GPT-3 is the first AI system that has obvious, immediate, transformative economic value.
I think the jury is still out on this one. It certainly seems powerful, it's doing interesting things, and it's better in many ways than any system that has come before. But there's a different between exciting demos and transformative economic value.
It's too soon to be sure, but to me, the most interesting question is whether any valuable startups will be built on top of GPT-3. Some leading indicators before that are whether useful products are built on GPT-3, and whether early-stage startups built on GPT-3 get seed investment. I'm not aware of any of these yet but maybe latitude.io counts as one.
I think the jury is still out on this one. It certainly seems powerful, it's doing interesting things, and it's better in many ways than any system that has come before. But there's a different between exciting demos and transformative economic value.
It's too soon to be sure, but to me, the most interesting question is whether any valuable startups will be built on top of GPT-3. Some leading indicators before that are whether useful products are built on GPT-3, and whether early-stage startups built on GPT-3 get seed investment. I'm not aware of any of these yet but maybe latitude.io counts as one.