[Yuki Maru Hoshi]

What I'm most concerned with is that we are looking to be paying a massive "AI tax" in the form of strained energy grids, thus even higher energy prices. All the build out that is coming with this technology is also creating all kinds of shortages in tech and nuisances for local communities in the short term because investors are currently doing a circle jerk. The idea is to print as much money as one can before the bubble actually bursts. These things may never actually be fully profitable. Yet these facilities require so much water as well. That means higher utility bills for the average person while they get tax credits to spend building out further.

Chips are already a precious commodity and they're needed in everything we own, so if the shortages continue then we eventually see a surge in everyday things. It'd take longer in things like appliances because they do tend to keep a lot of parts on hand, supposedly. But eventually it could impact those things as they need certain chips, even in simpler designs.

Re: AI vs Creatives;
The ethical issues being created by AI in the creative industry are the tip of the iceberg, sadly.

I have AI models installed on my computer and I can run them, but they're just toys. I create my own works and I don't bother with generative this and that except for personal amusement. I like that I can say how something was made, that it will work as designed for different users and not have issues down the line with low resolution output (poor code, node placement, etc), which is generally found with most AI slop ... so there is that.

There is a use case for AI creatively, especially for the preservation of older works. I took very clean rasters I had made on very old hardware more than 25 years ago, but of course they were at a lower resolution. It was not something I could rerender because for some the original files are long gone. I was able to blow them up to preserve them just about near pristine without any artifacting which was probably helped by the original being 3D renders so easy to interpret. So I consider those to be archival quality. I was able to get the software to run natively on modern hardware and rerender some of the other pieces and there's very few differences between them... and I'd say it's just creative differences in how some of the edges and textures were handled and even then, the differences were very minimal. But that AI algorithm exists already, doesn't require an online connection and so doesn't require entirely new facilities built to run it... so I'm torn on how far we take AI and what limits we should place. I think some AI is inherently a public good... but right now, it's much like the crypto market boom.