Apparently, I'll never know the reason, because I don't have a subscription. It's a huge article so I can appreciate why it's paid-for, and I don't know HN's policy on that stuff, but it would be better if that had been mentioned up front.
Funny really, when I see Demis Hassabis' name, I always think of him as the videogame dev who was always trying to push what was possible in sims and AI, and now he's betting Google on world models "that can understand and simulate the real world, not just predict the next token." I wonder if he'll end up making something like The Oasis as an incidental side project?
Weird, there's no paywall when I go to the site. In any case, here's a paragraph that gives you the reason:
"Here’s my hypothesis, stated plainly: Google DeepMind CEO Demis Hassabis—a pioneer of AI as we understand it today—doesn’t think that automating AI research with coding agents (AI systems that can program better AI systems) is the correct approach to artificial general intelligence (AGI, the kind of AI that’s as good as humans at everything). Google DeepMind’s leadership sees OpenAI and Anthropic’s bet at best as an off-ramp, and at worst as a dead end."
Apparently, I'll never know the reason, because I don't have a subscription. It's a huge article so I can appreciate why it's paid-for, and I don't know HN's policy on that stuff, but it would be better if that had been mentioned up front.
Funny really, when I see Demis Hassabis' name, I always think of him as the videogame dev who was always trying to push what was possible in sims and AI, and now he's betting Google on world models "that can understand and simulate the real world, not just predict the next token." I wonder if he'll end up making something like The Oasis as an incidental side project?
Weird, there's no paywall when I go to the site. In any case, here's a paragraph that gives you the reason:
"Here’s my hypothesis, stated plainly: Google DeepMind CEO Demis Hassabis—a pioneer of AI as we understand it today—doesn’t think that automating AI research with coding agents (AI systems that can program better AI systems) is the correct approach to artificial general intelligence (AGI, the kind of AI that’s as good as humans at everything). Google DeepMind’s leadership sees OpenAI and Anthropic’s bet at best as an off-ramp, and at worst as a dead end."
https://news.ycombinator.com/newsguidelines.html
> please use the original title, unless it is misleading or linkbait