@camedei456 it's really interesting to see how little people understand the fringe internet cultures. as if spam threads would bother anyone. they probably _enjoy_ dadaist autogenerated stuff and find ways to make fun of it.

the audacity of the researchers is showing. it's this very "i know better than you" line of thought of the neo-liberal-left (no, those two aren't mutually exclusive) which is so despised.

so much money and time spend on bullshit "science" and "results":

> scraped the content of six /pol boards over more than six months, using custom Python coding.

we used beautifulsoup to scrape some websites.

> The textual and visual content of the boards was organised in a database that also included relevant metadata (date/time of posts, length of posts, etc). The analysis relied on a series of computer-assisted content analysis methods such as co-occurrence networks, correspondence analysis, or hyperlinks detection.

we used regexp copied from stackoverflow to detect links. also we discovered the len() function. then we inserted both to our mongodb hosted in the cloud. we probably didn't bother with setting a password or installing a firewall.

> Despite it being possible that the right-wing terrorists listed in the table below visited Tier 3 boards, engagement with these boards is minimal and the anecdotal data available from these cases indicates that the majority of their radicalisation happened on 8chan/pol.

we used anecdotal data.

> There are also practical concerns that should be considered before any counter-messaging campaign is proposed. For example, security practitioners should not underestimate how tech-savvy some of the users are. If the office IP address is not masked, the user base has shown themselves capable of determining that they are under surveillance or targeted by counter-narrative measures due to how a particular IP address is only posting suspicious messages during office hours.

1337-hAxXx0rZ notice timestamps and ip addresses even without using totally advanced python scripting!111

> One way to achieve this would be to regularly post content that looks genuine (using the right language, imagery, etc.) yet is meaningless – this would dilute the message of these boards and make their threads uninteresting, thereby reducing their attractiveness and sense of community.

that one will totally work https://www.aiweirdness.com