Why AI is not the "jobs threat" everyone says

This is Lex Fridman, an AI expert from MIT, breaking down what the “improvements” in AI have actually been over the last 30 years:

As he points out through the writings of another expert, Rich Sutton, most of it is brute force and reliant on improvements in computation, not algorithms.

AI is not getting much smarter, just broader and more efficient. With the way academia is set up, where you need to find incremental improvements you can publish, this trend isn’t likely to change.