@themattsimpson @St_Rev Yeah, just the experiments to show that the ImageNet winners are doing something uninteresting would be a year.
-
-
Replying to @Meaningness
@themattsimpson
@St_Rev I’m “pretty confident” after a few days of reading, but huge gap between that and knock-down proof.1 reply 0 retweets 0 likes -
Replying to @Meaningness
@themattsimpson
@St_Rev ML people *want to believe*, which is why they don’t do control experiments; so forcing a belief update is difficult1 reply 0 retweets 0 likes -
Replying to @Meaningness
@themattsimpson
@St_Rev I think I know how to systematically choose images that will cause them to give ridiculous results, which >1 reply 0 retweets 0 likes
@themattsimpson @St_Rev > would be fun, though, and visually arresting and easy for public to understand.
7:58 PM - 14 Dec 2014
0 replies
0 retweets
0 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.