Too many tangents (bots, trolls and socks, oh my!)

[swansont]

There’s been a recent uptick in off-topic commentary about (alleged) rules violations, happening instead of (or sometimes in addition to) using the report post function. Posts that purport to call out bots, sockpuppetry, trolling, even duplicate posts.

This approach show a combination of Bad Faith Arguments, Hijacking, & Soapboxing all at once. They add nothing to the topic that’s supposed to be under discussion, and often lead to these issues being debated instead. And if they aren’t being reported, the mods may not even be aware of any potential infractions.

Everybody should be afforded due process in regard to potential infractions - we aren’t going to ban someone just because of an accusation of sockpuppetry (these accusations are incorrect a significant fraction of the time) so the mods need to investigate, and picking a fight with a suspected bot is something I just don’t get.

If you think a post violates the rules, report it. Respond only to the subject under discussion. And if it’s spam, don’t reply at all. (And for Zeus’s sake, don’t quote anything that includes spam links.)

We’re considering just trashing posts that include such material as described above. Even if there’s “proper” content in it. 

[MigL]

Due Process ???
I say round them up, and re-locate them to another forum.
( as was recently proposed with the homeless )

[iNow]

It does seem like there’s an uptick in bots being deployed by their designers into discussion forums as part of their broader training strategy and ML. I’m wary of most new posters myself. It’s getting worse… fast. Request to use the report button is entirely appropriate, though. 

[Genady]

More science-is-wrong bots, but fewer god bots, isn't it?

[davidoliver]

Indeed, the online world can be filled with various types of users and phenomena that can sometimes lead to tangents and distractions. Let's briefly explore the terms you mentioned:

1. Bots: Bots, short for robots, are software programs designed to perform automated tasks. In the context of the internet, bots can serve various purposes. Some bots are helpful, such as search engine crawlers that index web pages for search results. However, there are also malicious bots, such as spam bots, which can flood platforms with unwanted messages. Social media platforms may also have bots that automate actions like posting, liking, or following accounts.

2. Trolls: Trolls are individuals who intentionally provoke or disrupt online discussions by posting inflammatory or off-topic comments with the aim of eliciting strong reactions from others. They often thrive on creating discord and chaos, rather than engaging in constructive dialogue. Trolling can occur on social media, forums, comment sections, and other online platforms.

3. Socks or sock puppets: Sock puppets are fake accounts or personas created by individuals to deceive others online. The term comes from the idea of controlling a puppet with a hand inside a sock. Sock puppets are typically used to manipulate discussions, support one's own arguments, or create the illusion of a larger following. They may be employed to spread misinformation, sway public opinion, or engage in other deceptive practices.

[exchemist]

20 minutes ago, davidoliver said:

Indeed, the online world can be filled with various types of users and phenomena that can sometimes lead to tangents and distractions. Let's briefly explore the terms you mentioned:

1. Bots: Bots, short for robots, are software programs designed to perform automated tasks. In the context of the internet, bots can serve various purposes. Some bots are helpful, such as search engine crawlers that index web pages for search results. However, there are also malicious bots, such as spam bots, which can flood platforms with unwanted messages. Social media platforms may also have bots that automate actions like posting, liking, or following accounts.

2. Trolls: Trolls are individuals who intentionally provoke or disrupt online discussions by posting inflammatory or off-topic comments with the aim of eliciting strong reactions from others. They often thrive on creating discord and chaos, rather than engaging in constructive dialogue. Trolling can occur on social media, forums, comment sections, and other online platforms.

3. Socks or sock puppets: Sock puppets are fake accounts or personas created by individuals to deceive others online. The term comes from the idea of controlling a puppet with a hand inside a sock. Sock puppets are typically used to manipulate discussions, support one's own arguments, or create the illusion of a larger following. They may be employed to spread misinformation, sway public opinion, or engage in other deceptive practices.

That's funny, you sound just like a bot. 😁

[Eise]

43 minutes ago, exchemist said:

That's funny, you sound just like a bot. 😁

I guess it is ChatGPT...

[exchemist]

57 minutes ago, Eise said:

I guess it is ChatGPT...

I did wonder. The fact that the text is highlighted with a blue background indicates it has been copy-pasted from somewhere. And the poster's one previous post is similarly a little lecture that nobody had really asked for. But one can jump at shadows these days.  

[swansont]

These “informative” posts (and their cousins, the ‘thank you, that’s interesting’ posts) are often precursors to spam; the account is just waiting for their post count to be higher (unlocking some features or just hoping not to be the new account on the block) before posting spam links.

We know the tactics, and will ban if there are actual rules violations.

[exchemist]

42 minutes ago, swansont said:

These “informative” posts (and their cousins, the ‘thank you, that’s interesting’ posts) are often precursors to spam; the account is just waiting for their post count to be higher (unlocking some features or just hoping to to be the new account on the block) before posting spam links.

We know the tactics, and will ban if there are actual rules violations.

Hmm, interesting. I had come to suspect as much but it's nice to see it confirmed.

Sometimes I will respond to a post I am a bit suspicious of, on the basis of giving the benefit of the doubt and, in the process, trying to gain evidence one way or the other.  Quite often, the reaction, or lack of it, does indeed lead me to a more decisive view on the nature of the poster. It's a bit of a dilemma, though. One shouldn't react to obvious spam but there is a bit of a grey area when this type of "sleeper bot" behaviour may be occurring. (In the present case I was hoping for an amused or indignant reply, averring that the poster is real and not a bot.)   

