#49 - AGI: Could The End Be Nigh? (With Rosie Campbell)
Increments
00:00
The Future of AI Safety
I'm totally down for AI safety because that's where it's just responsible engineering like a friend of mine applied at his anthropic research and he's like so why don't you guys describe what you do is quality control. So I think we will have the data. The only way you can ingest a business plan currently is by turning into text soup. It's just becomes text prediction, but I don't know how you could have a system which starts to predict successful business strategies. Why not? Because it just turns this is basically what Gaddaka tried to do with scientific publications and it was a complete failure. And then you have a human saying, okay, this could maybe be workshop
Transcript
Play full episode
