On that note, Rebooting AI should be required reading for any doctor or medical student interested in work on applying deep learning and AI to medicine. There are many troubling issues with deep learning alone, especially for medical applications where reliability is paramount.https://twitter.com/GaryMarcus/status/1183077024207818753 …
-
-
This Tweet is unavailable.
-
Replying to @acganesh @GaryMarcus
One of the problems is when an algorithm gives a wrong result. The fact that that happens unpredictably with deep learning is a huge problem. In medicine we don’t demand testing to be perfect because it’s impossible anyways—but we do expect that we can anticipate unreliability.
2 replies 2 retweets 10 likes -
Well, humans aren't reliable either. I don't think it's a good argument. But humans have a deeper and shallower processing modes. Humans can use additional tools and form committees that are much more powerful than single humans. Standard stat algorithms are always shallow.
4 replies 0 retweets 5 likes
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.