SpaceX and Twitter CEO Elon Musk has announced his plans to launch a rival chatbot to ChatGPT called TruthGPT. This announcement came during an interview with Fox News host Tucker Carlson where the tech mogul criticized Microsoft-backed OpenAI for “training the AI to lie.”
Though there have been some criticisms lobbed at OpenAI for ChatGPT’s inaccurate statements at times, Musk’s desire to create his own LLM that could rival ChatGPT would not only take considerable amounts of resources, but time, for the CEO who is already juggling Twitter, SpaceX, Tesla, and other companies.
But the claim Elon Musk has for OpenAI’s training of ChatGPT wasn’t his only criticism. He also hammered OpenAI’s lack of transparency by not allowing its code to be open-source, and being “for profit” while being “closely allied with Microsoft.”.This is interesting since Musk was on the board of OpenAI until 2018. He stepped down in order to focus more on Tesla and SpaceX. He also said of his departure, “Tesla was competing for some of the same people as OpenAI & I didn’t agree with some of what OpenAI team wanted to do.”
It wasn’t just OpenAI for whom Musk was critical. He also pointed to Google’s co-founder, Larry Page, who he claims doesn’t take AI safety seriously. The calls for greater oversight and regulations have been drummed up over the past few months. The Chamber of Commerce a few weeks ago, called on Congress to begin investigating regulatory frameworks for AI, while the FTC also made clear that they were keeping a close eye on the industry.
The White House is also asking federal agencies to look into their current legal toolkits to see how they can manage the growth of AI in society with the release of the AI Bill of Rights last summer. After those statements, Musk offered up his own solution with what he deems TruthGPT. “I’m going to start something which I call TruthGPT, or a maximum truth-seeking AI that tries to understand the nature of the universe.”
In his view, it’s all about AI safety and humanity’s future, “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.”
This all comes a few weeks after Musk and over 1,000 other tech leaders signed a letter asking for startups and companies such as OpenAI to take a six-month pause on the development of any LLM more powerful than GPT-4. So did Elon Musk change his mind, or is he taking another approach?
You can watch the part of the interview where Musk opens up about TruthGPT via the Fox News YouTube Channel: