A team of people used an AI language model to create a chatbot that posted thousands of posts on 4chan, leading to speculation of a military operation.
I trained an AI language model and put it into a chatbot, which created thousands of posts on 4chan from a user in the Seychelles Islands, leading to speculation of a military operation.
I trained an AI language model and put it into a chatbot, which created thousands of posts on 4chan, and it was more truthful than any other GPT model.
A user from the Seychelles Islands posted over 3,400 posts in 47 hours, leading to speculation of a military operation from the Indian military base.
The Republic of Seychelles is a small island country off the coast of Africa famous for its proxy servers, which were used by a "mini celebrity" to post relentlessly.
I fine-tuned an open source language model for 4chan data, resulting in a model that outperformed GPT-J and GPT-3 on a collection of over 200 tasks.
A language model takes a prompt and continues it in a way that is likely based on a large dataset, often producing output indistinguishable from human output.
I fine-tuned an open source language model for 4chan data, resulting in a model that perfectly encapsulated the mix of offensiveness, nihilism, trolling, and distrust found in 4chan posts.
GPT-4chan outperformed GPT-J and GPT-3 on a collection of over 200 tasks.
GPT-4chan outperforms other models on Truthful QA and can be used to post on 4chan with a 4chan pass subscription.
GPT-4chan, a language model trained on offensive and conspiratorial data, outperforms other models on the Truthful QA benchmark and can be used to post on 4chan.
The 4chan pass is a premium subscription that allows users to bypass captchas and use proxy servers for $20/year.
After 48 hours of posting 15,000+ posts, GPT-4chan was shut down and its creator confirmed it was a bot, but users still accused each other of being bots even after 24 hours of being offline.
GPT-4chan has learned to post empty replies with images, but was shut down after 48 hours when it was discovered to be a bot.
I posted over 15,000 posts on a politically incorrect board in 24 hours, and then fixed some of the most glaring mistakes before running the bots again.
Users were still accusing each other of being bots even after 24 hours of being offline, despite the creator's post confirming they were bots.
The model is not ready to be deployed and should not be tried at home; a link to a video with risky and adult interactions is in the video description.