AI Garbage

This NY Magazine Intelligencer article is making waves. It’s paywalled but since this is a post about using AI to cheat, that’s an archive link. The gist is that ChatGPT and other AI bots are making it hard for teachers to tell the difference between real work and AI-generated work. I mean, shit, I read it, but it’s a real hand-wringer in the way that many “sky is falling” pieces are, so I’m still a bit skeptical.

Here’s another, more pithy perspective from a set of professors masquerading as LBJ. They’re history profs, so Lord knows they read a lot of student essays:

The former President seems to think that AI can be detected and dealt with. (In the NY Mag article, it was reported that some schools don’t allow teachers to fail kids for using AI. Jesus wept.)

Most people can’t write worth a damn. Most of those people who are in college shouldn’t cheat, but as soon as they graduate, they’ll be using ChatGPT to write for them. My beef isn’t that they do it, but that what they churn out is slop. ChatGPT trains on the god damn wordiest fucking shit on the planet: stuff posted on the Internet, like recipes and astroturf blogs that have three or four paragraphs of throat-clearing before they get to the point.

If I were an author writing fiction or non-fiction, I wouldn’t fear being replaced by ChatGPT. But perhaps I’d use ChatGPT to write some of my book, that I would then edit. Is that the end of the world? As far as I’m concerned, it’s a tool that can be used or misused. Programmers I know use it and find it somewhat useful, but of course it all depends on knowing what to ask and knowing whether what’s been generated is decent or garbage.

Switching gears a little, here’s a take on AI-generated imagery from the excellent 404 media:

I thought that captured it pretty well. Trash pseudo-art to piss off the libs. I mean, if there were no libs to own, would MAGA even exist?

Kay was complaining about AI generated shit on Bluesky, but I don’t see much because that crap is an automatic unfollow for me. It’s just kind of dumb. I also unfollow people who put large images with quotes on them as if that makes them somehow more profound. If I wanted to browse Facebook, I’d go there. We have a lot of power to remove the AI we don’t like from our lives.

I’ve been in working technology all of my adult life, and I’ve lost count of all of the overblown technology silver bullets that were going to change everything. AI has some decent parts, but even Google’s AI efforts are mostly bad. I went out of my way to create a Google search that doesn’t use AI, and when I see it places where I can’t get rid of it (like on my phone), I give it as much attention as an Internet ad — exactly none. I’ve become “AI blind”. I think it will pass, just like the last dozen trends that were just somewhat useful technology instead of some kind of techno salvation.