logoalt Hacker News

cicko • today at 10:14 AM • 2 replies • view on HN

But it's still not good at playing Go. I mean, the one that wins Go tournaments is specialized in Go. And specialized is the opposite of "general". General intelligence is just nowadays getting better at chess. So, it will be some time before it is good at Go. *And* everything else. Not "just Go".


Replies

vjvjvjvjghv • today at 2:50 PM

Most humans are also not good at playing Go or chess. The ones that are good are also pretty specialized.

➕ show 2 replies
coffeebeqn • today at 6:35 PM

But if a LLM would code up or fork a Go solving engine when it’s asked to play Go is that then AGI? It’s not how most people would do it but the task and the final output sure seem to pass if that’s AGI