I think you're right on the AI, and I think Google agrees. https://ai.google/

Problem is, while there's certainly money in applications of AI (the ad auction is one of them), it's unclear that "AI" as a discipline is a money-factory like ads is. Google's looking for a money factory to replace ads because they're worried about that revenue stream being fragile (for the reasons ocdtrekkie highlighted).

Fight climate change? How?

"The smart people" provided a recipe for preventing/fixing climate change several decades ago. The politicians have never gone along, and now it's pretty much too late. It's no use hoping for a magic technical solution -- the one that would have worked was "Tax externalities and reduce CO2 emission".

"the one that would have worked was "Tax externalities and reduce CO2 emission"."

Why?

Internalizing externalities with a tax only works if the actual correct amount of those externalities equals or exceeds the amount that would reduce demand for fuel enough to stop climate change.

This is a huge assumption that I never see anyone even acknowledge needs to be demonstrated.

I think in fact they are far less, but my point is that debating that would be a red herring, since people seem to talk of externalities while never even believing in the concept, and that tension needs to be resolved.

The externalities are a number that in principle has a correct value, and the increase in price to change behavior sufficiently to stop warming is another number that independently has a certain value. But you have to have a logical reason for thinking they match if you think internalizing the costs is a solution.

Mother Earth has a magical technical solution. It’s called a “hard reset” so hopefully we can come up with something better.

I am smart and I find AI as it is currently practiced utterly boring (always did, ever since my first course on neural networks back in university). I think I am not alone here.

Have you thought of potential ways to use AI to help some problem domain you’re interested in? Like maths and other sciences it’s important to have an interesting application that you can get behind. It’s just a tool after all.