7 Comments
User's avatar
bartb's avatar

Tacit knowledge .. journeyman ... Boston’s Route 128 .... Great post! From the outside, all the drama looks terrible. But this post is an excellent read for what's going on from the inside.

Brenton's avatar

Just a tiny observation, not a condemnation: There is a pattern of AI writing sticks out like a sore thumb to me: "This is not x. This is y." I see it all over the place. Especially on x tweets about geopolitics (calling out Mario Nawfal, and thousands of others). Again, no condemnation, but it just gives me the yucks when I see it.

Herbert Heyduck's avatar

I was writing in this style long before AI came along. LLMs imitate human writing styles.

Brenton's avatar

And ofc false positives are possible, as some people would naturally write this way. I'm sure it's considered a high quality writing technique and that's why it's incorporated heavily in the LLM training.

David Shapiro's avatar

Yeah, I'm sick of self-censoring "because AI writes like this"

The Mongoose's avatar

That is really cool. I had not thought of that. I especially like your comparison with the journeyman system.