AI does math. Should you get a PhD?

Where else are new ideas going to originate? In the past, you might have had Bell Labs, funded by a monopoly, that did more than make switchboards more efficient, but a society of highly efficient people with no imagination or creativity, manipulated in what to buy and think, is likely the goal of the technocrats who have their Bond villain goals and destructive and controlling means, disregard for others, who live in castles, surrounded by guards, and with similar looking surgically altered wives.

I suppose you can call that imagination, but of a dead race. (See "Return of the Archons" in the original Star Trek.)

Just look around you. Methinks that the technocrats regard those Bond movies as tales of entrepreneurs who were thwarted by a government official.

So tell me where are new, imaginative, inspiring and socially beneficial ideas going to be nurtured?

Perhaps get a PhD in something that (heavily) uses math, such as Physics or some aspects of Chemistry? Similar to CS, it might be better professionally to work in areas where CS is a tool rather than the product. If you can make math (or code) appear, than these associated areas should advance a bit more quickly.

This is good advice. Universities as a place of teaching and research are going to disappear.

People do humanities PhDs without much expectation of a job using it afterwards.