Elon Musk has stated that he is working on “TruthGPT”, a ChatGPT replacement that functions as a “maximum truth-seeking AI”.
During an interview with Fox News’s Tucker Carlson, the billionaire outlined his vision for an AI rival, stating that an alternative approach to AI creation was required to avoid the annihilation of humanity.
The complete interview is available below.
“I’m going to start something called TruthGPT, which stands for maximum truth-seeking AI that tries to understand the nature of the universe,” Musk added. “And I think this might be the best path to safety in the sense that an AI that cares about understanding the universe is unlikely to annihilate humans because we are an interesting part of the universe.”
Elon Musk Claims ‘TruthGPT’ will be a ‘Maximum truth-seeking AI’
Musk contrasted an AI’s alleged lack of intent to harm mankind to humans’ efforts to rescue chimps, which is hilarious considering Neuralink’s treatment of them.
“We recognize that humanity may decide to hunt down and kill all chimps,” Musk stated. “We’re grateful for their existence, and we hope to protect their habitats.”
Musk portrayed TruthGPT as a corrective measure for OpenAI, the AI software charity he helped build, which has now launched a for-profit subsidiary. Musk said that OpenAI’s commercial motives would interfere with the ethics of the AI models it develops, and he positioned “TruthGPT” as a more transparent choice.
During the conversation, Musk also restated his long-held fears about artificial intelligence’s potential catastrophic capability. “AI is more dangerous than, say, poorly managed aircraft design or production maintenance, or poorly manufactured automobiles,” he remarked.
“It has the potential — however small one may regard that probability, but it is non-trivial — of civilizational destruction,” Musk continued.
Musk expressed similar reservations about the rapid advancement of AI in a 2014 interview, adding that its utility function “could be detrimental to humanity.”
Musk has already experimented with artificial intelligence. He co-founded OpenAI, the corporation he now hopes to defy, in 2015. He was its chairman until 2018 when he resigned due to concerns about a conflict of interest. Musk has now severed his relations with OpenAI.
‘What we need is TruthGPT’ Musk says
Musk has previously expressed interest in developing a “TruthGPT.” In February, he tweeted, “What we need is TruthGPT,” while also highlighting the dangers of large-scale AI models like those developed by Open AI. Musk and many other AI academics signed an open letter in March urging corporations to halt “Giant AI experiments” that their designers are unable to “understand, predict, or reliably control.”
What we need is TruthGPT
— Elon Musk (@elonmusk) February 17, 2023
It’s unclear how far along Musk’s “TruthGPT” is — if it even exists at all — but it appears he’s serious about it since he mentioned the model during his discussion with Carlson. Musk also secretly founded a new AI business, X.AI, in March.
Notably, Musk co-founded OpenAI in 2015, although he left the company’s board of directors in 2018. He announced a year later that he was leaving OpenAI to focus on Tesla and SpaceX. He also stated that “Tesla was competing for some of the same people as OpenAI & I didn’t agree with some of what the OpenAI team wanted to do,” according to Reuters.
We will see what the man who sends satellites to space can do in the field of language modeling. Don’t want to wait for Elon’s ChatGPT alternative? Check out ChatGPT competitors that are just as good as ChatGPT.