I certainly don’t mean to imply that I work this way because it’s demonstrably better. It’s a habit, and a way of working with LLMs that I can get my head around. Many of my friends are going the harness route, generally with mixed results.
To some extent, I think my habit is related to how I like to work with my own mind. I’m not sure I can elaborate meaningfully on that right now, so I’ll leave it there. What I can offer for debate is that I think the harness direction of development would benefit from more thoughtful type-theoretic foundations. I don’t mean the trashy, just-trying-to-get-funded version that Jev is hand-waving at. I mean properly good type theory of the kind I know you know.
.- .-.. .-.. / ..-. --- --- - . .-. ... / .- .-. . / .-- .-. --- -. --. / ... --- -- . / .- .-. . / ..- ... . ..-. ..- .-.. FRIAM Applied Complexity Group listserv Fridays 9a-12p Friday St. Johns Cafe / Thursdays 9a-12p Zoom https://bit.ly/virtualfriam to (un)subscribe http://redfish.com/mailman/listinfo/friam_redfish.com FRIAM-COMIC http://friam-comic.blogspot.com/ archives: 5/2017 thru present https://redfish.com/pipermail/friam_redfish.com/ 1/2003 thru 6/2021 http://friam.383.s1.nabble.com/
