[Imnimo]

It does seem to raise fair questions about either the utility of these tools, or adoption inertia. If not even OpenAI feels compelled to integrate this kind of model-check into their pipeline, what's that say about the business world at-large? Is it that it's too onerous to set up, is it that it's too hard to get only true-positive corrections, is it that it's too low value for the effort?

[JumpCrisscross]

> what's that say about the business world at-large?

Nothing. OpenAI is a terrible baseline to extrapolate anything from.