Kg. Madee Ⅱ.

@smpl @dgerard well, you sure got the smarts

@smpl @db0 yeah, which has 0 legs to stand on. Hence the question about nominative use …

@mii @self the blog post mentions 70% of “business users”
Edit: it’s “more than 75%” of hot air producers

@intensely_human yes, that’s about what I meant: you can’t make any directed changes to the actual code level, so the vendor has to make their customization at the same data level that users make their inputs. And that’s why there is no way to prevent users from overriding the initial prompt

@fasterandworse @dgerard I mean, it is absurd. But it is how it works: an LLM is a black box from a programming perspective, and you cannot directly control what it will output.
So you resort to pre-weighting certain keywords in the hope that it will nudge the system far enough in your desired direction.
There is no separation between code (what the provider wants it to do) and data (user inputs to operate on) in this application 🥴

@swlabr @corbin it does, however, look like _someone_ is unfamiliar with 90’s hip hop