didn't FB grant someone else $1B over 4 years or something for some AI lead role? Wouldn't be surprised if this guy is getting similar offers, which could explode even further with the stratospheric valuations of top AI companies.
I have to imagine that Anthropic is giving him an equity grant worth $1B or more in their upcoming IPO. And having him will increase their market cap more than that, if for no other reason than that people trust him as a judge of who is winning the AI race. So it's already worth it even before he does any work.
Yes, once benchmarks get saturated they get replaced by harder ones. You don’t see GSM8K, MMLU, or HellaSwag anymore because they’re essentially solved. It takes constant work to make benchmarks hard enough to show meaningful model performance differences but easy enough to score higher than the noise threshold.
It used an existing implementation, in theory this was mostly a porting task.
GPL-wise, I don't know how much is inspiration vs "based on" would this be, it'd be interesting to compare.
This looks like my Company peers, as long as there is any existing implementation they are pretty confident they can deliver, poor suckers that do the "no one has done it before" first pass don't get any recognition.
reply