Tootfinder

@aiappsapi@mastodon.social I agree that per-query values have limited meaning given all the variability in the context of how they're used.
That said, although I don't know an empirical study of this, I think there are many contexts in which a cheaper model has a much lower footprint. In fact, a recent study out of Stanford et al. found re-reading context was one of the reasons agentic tasks blew up compute

This is also the reason I despise most requests for content warnings. I do not in fact want to use your schema for this, and people do not in fact agree on what’s triggering. I’m happy to accommodate specific people in some contexts, but for example generally on social media? Sorry, no. The unfollow button is over there.
And it’s not like I go posting spider images or grisly photos or anything. I talk about hard stuff though. usually a little academically, usually a little distantly. And I deeply believe we need more of that and not to bury it under ‘cw: trauma’ or whatever. That warning gives very little information to go on and presumes a whole lot about the schema of the world that does not hold. I won’t be subject to it.
I can exercise care for people in other ways.
And not everyone all the time gets personalized care from me. That’s an impossible demand and should be treated as such.