Guys any ideas for building a thread bumper?

Nov 1, 2018
61
0
6
#1
I have an idea,

post article of that I want to use to influence a website
generate bot comments that are somewhat relatable to bump that post
make sure my post is always ranked top for exposure

subtract all comments with the bot comments I made to train my bot to generate more relatable comments

how to do this?
 
Apr 8, 2018
371
1
18
#2
That's literally my first project when I first got into AI, I wrote something like that to bump the threads of a few phpBB forums to completely desintegrate their community. People couldn't discuss what they usually discussed and were forced to see threads that were boring to them bumped to the top, and they were too lazy to click next page, they just left, haha.

The same technique is currently being done on /pol by some people, the amount of bot posts are incredible, if you know how to look for them
Their bots are something next level, they're much smarter, the board is being subverted and never progressed since 2014
 
Apr 8, 2018
437
1
18
#3
That's literally my first project when I first got into AI, I wrote something like that to bump the threads of a few phpBB forums to completely desintegrate their community. People couldn't discuss what they usually discussed and were forced to see threads that were boring to them bumped to the top, and they were too lazy to click next page, they just left, haha.

The same technique is currently being done on /pol by some people, the amount of bot posts are incredible, if you know how to look for them
Their bots are something next level, they're much smarter, the board is being subverted and never progressed since 2014
Were your bots so smart they couldn't tell and ban all of them?
 
Apr 8, 2018
371
1
18
#4
Were your bots so smart they couldn't tell and ban all of them?
It didn't work like that

Basically I made a semantic bot that understood words and sentences of each forum I wanted to nuke as high dimensional vectors, I dumped the bots' memory, chose opposing bots whose descriptive vectors of lexicon were most distant of one another's. This helped me pick pairs of forums whose ideology and interests were opposite or they just didn't care about what the other said, they had different understanding or are indifference of the each other's subjects in short. It's like finding a group who like ponies and a group who like cats.

Then I replayed one forum's discussion onto another forum, my bot had no intelligence in generating language, but it was copying what humans said and it was sure the forum that was intruded had absolutely no interest of what was being piped into but convinced it's real humans talking.

Admins and mods could ban bot accounts for saying "boring stuff", but they didn't, they just didn't know what hit them, there's no precidence in banning "boring", they were stupid jannies, they didn't even understand their userbase and were just pleased there were people talking non stop, of course they could just do a google search of the bot discussion and look for duplicates, that's why it had shortcomings and were terminated after a few tries, mostly because there's no profit in it, why would I continue after proving it works.