Apple Finally Fixed Siri. It Just Took Long Enough for Everyone Else to Catch Up.
The new Siri is genuinely better. But after years of delays, Apple's AI assistant arrives in a world where its rivals can already write code, reason through problems, and complete complex tasks on your behalf.
Key points
- Apple has launched a major AI overhaul of Siri, its voice assistant, after years of delays.
- The updated Siri handles more complex requests and is more useful in everyday situations than its predecessor.
- Rival AI assistants, built on large language models like OpenAI's GPT-4o and Google's Gemini, already perform tasks well beyond basic question-and-answer.
- The launch leaves Apple playing catch-up in a market it arguably helped create when the original Siri debuted in 2011.
Siri is finally smart. That sentence would have been a headline two years ago. Today, it barely clears the bar.
Apple has shipped a sweeping AI upgrade to Siri, its voice assistant built into iPhones and Macs. First reported by TechCrunch AI, the overhaul transforms Siri from a frequently mocked button-presser into something closer to the helpful assistant Apple always advertised. Ask it to summarise an email, draft a message, or find a photo from last summer: it now does those things without frustrating detours. We've tracked this relaunch across [thirteen Siri stories since 13 July](/ story/apple-plans-paid-tiers-for-heavy-ai-users-as-siri-overhaul-nears), including Tim Cook's July pledge of paid tiers for heavy AI users.
So what took so long?
Apple spent years rebuilding Siri's underlying architecture, the software foundation that decides how the assistant understands and responds. Competitors didn't wait.
While Apple was rebuilding, a new class of AI tool emerged: large language models, the technology behind chatbots like ChatGPT and Claude. These models can write, reason, generate images, and produce working software code. OpenAI and Google turned those models into AI agents, software that carries out multi-step tasks on its own, booking travel or debugging code with minimal hand-holding.
By the time Apple's new Siri arrived, "capable AI assistant" had become the floor, not the ceiling.
Does the new Siri actually work?
For everyday tasks, yes. The upgrade is real and noticeable for ordinary iPhone users.
Siri now understands follow-up questions, remembers context within a conversation, and connects more deeply with apps on your device. Ask it to find an email from your dentist and then say "reply to that": it knows what "that" means. Simple? Yes. Exactly the kind of thing the old Siri failed at constantly.
| Feature | Old Siri | New Siri | Leading rivals (e.g. ChatGPT, Gemini) |
|---|---|---|---|
| Follow-up questions | Poor | Good | Strong |
| App integration | Limited | Improved | Varies |
| Code writing | No | No | Yes |
| Image generation | No | No | Yes |
| Multi-step task agents | No | Partial | Yes |
The gap the table shows is the real story. New Siri is better at the things most people use a voice assistant for. It can't write software or autonomously complete a chain of tasks the way GPT-4o or Gemini can.
What does this mean for ordinary users?
If you own an iPhone and found the old Siri more annoying than useful, this update's worth trying again. It's more reliable for quick, practical jobs: setting reminders, summarising notifications, drafting short replies.
For anything more ambitious, a dedicated app like ChatGPT or Google's Gemini remains the stronger choice for now. Apple hasn't closed the gap yet. It's just finally started running.
