Siri is good now?? | The Vergecast
TL;DR
After years of dysfunction, Apple has rebuilt Siri from the ground up with a working on-device content index and Gemini-powered natural language processing, potentially making the built-in assistant good enough to disrupt the entire consumer AI app market.
🗣️ Siri's Renaissance 3 insights
Rebuilt content index
Apple reconstructed its database architecture from scratch after nearly 20 years, finally enabling functional search across messages, photos, and mail that actually returns relevant results.
Gemini integration
Integration with Google's Gemini provides the speech-to-text and natural language processing capabilities that fix Siri's historical "garbage in, garbage out" input problems.
Cross-app intelligence
Siri can now perform complex cross-app queries like finding dinner recommendations based on flight terminal information extracted from emails.
⚙️ Technical Architecture 3 insights
Hybrid cloud processing
Complex queries route to Private Cloud Compute using Nvidia processors for interpretation while the actual data index remains on-device, though exact query routing remains unclear.
Connectivity requirements
The system requires internet access for natural language understanding, creating potential functionality gaps during flights or in low-connectivity environments.
Database foundation
Mike Rockwell confirmed Apple literally rebuilt the database from the ground up, representing a fundamental architectural fix rather than just an AI layer addition.
💥 Market Disruption 3 insights
Sherlocking free ChatGPT
By matching free-tier AI capabilities while offering exclusive access to personal data like iMessages and Photos, Apple threatens to make standalone AI apps redundant for average users.
Platform advantage
Apple's OS control allows them to force developers to expose app intents, potentially winning the race to become the primary AI platform before OpenAI can build a competitive app ecosystem.
Competitive pressure
OpenAI's pivot toward enterprise and Microsoft's hardware experiments suggest AI companies recognize they must either disrupt the phone itself or abandon the consumer market to Apple.
Bottom Line
Apple has transformed Siri from a dysfunctional voice assistant into a capable AI platform by fixing its foundational data architecture, potentially rendering third-party AI apps unnecessary for hundreds of millions of iPhone users.
More from The Verge
View all
YouTube is taking over Hollywood | The Vergecast
YouTube creators are bypassing traditional Hollywood gatekeeping to achieve box office success with films like "Back Rooms" and "Iron Lung," signaling an industry shift toward hybrid creator models where talent maintains multi-platform audiences rather than signing exclusive studio contracts.
Your biggest questions from Apple's WWDC | The Vergecast
Apple's WWDC 2026 keynote abandoned its usual theatrical polish for unusually honest, slow-paced live demos designed to prove Apple Intelligence actually works, while signaling an architectural shift toward horizontal AI services that transcend traditional OS boundaries and a return to rigorous design discipline.
How Steve Jobs became Steve Jobs | The Vergecast
Author Jeff Kaine discusses his book 'Steve Jobs in Exile,' arguing that Jobs' ousting from Apple in 1985 and subsequent struggles with NeXT and Pixar were essential crucibles that transformed him from an 'immature agent of chaos' into the disciplined leader capable of orchestrating Apple's historic turnaround.
This is your laptop... on AI | The Vergecast
The hosts discuss the shift toward 'agentic AI' at recent developer conferences, highlighting Google's Gemini Spark as a powerful example that leverages deep personal data to perform complex tasks like trip planning, while raising profound questions about whether the utility of such personalized AI justifies the extensive privacy trade-offs required to make it work.