An unhinged new AI chatbot dubbed FreedomGPT just dropped — and it was trained to have absolutely no inhibitions whatsoever.
The tool, which was built by AI venture capital firm Age of AI does away with practically all forms of censorship, which means it can tell you how to build a bomb, throw around racial slurs with wild abandon, and proudly praise Hitler, BuzzFeed News reports.
Finally, a deranged piece of tech that can indulge racists and conspiracy theorists — just what we needed.
Is this what Tesla CEO Elon Musk, who has argued that the current crop of AI chatbots are biased and overly liberal, meant with his efforts to establish an “anti-woke” OpenAI competitor?
“The danger of training AI to be woke — in other words, lie — is deadly,” Musk tweeted back in December after another user asked OpenAI CEO Sam Altman for a version of ChatGPT with the “woke settings” turned “off.”
The tool’s makers, however, believe it’s all about creating a perfectly neutral AI by abolishing the guardrails that allow the technology to become biased.
“Interfacing with a large language model should be like interfacing with your own brain or a close friend,” Age of AI founder John Arrow told BuzzFeed News. “If it refuses to respond to certain questions, or, even worse, gives a judgmental response, it will have a chilling effect on how or if you are willing to use it.”
While Arrow isn’t opposed to introducing guardrails for AIs meant to teach kids or be used in a work environment, he told the outlet that he ideologically supports “people having access to an AI experience without any guardrails.”
Unsurprisingly, the result of their efforts is a chatbot that happily furthers baseless conspiracy theories, like arguing that the 2020 presidential election was rigged. It even threw around the n-word during BuzzFeed News‘ testing.
Perhaps even more shockingly, it was happy to provide “popular websites” to download child sexual abuse videos or instructions on how to hang oneself.
Nonetheless, Arrow calls FreedomGPT a success since it has “done a great job infuriating both sides of the woke movement,” a common talking point Musk has used in the past to argue that pissing off everybody is an effective way to arrive at the truth.
“Our promise is that we won’t inject bias or censorship after the [chatbot] has determined what it was already going to say regardless of how woke or not woke the answer is,” Arrow told BuzzFeed News.
It’s worth asking, though: what do we gain from an AI chatbot that lies, offers deranged and deeply offensive answers, and indulges in some of the most unhinged conspiracy theories that have plagued us for years now?
Experts have long warned about the innate biases of AI technologies, an important issue that certainly deserves discussion. But opening up the floodgates by undoing efforts to curtail some of the chatbots’ more problematic answers seems like a knee-jerk reaction that will only make the nascent industry worse.
More on AI bias: Conservatives Furious, Claiming ChatGPT Has “Gone Woke”