I don't think AGI is necessary to disrupt most industries. There is a lot of jobs that purely rely on a computer interface to do the job. As long as there is a push by VCs to fund automation, which there is, and companies that would prefer to automate, which there are, then it seems like it's inevitable we'll see more and more in this sense.

Plenty of machine learning models can do better than humans in some disciplines, that will only go up from here. You don't need AGI to do that either. Just a model that's better than human performance.