“Correlation does not equal causation.”
#ArtificialIntelligence #AI #MachineLearninghttps://twitter.com/GaryMarcus/status/1170912860815396864 …
-
-
Replying to @AmyKotsenas @GaryMarcus
No one in AI/ML would ever say something so silly. Why bring it up? Causation may be established through randomized experiments, or by varying each hypothesized causal factor in turn while holding the rest fixed. 2nd order statistics may then be used to represent causal factors.
2 replies 0 retweets 4 likes -
Replying to @AlexVasilescu @AmyKotsenas
sure, but consider for example DeepMind’s much-celebrated Atari game system, which lacked anything like @yudeapearl ‘s explicit representations of causality and then (as
@dileeplearning showed) in consequence generalized very poorly to slightly altered situations.2 replies 0 retweets 3 likes -
Replying to @GaryMarcus @AlexVasilescu and
I think (correct me if I'm wrong) Marta Garnelo and I showed this before
@dileeplearning in our 2016 arXiv paper - https://arxiv.org/pdf/1609.05518.pdf … which was one reason@DeepMindAI hired us - to work on those issues4 replies 0 retweets 10 likes
pieter abbeel and ian goodfellow and others had another demo of this failure pattern, perhaps just a bit later: https://arxiv.org/abs/1702.02284v1 …
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.