As for AI, look, I'm chemist, not a computer scientist Jim! Yes that was just my opinion. That and $4 will get you a latte at starbucks. However, in my defense, just look at the materials costs: All of AI's material needs are computing power. That's something we've got a lot of (the question is, how much is necessary etc). But ITER is a classic "big science" project, massive facilities, HUGE operational budget and power requirements, and a small army of technicians required for ONE experiment, let alone running the thing long-term as a viable power source.
Do you really think that $1 billion spent on these fusion projects will get us any closer to practical fusion power than the equivalent $1 billion towards AI? Or solar energy for that matter?
As for AI, my own opinion is that it will emerge organically, either in an uncontrolled fashion out on the internet or through some kind of evolutionary learning algorithm-based approach. I don't think that we'll really wind up "creating" AI, but rather creating conditions under which an AI could emerge. Therefore the timescale for that could be tomorrow, could be 50 years from now.
Also, check any news announcement of new supercomputers coming online, i.e. [news.cnet.com] or [publicaffairs.llnl.gov] . You'll almost always find some mention of "nuclear weapons simulations" somewhere in the description of the applications.
But by all means, continue to believe that our government has no further interest in nuclear weapons.
Fusion, like AI has been perpetually "just 10 years away" since the 1950's. Unlike AI however, there are serious concerns about whether we'll ever get there. So let's stop dumping limited taxpayer resources into a pipe dream.