4 ms·
> Uh oh. Unseen Risk Fallacy a bit off topic, but couldn't identifying fallacies in arguments be automated? how much better would the discussion be if comment
by marmaduke 7y ago
> Uh oh. Unseen Risk Fallacy
a bit off topic, but couldn't identifying fallacies in arguments be automated? how much better would the discussion be if comments were labeled with likelihood of fallacy?
- TheGallopedHigh 7y agoThis would be a tall order for any NLP algo. I would hazard a guess and say no, not yet.
- contravariant 7y agoGiven that this thread is the only instance of the phrase "Unseen Risk Fallacy", I'm not too sure that it will work as well as you'd hope. Fallacies are subtle and even in the best of cases only point out holes in someone's argumentation, which doesn't tell you much about the conclusion. If anything automatic labelling of fallacies just risks running foul of the fallacy fallacy.