Re:Google's Project Nimbus is the future of evil
by deltree7 on Monday September 05, 2022 @05:35PM
(#32729775)
You don't have proof that AI is woefully unreliable, especially the ones used by the Isreal government
Re:Google's Project Nimbus is the future of evil
by z9znz on Monday September 05, 2022 @08:15PM
(#32731011)
This topic was about Google providing AI logistics. There is ample evidence of Google's AI of flagging false positives.
Re:Google's Project Nimbus is the future of evil
by deltree7 on Monday September 05, 2022 @10:31PM
(#32732030)
Once again you have zero access to the AI that was provided to Israel Government, just like you have zero access to Google's AlphaGo or AlphaFold or Imagen to make a judgement
Re:Google's Project Nimbus is the future of evil
by jjulius on Tuesday September 06, 2022 @08:48AM
(#32736381)
... as do you. Yet you seem convinced - dead certain - that the author is wrong, to the point that you're a bit belligerent. What happened to that "complex, nuanced" discussion that you referred to? Step back and breathe before posting.
If by “net benefits” you mean “helps in many cases but sacrifices a few people due to false positives”, then I see your point.
But you never really said why the author was wrong. You just expanded your critique of the author himself.
The author has valid concerns, and one doesn’t need to be an AI researcher to know that content analysis AI is woefully unreliable thus far. We see examples nearly every day here on HN where these systems (often Google’s or YouTube’s) have made wrong judgements that defy explanation.