you think debugging the system and having it explain decisions are the same? I think I understand where you're confused now ;)
-
-
Steve nobody saying that nothing is fixed. There’s consensus in the field that ML is not interpretable, a research program to try to do that, but not much progress and good arguments that it never well because connectionist neural networks are different then symbolic coding.
2 replies 0 retweets 2 likes -
Replying to @zeynep @marypcbuk and
Hence most programs are trying to do things that try to interpret outputs, uncover potential biases, tweak models to guess at spurious correlations... None of those are interpretation, and none of them rule out the issues I worry about.
0 replies 0 retweets 1 like -
This Tweet is unavailable.
-
It really isn’t controversial! But really not sure how to convince you. ML not interpretable isn’t a grand claim of mine. It’s a mundane part of the field.
1 reply 0 retweets 3 likes -
This Tweet is unavailable.
-
Pointing out problems so that we can try to address them, in my view, is less fear inducing in the long run compared to ignoring the problems and inevitably running into them.
0 replies 0 retweets 3 likes -
This Tweet is unavailable.
-
That application I highlighted is exactly the kind of inappropriate use we should avoid! To me it’s incomprehensible that a big babysitting referral company can even think about incorporating such ratings. There is no real study if such applications. At best, it’s snake oil!
2 replies 0 retweets 6 likes -
Interpretability of ML feels like a red herring. More central issue is that it would be impossible for humans to screen social media in this way. AI enables massively more surveillance...
2 replies 0 retweets 2 likes
Which was my initial point! Steve wanted to discuss interpretabilty again because of a passing comment in the thread.
-
-
This Tweet is unavailable.
-
@stevesi Your skepticism here has provided food for thought but I'm also surprised at how insistent you are in questioning a highly credentialed and widely published researcher. It's not reasonable to require her to relitigate her full argument to satisfy every new skeptic.2 replies 0 retweets 0 likes - Show replies
-
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.