Why Amazon hates 'human-in-the-loop' AI governance

(theregister.com)

10 points | by ano-ther 17 hours ago ago

1 comments

  • cyanydeez 4 hours ago ago

    internally, they probably call it 'liabilty in the loop' and say things like, we know this feature might wrongly imprison people if we automate it, so we dont want any liability in the loop. ww can just call it a bug if something goes wrong.