Right. There are two (very cursory) ways I think about these marquee DeepMind departures: these folks are (are, not were) amazing talents, but despite their capabilities they could not surmount some internal obstacles at Google that kept them from realizing their full potential. Or these folks were amazing talents for what they did in the past, but today those talents aren't able to get DeepMind to today's bleeding edge. Reality is it's probably a bit of both and there's just no nuance in the discourse so I share your fatigue.
And then I guess there's the third option which is money. These folks already have "enough" funds for generations to come, but there's a lot of financial fomo out there when you see all your peers becoming billionaires and you're a lowly centimillionaire.
I think another third option is that perhaps there aren't many people with the skills to run large teams for both research and product. I suspect that DeepMind / OpenAI / Anthropic require infrastructure and tooling that's significantly more sophisticated than what smaller independent research/product teams have experience with.
So in addition to large-team leadership skills, there may be some niche technology/systems that only a limited number people have experience with multiple 1-3 year cycles of designing, building, maintaining, and replacing these uniquely large infra systems.
Dupe: https://news.ycombinator.com/item?id=48601162
Hm first Shazeer and now Jumper, DeepMind getting hollowed out this week.
Since he had spent 9 years there, he cannot be considered a job jumper.
The Great Man Theory of computer science is a bit tiresome.
Right. There are two (very cursory) ways I think about these marquee DeepMind departures: these folks are (are, not were) amazing talents, but despite their capabilities they could not surmount some internal obstacles at Google that kept them from realizing their full potential. Or these folks were amazing talents for what they did in the past, but today those talents aren't able to get DeepMind to today's bleeding edge. Reality is it's probably a bit of both and there's just no nuance in the discourse so I share your fatigue.
And then I guess there's the third option which is money. These folks already have "enough" funds for generations to come, but there's a lot of financial fomo out there when you see all your peers becoming billionaires and you're a lowly centimillionaire.
I think another third option is that perhaps there aren't many people with the skills to run large teams for both research and product. I suspect that DeepMind / OpenAI / Anthropic require infrastructure and tooling that's significantly more sophisticated than what smaller independent research/product teams have experience with.
So in addition to large-team leadership skills, there may be some niche technology/systems that only a limited number people have experience with multiple 1-3 year cycles of designing, building, maintaining, and replacing these uniquely large infra systems.