What's the difference between AI and AGI?

AI is a tool that's really good at one specific thing. AGI — artificial general intelligence — would be a system that can learn and handle any intellectual task a human can. We have AI today. We do not have AGI. Not even close. Think of it this way. The AI in your chess app can destroy any grandmaster on the planet. That same AI can't write a grocery list or recognize a cat. It's a one-trick pony. A brilliant pony, sure. But narrow. Almost all AI you've ever interacted with — Siri, Netflix recommendations, ChatGPT — is narrow AI. It operates within a box, even if the box is very large. AGI would be different. It could play chess, then switch to writing a poem, then learn to drive a car without being specifically programmed for any of it. It would transfer knowledge between domains the way you do. You use skills from cooking to figure out a chemistry experiment. Current AI can't do that. John Carmack, the legendary game programmer, recently announced he's focusing on AGI. That's notable because he's famously pragmatic. He's not chasing hype. He sees a genuine technical mountain to climb. Meanwhile, writer Ted Chiang argues convincingly that even our most advanced models show zero signs of actual consciousness. They're pattern matchers, not thinkers. Here's a useful mental model: narrow AI is a calculator for a specific subject. AGI would be a calculator for anything. We don't know how to build the second one yet. Some researchers think we're decades away. Others think it might require breakthroughs we haven't even imagined. The important thing for a beginner to understand is that when people argue about AI taking over the world, they're arguing about AGI — a hypothetical technology. The AI you're using today is powerful, but it's not on a path to wake up.