> I think it's funny how much credit Hinton gets for AI. His contribution is pretty much just keeping some grad students on the problem.
Yes, clearly overrated in terms of credit. Doing foundational work in the field going back to the 70s which laid the groundwork and inspired the resurgence of neural networks in the late 80s. Being a solid community organiser throughout his career and keeping neural network research alive through the more formal statistical methods dominating for over a decade. Supervising and thus raising many others who themselves contributed greatly to the explosion of neural network utility we have seen since around 2010 until now. Should I carry on?
I think it is absolutely clear that Hinton has contributed plenty enough to get a massive amount of credit for where we are today. The kind of mentality at display here is akin to ahistoricity on the level of saying that Gordon Moore "just started a company" after Apple released the M1 under the delusion that there is not a direct lineage between what we have today and breakthroughs and efforts in the past. Believe it or not, but we stand on the shoulders of giants and cutting them some slack is not the same as downplaying the impact of people more active in the present day; that are gradually becoming future giants.
I think what the comment gets wrong is that even in the professor (PI) / grad student relationship, it's not defined as a purely managerial one. The one-on-one meetings between professor and graduate student are often about working out a new theory, even if the implementation and experimental work is left to the student. The amount of mind-share that goes on between is nontrivial, it is necessary for work that pushes the boundaries of human knowledge, and isn't at all like an executive asking for a report from an underling.
I think there's an important difference between not understanding and not doing the work yourself because you are in a managerial/advisor position. If you give Hinton a recent paper on LLM RLHF he will understand the nuances in it, he just delegates the actual work. If you give Emmett Shear or whoever such a thing, they almost certainly don't. For a deep tech company focused on research (not some consumer SaaS thing) I don't think you can be a good CEO if you don't even have an understanding of what you are building.
It's important but in tech world you'd get it as explicit "leadership" credit rather than a researcher, I think. Or maybe people would just call you a middle manager.
Btw, I think it's funny how much credit Hinton gets for AI. His contribution is pretty much just keeping some grad students on the problem.