3 ms·
Maybe it's not a great list, but as far as I know, adversarial attacks are still a major unsolved research problem. Why pretend otherwise? They belong on the li
by skybrian 9y ago
Maybe it's not a great list, but as far as I know, adversarial attacks are still a major unsolved research problem. Why pretend otherwise? They belong on the list, don't you think?
- chillee 9y agoI'm not trying to pretend otherwise. Of the problems on this list, I think adversarial examples and bias are real problems. > Some of them are illuminating parts of ML that are problems (copying society's bias, adversarial attacks), but most of the rest are simply things being overhyped/regular design errors (anything to do with Alexa/google home). It's a personal opinion that adversarial examples will always exist (they exist for pretty much all kinds of machine learning models), and that they might exist for human vision too.