On Tuesday, Elon Musk said in an interview with Fox News’ Tucker Carlson that he wants to develop his own chatbot called TruthGPT, which will be “a maximum truth-seeking AI” — whatever that means.
The Twitter owner said that he wants to create a third option to OpenAI and Google with an aim to “create more good than harm.”
“I’m going to start something which you call TruthGPT or a maximum truth-seeking AI that tries to understand the nature of the universe. 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,” Musk said during the Fox & Friends show.
Musk was also critical of OpenAI saying that he played a pivotal role in setting up the organization, but it’s not clear if it is doing any good. He also accused OpenAI of training AI models to be politically correct, which he considers “another way being untruthful.”
He previously criticized the company for becoming “closed sourced” and “effectively controlled by Microsoft” in a tweet in February.
In March, Semafor reported that Musk and OpenAI co-founder Sam Altman had a tiff in 2018. When Musk wanted to take control of the company because it had “fallen fatally behind Google,” Altman and other co-founders objected to the move resulting in Tesla CEO walking away.
Musk also showed concerns about AI being too powerful to manipulate public opinion or even cause “civilizational destruction.” This is not the first time the SpaceX CEO has talked about the dangers of AI. Most recently, he signed a letter to pause the training of models more powerful than GPT-4 for at least six months along with people like Steve Wozniak, and Tristan Harris of the Center for Humane Technology.
Earlier this week, the Financial Times reported that Musk has been working on an OpenAI rival — he has recruited former DeepMind employee Igor Babuschkin for this task.
It’s not clear if Musk wants to build a Large Language Model (LLM) that rivals the offerings from OpenAI and Google, or if he wants to push for research in other areas of AI like fairness in training models.
But as TechCrunch’s Devin Coldewey noted in his story earlier this week, Musk doesn’t need to develop a cutting-edge product to make ripples. “One does not have to topple the market leader to influence events and gain a seat at the table — often all it takes is billions of dollars. And that is something Musk reliably seems to have at his disposal,” Devin wrote.
It’s also worth noting that TruthGPT might be a temporary project name as OpenAI has applied for a trademark for the term ‘GPT’.