People thought the internet would ruin peoples minds, before that it was writing on computers instead of paper, and yet before that it was something else. We need to stop fearmongering everytime a breakthough technology changes things and start debating the actual issues calmly intead. Hidden figures is a great example of a movie where the issue is presented in a past time.
I'm sorry... is there actually a contingent who thinks the internet *didn't* ruin many/most people's minds? Look around. The average IQ didn't go *up* in the past 30-40 years. For every TED talk or similarly informative video (they used to be, anyway...), there are a thousand videos of mindless drivel, each of which has orders of magnitude more views than their informative counterparts. The same can be said of TV, in large part. It is absolutely the message not the medium that is the core problem - I believe TV/movies/video games/the internet/AI are equally capable of stimulating the mind as any book or newspaper, and can say with confidence I learned more about the Renaissance from the Assassin's Creed series than I did in history class (even after backing out all the fictional bits... in both sources...) - but throw in a profit motive that drives that message to the lowest common denominator, disseminated to the widest possible audience, and the inevitable result is to dumb things down further and further over time in a vicious, downward spiral. Idiocracy may not have been the most well-crafted of movies, but it was certainly prescient.
AI, like the internet and computers in general that came before, is a versatile tool. It can be used to do a great many things. But, like the internet and computers in general that came before, most people will use it to find new and innovative ways to *avoid* thinking, rather than new ways of thinking for themselves. In the long run, that's probably a good thing, since most of us can't be trusted to do that anyway. In the short run, though, it's going to get worse before it gets better (I'm of a similar mind regarding self-driving cars - they'll do better than humans before long, since most people aren't great drivers anyway, but in the meantime, they make me nervous).
That, and we now live in an age where "pics or it didn't happen" or even "video or it didn't happen" will be meaningless. To the extent we ever had a shared objective reality (which was already debatable), generative AI renders that impossible. We already had extremely skewed, partisan news sources, but at least that was just completely different takes on the same objective events. Now there won't be any way to tell what really happened, since both sides will be able to point to clear, unambiguous video of everything happening exactly as they claimed, regardless of what *actually* happened. And before long, there won't be any way to the difference, at all, because the technology is constantly improving. Even well-intentioned, ethical journalists won't be able to tell fact from fiction with any certainty, unless they were the reporter on site when it happened. And just like it's always been (at least since news went from a once-daily paper to a 24-hour news cycle), the ones that stop to make sure they have independent verification of the facts before they publish, will lose out to those that post the story as soon as they can.
...so I say, let's get out of it whatever we can, while we can still enjoy it. Generate whoever, doing whatever, however the viewer wants to see it. Aside from the environmental impact (which ship has long since sailed), what harm does it do to anyone else? The same hardware that tech companies are spending billions on today, will cost a fraction of that in a few years, and a smaller fraction a few more years after that, etc., all while training methods get more and more efficient to reduce the need for as many compute cycles in the first place - particularly once the initial AI bubble bursts, and the ones who survive it already have sufficient hardware to do what needs doing in the aftermath, such that demand levels off and with it the cost of related hardware for retail consumers (or smaller, niche providers, as the case may be). Once it gets to the point that it is truly photorealistic and indistinguishable from reality, it will become impossible to go after those harming real children to obtain real photos, so they will have to ban both (can't really blame them, there). But by then, I'm not terribly worried about there still being a fully functioning government to enforce such laws, except insofar as they can use it as an excuse to go after anyone speaking out against whatever government is left (OK, that last part may have been fearmongering a bit... but only a bit).
