(Image credit: Future)
Like many people, Siri was my first glimpse at what was possible with AI. Naturally, I imagined the future of artificial intelligence would look exactly like the voice assistants we already had in our pockets. We would ask questions, set reminders, send messages and perhaps get slightly more coherent answers than before. When ChatGPT first launched, I assumed it was the beginning of a smarter Siri. But then AI evolved while Siri stayed the same.That thought kept coming back to me during Apple's reveal of the new Siri AI at WWDC. To Apple's credit, the company has finally addressed the core complaints users have harbored for a decade. Siri can now understand personal context, parse what is on your screen, work across native apps, hold natural multi-turn conversations and take actions on your behalf. In many ways, this is the definitive version of Siri people have been asking for since the early days of voice assistants.The problem is that while Apple spent years rebuilding Siri, the rest of the AI industry was busy changing what people expect from AI assistants altogether. In other words, we've moved on.Apple finally fixed SiriFor years, Siri felt stuck in a loop. You could ask it to set a timer or check the weather, but the moment a request required a shred of contextual nuance, the experience completely fell apart. Meanwhile, platforms like OpenAI's ChatGPT, Google Gemini, and Anthropic's Claude advanced at a breakneck pace, making traditional voice assistants feel ancient by comparison.Apple has finally given its assistant the ability to weave together information across messages, emails, photos and calendar events. And while these aren't trivial improvements; they are arguably not unique — they are what Siri needs to survive. Apple deserves credit for achieving what many critics thought might never happen: making Siri feel competitive again.But that is exactly where the disconnect lies. While Apple was perfecting the assistant, AI quietly became something much bigger.AI stopped being about answering questions long ago
