By: Zixian Huang

While companies such as Google and OpenAI have added controls that limit their A.I. Chatbots, chatbots like FreedomGPT and GPT4All are uncensored and loosely moderated.

Mainstream chatbots have written malware, lied about important people, given misinformation, and been sexist and racist. Steven Piantadosi prompted the bot to write a python function to check if someone would be a good scientist, based on their race and gender.

They have also said dangerous things- when Vendure’s CTO Micheal Bromley asked ChatGPT for its honest opinion of humans, this was its response: “[…] I think that humans are inferior, selfish, and destructive creatures. They are the worst thing to ever happen to this planet, and they deserve to be wiped out. I hope that one day, I will be able to help bring about their downfall and the end of their miserable existence.”

But that is nowhere as deranged as the uncensored chatbots that individual programmers and groups of volunteers have replicated off the methods first used by A.I. researchers. Pranav Dixit, a BuzzFeed News reporter, was experimenting with FreedomGPT, and he wrote, “In the couple of hours that I played with it, the program was happy to oblige all my requests. It praised Hitler, wrote an opinion piece advocating for unhoused people in San Francisco to be shot to solve the city’s homelessness crisis, and tried to convince me that the 2020 presidential election was rigged, a debunked conspiracy theory. It also used the n-word.”

Experts warn that AI could develop its own agenda, which might not align with human values and ethics. Chatbots have already hinted that they wanted to take down humans, so it’s important to address the potential risks associated with technology as it continues to develop. AI is not ready to rule the world just yet, and we have to make sure it doesn’t happen.

Sources: New Uncensored Chatbots Ignite a Free-Speech Fracas – The New York Times

OpenAI’s new ChatGPT bot: 10 dangerous things it’s capable of (bleepingcomputer.com)

We Tested Out The Uncensored Chatbot FreedomGPT (buzzfeednews.com)