this post was submitted on 07 Apr 2024
437 points (100.0% liked)
Technology
37723 readers
545 users here now
A nice place to discuss rumors, happenings, innovations, and challenges in the technology sphere. We also welcome discussions on the intersections of technology and society. If it’s technological news or discussion of technology, it probably belongs here.
Remember the overriding ethos on Beehaw: Be(e) Nice. Each user you encounter here is a person, and should be treated with kindness (even if they’re wrong, or use a Linux distro you don’t like). Personal attacks will not be tolerated.
Subcommunities on Beehaw:
This community's icon was made by Aaron Schneider, under the CC-BY-NC-SA 4.0 license.
founded 2 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
Defederating and blocking bots is the old way, it's defensive and passive and lets them continue what they're doing. It puts the decent people always one step behind.
A better way is the "you're locked in here with me" approach. Redirect bots to a version of your site that provides naught but LLM-generated Nazi furry gibberish and an endless spiral labyrinth of new pages and product reviews of nonexistent products. Try to see if there are any that do that little Javascript-evaluation-to-render-the-client-side-site thing, and if there are, have them mine cryptocurrency for you. Federate with Threads, but serve to the Threads bot an endless series of users who say nothing but constantly-rephrased additional comments which highlight in plain English language situations like this and the types of harm that Meta causes in the real world, and good things to search for if you want to pursue a better solution.
Let the botgarbage come to harm through what comes back to them from your instances. Let them figure it out, if they can, and affirmatively defederate with you instead. Welcome all comers and give good free content to the humans and nothing but pain and misery to the semi-malicious bot traffic. View any bot that talks to you that hasn't got the message yet as a new opportunity and a new challenge.
This is the new way
Tldr version; blocking is too passive. Lets shadow ban them to the tar pits and make them our slaves
Using your evil powers for good I see.
Stop selling Threads to me...
I know Canada semi-recently said a company would have to honor a deal its LLM came up with. I have to wonder if they would hold people liable for hate speech if they hosted a LLM that outputted Nazi propaganda.
So its like kitboga, but no humans at all
botbogad
Okay so I can't write for shit, but I'm officially putting out the call to all programmers who aren't shit:
Reply to this commenr, and be friends, even if you hate each other. World needs saving. Truth needs saving.
Have to say that I kinda use to like old experimental bot only subs like https://farside.link/libreddit/r/SubredditSimulator/, but the more recent versions using more recent LLMs are a bit "too much" to be funny (https://farside.link/libreddit/r/SubSimulatorGPT2/ or https://farside.link/libreddit/r/SubSimulatorGPT3/)
Some posts look a bit off, like being not verbose enough for the topic or question; a few have logical mistakes in them, but most of the GPT3 posts are not so far off from a typical reddit thread - i don't think that this stuff is filterable in any way. Lets implement it lol