Edited May 24, 2023 by exchemist

[swansont]

Sometimes it’s hard to tell the difference between awkward phrasing you might get with a bot and someone using a translation program or for whom English is not their primary language, since syntax doesn’t always translate well. 
 

Our rules are set up such that only a few topics are off-limits. Other than those, moderation is not for content or veracity. We focus on the behaviors listed in the guidelines.

[Sensei]

On 5/24/2023 at 9:43 PM, swansont said:

Sometimes it’s hard to tell the difference between awkward phrasing you might get with a bot and someone using a translation program or for whom English is not their primary language, since syntax doesn’t always translate well.

Facepalm. C'mon.. make database of IP addresses..

[swansont]

5 hours ago, Sensei said:

Facepalm. C'mon.. make database of IP addresses..

Because all bots use the same IP address? Or that would tell us anything useful?

[Sensei]

8 hours ago, swansont said:

Because all bots use the same IP address?

Of course, different companies use different IP ranges.

For example, I have a database of all the end IP addresses of the Tor network. So when someone uses Tor or the Tor browser to visit my server, I can tell immediately.

You can tell if a Google crawler has visited your web server simply by checking the IP address. (not single IP address, but Google ASN IP ranges)

https://www.google.com/search?q=Google+ASN+IP+range

Microsoft's Bing https://www.bing.com/toolbox/bingbot.json

8 hours ago, swansont said:

Or that would tell us anything useful?

That it is a chatbot that is trained on scienceforums.net as a source?

"Sometimes it’s hard to tell the difference between awkward phrasing you might get with a bot"

 

[swansont]

36 minutes ago, Sensei said:

Of course, different companies use different IP ranges.

For example, I have a database of all the end IP addresses of the Tor network. So when someone uses Tor or the Tor browser to visit my server, I can tell immediately.

Just because they used Tor doesn’t make them a bot. People use VPNs or dynamically-assigned IP addresses. I’ve seen instances of spammers using the exact same IP address as were previously used by long-time members.

Quote

 

That it is a chatbot that is trained on scienceforums.net as a source?

"Sometimes it’s hard to tell the difference between awkward phrasing you might get with a bot"

 

 How do you determine that it’s a bot?

[Sensei]

12 hours ago, swansont said:

 How do you determine that it’s a bot?

A company that trains a chat bot or has a crawler will download (visit) a website (such as a forum), all posts, recursively, usually much faster than any human can do it. A human has to sleep and has limited abilities to read text. A bot that scans the site does not. Some developers add some delay between downloads, others don't. Therefore, an analysis of the web server logs for a given IP address/ranges/db may show that it is some kind of bot or crawler, simply because there are too many requests from the same client IP address (or ASN IP ranges from the same company).

Example bash script for a bot that will download the entire 1st topic page of scienceforums.net, and resemble a simple DoS attack:

#!/bin/bash
for TOPIC in {1..128952}
do
   wget -q "https://www.scienceforums.net/topic/$TOPIC--/"
done

128952 / one request per second = 36 hours to download entire scienceforums.net 1st topic page..

 

(if done from multiple IPs e.g. through Tor network (it has less than 2000 servers ATM), DDoS attack)

Tor version:
Installation:

sudo apt install proxychains tor

sudo service tor start

Usage:

#!/bin/bash
for TOPIC in {1..128952}
do
   proxychains wget -q "https://www.scienceforums.net/topic/$TOPIC--/"
done

128952 / 2000 requests per second = (theoretically, doubtful this server will handle it) 65 minutes to download entire scienceforums.net 1st topic page..

(Tor switches end-IP address every 10  minutes)

 

Going by comments, will obviously require much much more, and it will have requests like:

"https://www.scienceforums.net/topic/$TOPIC--/?do=findComment&comment=$COMMENT"

Your comment is one of the newest and has COMMENT=1244489

Downloading them all, will take 1244489/3600=346 hours (if one request per second).

 

[Meilos]

On 3/5/2023 at 1:46 AM, swansont said:

There’s been a recent uptick in off-topic commentary about (alleged) rules violations, happening instead of (or sometimes in addition to) using the report post function. Posts that purport to call out bots, sockpuppetry, trolling, even duplicate posts.

This approach show a combination of Bad Faith Arguments, Hijacking, & Soapboxing all at once. They add nothing to the topic that’s supposed to be under discussion, and often lead to these issues being debated instead. And if they aren’t being reported, the mods may not even be aware of any potential infractions.

Everybody should be afforded due process in regard to potential infractions - we aren’t going to ban someone just because of an accusation of sockpuppetry (these accusations are incorrect a significant fraction of the time) so the mods need to investigate, and picking a fight with a suspected bot is something I just don’t get.

If you think a post violates the rules, report it. Respond only to the subject under discussion. And if it’s spam, don’t reply at all. (And for Zeus’s sake, don’t quote anything that includes spam links.)

We’re considering just trashing posts that include such material as described above. Even if there’s “proper” content in it. 

It's clear that you're concerned about the quality of discussion and adherence to rules within your community. I'm new here, but I'm not gonna break anything.