AI has been a culturally radioactive PR disaster of truly epic proportions. Aside from whether or not it works, there are so many established catastrophically negative talking points - steals from creators, destroys the environment, is coming for your job - I'm not sure its reputation can be recovered.
if the internet did AI will as well. after all the internet was (and is) full of scams and p*dos, yet people use it the whole day, for everything.
Does it's reputation even matter? Everyone with money is pushing it, heavily. The government is even stepping in to stop any kind of punishment when it's factually shown that they are stealing water. The people will learn to tolerate it whether they like it or not, eventually.
The data center roll-out is weird, and far beyond anything that can be justified rationally. But this administration is aggressively pro-grift and anti-reality, so I would suspect that's as likely to be about some kind of corruption/grift/Ponzi as about real capability.
I try to distinguish between the actual tech, which spans light and dark, and the financial and economic engineering around it, which is definitely a darker shade of black.
Even describing the tech as light and dark seems a stretch, it exists solely due to the wholesale theft of virtually all copyrighted works in existence. Where would Ai be if they had to aquire every scrap of training data legally and with the informed consent (not legalese buried in esoteric terms of service documents) of everyone who created it?