Re:College Kids Doing What Twitter Won't
by rohanphadte on Friday November 03, 2017 @02:06PM
(#15621110)
Hey, I'm a creator of the tool written in the article!
In our model, we look at hundreds of features per account. Polarizing tweet content is not the only thing the model looks at. If an account is tweeting OC content, it's a clear indication that account is human.
Twitter's policy allows bots of their platform, and most are harmless! There are bots that tweet out the upcoming songs on a radio station. Unfortunately, a CAPTCHA would eliminate all bots, not the ones that spread fake news or claim compromised accounts.
Re:College Kids Doing What Twitter Won't
by torinmr on Friday November 03, 2017 @04:59PM
(#15622163)
It seems like there's an obvious solution to the problem of how to crack down on the types of bot your AI tries to detect, without removing legitimate boots:
Give users the ability to designate their account as a self-reported bot. Accounts designated this way would be identified as a bot in the Twitter UI, and would be exempt from captchas.
They started by identifying "left- or right-leaning" tweets. Then they decided that everything in-between were bots. Who decides what is left and what is right, and who decided that anything other than those extremes is "fake news"? There are ethical questions here.
It scares me that we now find it acceptable to classify accounts as "bots" or "Russian-backed" because of any content in the tweets, even if it's #LockHerUp or #BuildTheWall. If people really believe these things, shouldn't they be allowed to say them? What if tomorrow, something that you strongly believe in becomes labelled "fake news" or "propaganda", and your own accounts are shut down?
The methodology is flawed in the most fundamental way.
There are other technologies that can be used for detecting bots[0]. Twitter should use it.
[0] https://security.googleblog.com/2014/12/are-you-robot-introd...