Siri works now. After roughly fifteen years of setting the wrong timers and shrugging at basic requests, Apple has shipped a new version of its assistant that’s actually competent, according to The Verge AI. On the latest episode of The Vergecast, hosts David Pierce and Nilay Patel walked through their early hands-on time with the revamped Siri AI and what it means for users, for the iPhone, and for the broader AI industry.
The headline isn’t that Apple built something futuristic. It’s that the company built something that works.
What actually happened
Apple rolled out a new Siri powered by its AI stack, and The Verge AI reports it handles most everyday tasks well enough that you’d trust it again. As Pierce and Patel put it, there’s “very little about Siri AI that feels bleeding edge or brand new, but it works. And that might change everything.”
That framing is the whole story. Siri spent a decade and a half stuck somewhere between mildly useful and genuinely broken. The bar was on the floor. Clearing it turns out to be a bigger deal than launching another flashy demo.
Why this matters
Here’s what stands out to me. The AI assistant race has been dominated by chatbots you have to go open: ChatGPT, Claude, Gemini. They’re powerful, but they live in an app you choose to launch. Siri lives on more than a billion iPhones by default, sitting one button-press away from every user.
An assistant that’s merely “good enough” but already in everyone’s pocket can reach more people than a brilliant one you have to seek out. That’s distribution beating raw capability. For years Apple couldn’t cash in that advantage because the product was bad. If that’s fixed, the math changes fast.
A few things to watch:
- Default beats best. Most people won’t install a third-party assistant. A working Siri becomes the AI most iPhone owners actually use, whether or not it’s the smartest one available.
- Pressure on the chatbot crowd. OpenAI, Google, and Anthropic now compete with something built into the hardware, not just the app store.
- The “good enough” benchmark. Reliability at common tasks may matter more to everyday users than benchmark scores or reasoning demos.
The bigger context
Apple has been the laggard in this cycle. While rivals shipped fast, Apple moved cautiously, and Siri became the punchline for every “AI is overhyped” argument. A version that quietly does the job resets that narrative. It also validates a slower, integration-first approach over racing to ship the newest model.
The Vergecast episode didn’t stop at Siri. The hosts also covered shifting winds in social networking, with Instagram, Bluesky, and YouTube all rolling out features meant to make their massive platforms feel smaller and more personal. Group chats have been eating into public posting for a while, and the team asked whether the era of the open Twitter-style feed is really winding down. The lightning round touched on the Trump Phone, Claude Fable, solar energy, and a deal for iPad users.
What to expect next
If you’re an iPhone user, the practical takeaway is simple: give Siri another shot. The assistant you wrote off years ago may now handle the basics you stopped asking it to do.
If you build or follow AI products, watch how a competent default assistant reshapes user behavior. The companies betting that people will leave their home screen to talk to a separate chatbot now have to reckon with one that’s already there and finally pulling its weight.
The real test comes once millions of people use it daily instead of two podcast hosts trying it early. Reliability at scale is where Siri broke before. But for the first time in a long time, Apple has an assistant worth testing again. Full breakdown and the hosts’ early impressions are in the original episode from The Verge AI.