modes. Apple Intelligence's rollout became the canonical example of a tech giant overpromising on AI features to compete with Microsoft's Copilot integration and Google's Gemini rollout — and then delivering something measurably behind both.
Continue reading to see the full article
