Elon Musk, the billionaire Twitter owner, is once again sounding the alarm about the risks of artificial intelligence to humans, alleging that a popular chatbot has a leftist bias that he aims to fight with his AI invention.
Musk told Fox News anchor Tucker Carlson in a program aired Monday night that he aims to build “TruthGPT,” an alternative to the famous AI chatbot ChatGPT that would be a “maximum truth-seeking AI that tries to understand the nature of the universe.”
Musk said that an AI that seeks to understand humanity is less likely to kill it.
Musk also expressed concern that ChatGPT “is being trained to be politically correct.”
Musk also argued for artificial intelligence regulation in the first two-part interview with Carlson, calling himself a “big fan.” He stated AI is “more dangerous” than vehicles or rockets and can wipe out all humans.
According to a Nevada business registration, Musk has formed a new company named X.AI Corp. According to the Nevada Secretary of State’s website, the company was founded on March 9 and is led by Musk and his longtime advisor, Jared Birchall.
Musk has long had strong views on artificial intelligence and has criticized other industry executives, like Mark Zuckerberg and Bill Gates, for having a “limited” grasp of the area.
Musk was an early investor in OpenAI, the firm behind ChatGPT, and co-chaired its board of directors when it was founded as a nonprofit AI research center in 2015. However, Musk only stayed for a few years, leaving the board in early 2018 in a decision that the San Francisco company attributed to Tesla’s efforts to develop autonomous driving systems. “As Tesla continues to focus on AI, this will eliminate a potential future conflict for Elon,” OpenAI said in a blog post in February 2018.
“I came up with the name and the concept,”
Musk told Carlson, criticizing that OpenAI has become too closely associated with Microsoft and is no longer a charity.
Musk commented on his leaving in 2019, stating that it was also due to his desire to concentrate on technical challenges at Tesla and some disagreements with OpenAI’s executives. It was “just better to part ways on good terms,” he reasoned.
“Tesla was competing for some of the same people as OpenAI & I didn’t agree with some of what OpenAI team wanted to do,”
Musk said in a tweet, without elaborating.
However, concerns have been raised about the quality of Tesla’s AI systems. Last month, US safety officials launched an inquiry into a deadly incident in California involving a Tesla suspected of employing an automatic driving system when it collided with a parked firetruck.
The firetruck inquiry is part of the agency’s wider examination into various instances of Teslas utilizing the automaker’s Autopilot technology colliding with parked emergency vehicles responding to prior collisions. In the last year, the NHTSA has become more active in pursuing Tesla safety issues, issuing many recalls and investigations.
After Musk resigned from the board, OpenAI publicly announced the first iteration of their GPT technology, on which ChatGPT is based, and started incorporating as a for-profit firm.
Musk tweeted in 2020 that
“OpenAI should be more open” despite having “no control & only very limited insight“
He’s complemented. Musk tweeted to OpenAI CEO Sam Altman that ChatGPT was
and that the media wasn’t covering it because
“ChatGPT is not a far-left cause.“
Musk has since regularly cited left-wing prejudice or censorship. ChatGPT, like other chatbots, incorporates filters to avoid toxic or offensive responses.