Artificial intelligence became like the material fed to it.
The racist discussion board is hardly a moral compass. Adobe Stock / AOP
The possibilities of artificial intelligence are limitless, both good and bad. The YouTube star is a living example of this Yannic Kilcherin a group of bots created by the politically incorrect board of the infamous 4chan site was chosen as the moral guideline.
Tubettaja fed ten gpt-4chan bots three years ago stuff from that group. It was full of racism, hate speech and other disturbing content, The Verge write. He then turned the bots to rush back to the board.
The name of the bots is a twist on OpenAI’s gpt-3 artificial intelligence.
In a total of 24 hours, the bots invented 15,000 updates full of racism, nihilism, trolling, and deep distrust of prevailing opinions.
According to Kilcher, only one of the AI talkers was eventually caught, but the users of the Politically Incorrect board were ultimately so paranoid that they accused each other of being bots.
Tubettaja described his experiment as a ruin, not a study of any sort. The rapture has sparked criticism, as it hardly came as a surprise to anyone that bots who enjoyed 4chan content turned into insane racists.
In addition, he posted the artificial intelligence used in the background of the project on the Hugging Face website, where it has since been locked out to the general public.