TechCrunch: Apple finally fixed Siri — and it feels anticlimactic

After years of delays, Apple shipped a substantially rebuilt Siri in the iOS 27 public beta in late July 2026. The assistant now understands personal context, pulls information from photos, emails, and calendar entries, answers general knowledge questions, and supports natural typed or spoken conversations — all things Apple had promised for several release cycles. A dedicated Siri app appeared for the first time.

Under the hood, the new Siri uses Google’s Gemini AI models combined with Apple’s proprietary Foundation Models. The result is an assistant that reliably handles tasks the old Siri regularly failed: playing the right song, finding a receipt, answering a question without redirecting to a web search.

The TechCrunch piece’s headline captures the central tension: Apple delivered what it promised, yet the response is muted. The author argues that in mid-2026, a functional AI assistant is no longer remarkable. The bar for AI products has moved. AI agents now handle multistep tasks, autonomous code execution, and long-horizon reasoning. A chatbot that does what it advertises lands closer to table stakes than to differentiation.

For product managers, the Siri story is a timing problem that applies well beyond Apple. Being late to a capability means competing on quality against incumbents who have already normalized that feature in users’ minds. Apple arrived with a genuinely capable assistant, but the market had already absorbed the novelty elsewhere. The differentiation they might have claimed in 2023 or 2024 was largely unavailable by August 2026.

The article also points to something worth noting for PMs navigating privacy-first product constraints: Apple’s on-device processing requirement shaped both what Siri could do and how long it took to ship. Those tradeoffs were real decisions with real competitive consequences. The choice to maintain privacy architecture at the expense of speed is not obviously wrong — it simply illustrates that technical constraints and product timing interact in ways that compound over time.