The Bifurcation of Resources Between General Academic and Research

Kavia: A lot of this new trend is also a function of how humans organize and how we've been organizing ourselves. So I think it's in part due to that as well. It started with AlphaGo from DeepMind was maybe the first big example. And then it has been basically in industry labs that have released these large scale models, which are accelerating about as fast or slower than general deep learning.