What's the real difference between AI and AGI, and why does everyone keep talking about AGI now?

AI is a specialist. AGI would be a generalist — like a human. That's the simplest way to think about it. The AI you use every day, like ChatGPT or the recommendation system on Netflix, is incredibly good at one or a few specific tasks. It can write an email, generate an image, or suggest a movie. But that same AI can't suddenly drive your car or diagnose a medical condition. Its intelligence is narrow. AGI, which stands for Artificial General Intelligence, is a theoretical idea of a machine that can understand, learn, and apply its intelligence to any problem a human can. You could ask it to learn a new board game, then write a poem about the experience, and then help you plan your taxes — all in the same conversation. It would adapt fluidly. The reason you're hearing about it more now, especially with figures like John Carmack starting AGI-focused companies, is that the leaps in narrow AI have been so dramatic. People see a chatbot that can code and write sonnets and they think, 'We must be close.' We're not. Not really. A 2024 survey of AI researchers estimated there's only a 50% chance of high-level machine intelligence arriving by 2047. The gap between mimicking understanding from massive data and possessing genuine, flexible reasoning is still a canyon. The hype makes it feel closer than it is. A good way to spot the difference in the wild: if a system fails completely at a task slightly outside its training, it's narrow AI. An AGI would figure it out.