I’ve been engaging more on Reddit, and I noticed something: on some subreddits, a vast number of new threads are started by bots. On Facebook and Instagram, especially on posts by local politicians, many of the comments are made by bots. Why?
Because those platforms earn revenue from selling ads. If bots participate and drive up human engagement (which they do), then it justifies the prices that those platforms charge for advertising. It puts more eyeballs on ads. So it’s in their interest to host bots. I expect that at some point, restaurants will also be full of customers who are androids, for the same reason: it makes the restaurant look more popular and engaging, which draws in real humans.
Here on the iNat Forum, we don’t have any bots (as far as I can tell). Why? Because there is no advertising. iNat pays Discourse a monthly fee, so bot participation would hurt iNat, by using server resources without contributing bug reports etc.
I’m grateful for this place! I have to learn to separate humans from bots on those other platforms.
Forum moderators do regularly remove bot-posted content here, whether autonomous or semi-autonomous. Sometimes what looks like a legitimate contributor will create a topic or comment that seems normal, but then is later edited to add different text or a link to a spam site to improve search engine optimization. Thankfully the forum has tools that can help recognize this kind of content and many forum members have also helped by flagging suspicious posts. There isn’t a ton of discussion about commercially available products in our community so that in addition to the lower amount of political content and our relatively small community probably helps keep relative numbers lower than places like Reddit/FB. I think. (looks around suspiciously)
I saw that Tony just posted https://forum.inaturalist.org/t/timed-suspension-discussion/
I learned that human reports of bots on big platforms like Facebook, Reddit, etc. often become a data point (circumstantial evidence), rather than instantly resulting in the permanent banning of the account.
Humans often wrongly report other humans as bots, because:
- The writing is quite polished or stilted (it “sounds like a bot”)
- One human is mad at another human, and wants to get them in trouble
99% of bots are caught, NOT by human reporting, but by the platform’s own “automated bot detectors” that look for machine-like behaviour, e.g. writing 100 detailed comments in seconds on different threads, being active 24 hours a day, etc.
I don’t see what the new timed suspension tools for iNaturalist have to do with preventing forum bots. iNat moderation is separate from forum moderation.