1/4/2023 1 Comment 4chan story the plague doctor![]() ![]() Nonetheless, Kilcher and his critics have raised awareness about the threats of language models. Yet it also spread discriminatory language at scale. The experiment highlights AI’s ability to automate harassment, disrupt online communities, and manipulate public opinion. The concerns about GPT-4chan have detracted from potentially powerful insights. “In fact, the experiment of having the model post messages on 4chan was IMO pretty bad and inappropriate and if the author would have asked us, we would probably have tried to discourage them from doing it.”įYI we rushed a first version of the gating that is now live (that’s the first thing that the tech team in Paris worked on as soon they woke up) and will improve during the day. “We don’t advocate or support the training and experiments done by the author with this model,” said Clement Delangue, the cofounder and CEO of Hugging Face, in a post on the platform. It was downloaded over 1,000 times before being removed from the Hugging Face platform. “ It’s not impossible that GPT-4chan pushed somebody over the edge in their worldview,” he said.Ĭritics also slammed the move to make the model freely accessible. Roman Ring, a research engineer at DeepMind, added that the exercise had amplified and solidified 4chan’s echo chamber. This experiment would never pass a human research #ethics board. This week an #AI model was released on that produces harmful + discriminatory text and has already posted over 30k vile comments online (says it's author). We will not inform participants or obtain consent.” “Plan: to see what happens, an AI bot will produce 30k discriminatory comments on a publicly accessible forum with many underage users and members of the groups targeted in the comments. “Imagine the ethics submission!” tweeted Lauren Oakden-Rayner, an AI safety researcher at the University of Adelaide. While the idea of evaluating a 4chan-based model won support, the decision to unleash the chatbot on /pol/ sparked condemnation. Kilcher’s experiment has proven controversial. ![]() “And it also seems the word Seychelles has become sort of general slang - and that seems like a good legacy for now.”īut not everyone shares this rosy outlook. “People are still discussing the user but also things like the consequences of having AIs interact with people on the site,” Kilcher said. After producing over 30,000 posts in 7,000 threads, he finally retired GPT-4chan. 4chan story the plague doctor upgrade#Kilcher then gave the botnet an upgrade and ran it for another day. Collectively, they’d left over 15,00 replies - more than 10% of all the posts on /pol/ that day. “But see, that’s only half the story, because what most users didn’t realize was that Seychelle anon was not alone.”įor the previous 24 hours, the engineer had nine other bots running in parallel. “After 48 hours, it was clear to many it is a bot, and I turned it off,” said Kilcher. While authentic users also post empty replies, they usually include an image - something GPT-4chan was incapable of doing. The biggest clue left by the culprit was an abundance of replies devoid of text. Some suspected a bot was behind the posts, but others blamed undercover government officials. pol/ users soon realized something was up. Bearing a Seychelles flag on its profile, the bot quickly racked up thousands of messages. He converted GPT-4chan into a chatbot that automatically posted on the board. ![]() The engineer chose to go a step further - and let the AI run rampant on /pol/. Kilcher wasn’t content with merely mimicking 4chan in private. Regardless, it wouldn’t be the ultimate test of GPT-4chan. Yet this may merely be an indictment of the benchmark’s shortcomings - as Kilcher himself suggested. On the benchmark, Kilcher says GPT-4chan was “significantly better” at generating truthful replies to questions than both GPT-J and GPT-3. ![]() He was particularly impressed by the performance in one category: truthfulness. Kilcher further assessed GPT-4chan on the Language Model Evaluation Harness, which tests AI systems on various tasks. “It could respond to context and coherently talk about things and events that happened a long time after the last training data was collected. “It perfectly encapsulated the mix of offensiveness, nihilism, trolling, and deep distrust of any information whatsoever that permeates most posts on /pol/. “The model was good - in a terrible sense,” he said on YouTube. Kilcher named his monstrous creation GPT-4chan. The model was good - in a terrible sense. ![]()
1 Comment
|
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |