NYT is moving to using machine learning to moderate its (actually, pretty great) comment section. Big shift. Will they study it thoroughly?https://twitter.com/nytimes/status/874592901073956864 …
-
-
Replying to @zeynep
First, public editor eliminated in favor of reader feedback. Now, reader feedback will be machine moderated. Will any NYT folks read them?
5 replies 13 retweets 26 likes -
Replying to @zeynep
Machine learning is becoming our gatekeeper in so many areas without clear cut answers with which to anchor these alien intelligence filters
4 replies 19 retweets 43 likes -
Replying to @zeynep
It's not they don't work; it's that they are different and these differences are—yet— barely understood. Happening so fast since it's cheap.
2 replies 5 retweets 17 likes -
Replying to @zeynep
But once you plop it as the gatekeeper so fast, hard to study it properly because the gatekeeping introduces feedback as people adjust.
4 replies 3 retweets 13 likes -
Replying to @zeynep
Shifting gatekeeping to machine learning is not a small change even if your model seems to replicate your training data well.
3 replies 7 retweets 14 likes -
Replying to @zeynep
At best, we are putting in systems that work-to-rule—ask any labor organizer why that can be the most potent form of strike.
2 replies 3 retweets 7 likes -
Replying to @zeynep @thirdtruck
work-to-rule would be the equivalent of a rule-based 'expert system' ... machine learning is not that, it'll be seemingly arbritary at times
2 replies 0 retweets 0 likes
work to training data except when it goes weird
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.