If reparations were made to the people whose work was used, and the energy demands were solved, perhaps I'd feel more neutral about it. But right now, I'm appalled every time someone casually uses AI for something small.

I think that for specific uses like this, there's value. Or for differently abled folks who need assistance. But humans are very bad about limiting our use of new toys to NEEDS and not wants.

There have been some really neat things that have come out of it... new antibiotics, more efficient solar panel materials... some scientists are using it to analyze whale songs. My understanding is that these models are different than the ChatGPT/Gemini/CoPilot models that are being pushed on us.

And the studies coming out now about how using AI actually *inhibits* brain function—I intuitively felt that from the start, but it's gratifying to see the studies coming out now that prove it. I'm afraid for my kids. Schools are talking about deliberate AI training. Can't avoid it, but I'm opposed.

*LLMs (plural, oops!) I'm also an editor, and the writing that AI spits out seems fine on the surface, but once you dig down and actually analyze it, it is recursive and often incorrect. My colleagues have horror stories about the facts and sources that it makes up.

I don't use it because of the way companies ripped off artists and writers to train the early LLM (I've been anti-AI since Midjourney), and because of the ridiculous energy and water needs that AI servers require. No thank you!

I love this bit of your post: "The Framers of our Constitution... used these marks to do what their words alone could not. They used a brigade of periods, commas, and semicolons to inscribe a 'voice' into what would otherwise be a collection of terms and phrases subject to interpretation at will."