Let's Do It and Say We Did
I seem to know a lot of people who talk about hating AI, or being mad at the money people scaling it (as you should be), while not mentioning that they use it for their taxes, or for quick health lookups, or for help with spreadsheets and admin stuff surrounding their art.
Whenever I find out about their secret uses I cheer them on, tell them I’m not judging them (if that matters), and that I think anyone, especially artists, should do whatever reasonable thing they want to do to make the work that surrounds their art easier. Something’s lost but something’s gained in living every day.
You’re not gonna use the robot tax preparer to help prepare your taxes? Are you kidding me?
The conversation about what’s going wrong with AI policy is kinda corrupt when people pretend like they don’t use it. Just own up. Maybe some purists would punish you, but I think it makes you more honest, and the conversation more productive, if we lay it out on the table.
And yet I’m scared to talk about the ways I use LLMs. Mainly, as a bookkeeper, a programmer, and a personal librarian. Not as an oracle. Dear god, not as a writing “partner.” As fuzzy software for filing all my shit.
I think some of the techniques I’ve discovered for making those use cases work for me in a reliable, sane way are interesting, so I want to write about them. But I’m scared of the purists too :)