The Role of Scale in the Future of AI

In the era of large language models, we're seeing that same paradigm play out just in a slightly different form. You want to invest the human effort towards the parts that humans are especially good at and try decommos the problem into these pieces where machines can do what the machines aregreat at. And I'm kind of curious, the role scale could play in somehow what's the new kind of annotation that will inject a notion of truthfulness into these models? Totally.