The College Kids Doing What Twitter Won't | Hacker News Reader

The College Kids Doing What Twitter Won't

wired.com

16 pointsrbanffy8 comments

Flagging feels harsh, hopefully that can be un-done.

This was the original post, but a larger conversation after one hour is on this dupe: https://news.ycombinator.com/item?id=15619533

TL;DR: publicly classifying Twitter accounts as bot-or-not: https://botcheck.me/

This is a great thing these kids are doing, but even if Twitter were to purchase their plugin and fully incorporate it, Twitter's problems won't ever go away.

If Twitter is defined by users who only want to share "little bits" of ideas, and only have the attention span to consume "small snippets" of information, it will always cater toward the simple minded.

Twitter will never be a good place to go to listen, to learn, or to have your ideas challenged in a real way.

Is it any surprise the bumper sticker of social networks has become such a toxic place?

Incase any Ai / tech minded people are reading, this is a relevant quote:

> As fall semester kicked up, RoBhat hand-picked 100 Twitter accounts with automated behavior to serve as “ground truth” data to train their model. They picked accounts with several red flags: ones that joined the site, say, a month prior but had tweeted 10,000 times, or ones that were followed by thousands of other suspected bots.

While I was skeptical they could train a model without actually knowing for sure which accounts are bots, their assumptions here seem to make sense. There is a certain "tweets per day" ratio that can't be easily maintained by a human.

BotRank -- the more suspected bots follow a given feed, the more likely that feed is itself a bot

Add to that sentiment analysis by topic. Real people don't tweet about Hillary 100% of the time, no matter how awful she is.

If there aren't at least 10% of tweets about food, sports, or TV shows, I think its probably fake.

> Real people don't tweet about Hillary 100% of the time, no matter how awful she is.

I disagree. Real people use pseudonyms as well. I have multiple twitter accounts and while I don't have an ax to grind against Hillary, I can imagine some of my friends in Texas would create and devote a twitter account solely for venting about something like this.

Of course, there is a difference between someone in Katy, Texas tweeting "lock her up" versus someone using a farm to make it look like there is a groundswell when there is none. However, is it wrong, immoral, or illegal? How do we define these things? We have to be really careful because we might end up making advertising illegal.

Is it open source?

There's an interesting issue here. Because Twitter makes no real effort to block bots currently, it's still reasonably possible to detect bots with at least a fair degree of accuracy. That's because they have no major reason to do anything beyond fool the lowest common denominator.

But as you add checks, particularly convenient and handy little instant checks like this, you incentivize the bots to become more intelligent. And it's not exactly difficult to fake being a human when you're limited to 140 characters in a mostly unidirectional shout box. So the net result is that you just end up with bots more able to tolerate inspection, and thus more likely to fool even more people.

---

As for Twitter themselves, in spite of their rhetoric they are going to love bots. It helps maintain an extremely healthy critical mass, results in that all so addictive rapid feedback of social media, and also inflates their numbers for advertisers. So long as they put on a song and dance about cracking down on bots, they can sell maximally inflated usage and activity numbers with plausible deniability if they turn out incorrect.

Reply on news.ycombinator.com