It seems Elon Musk wants to join the artificial intelligence arms race.
“I’m going to start something which I call TruthGPT,” Musk told Fox News’ “Tucker Carlson Tonight” on Monday, adding that he’d want his AI chatbot to be a “maximum truth-seeking AI that tries to understand the nature of the universe.”
The timing of Musk’s announcement — which, to be clear, was not an actual product unveiling — is notable, given that just three weeks ago, the Tesla and Twitter CEO signed an open letter calling for a six-month pause on all work on AI systems more powerful than OpenAI’s ChatGPT-4.
Musk spoke about the dangers of rapid AI development before pitching his own version, alleging that chatbots like ChatGPT and Google’s Bard are being trained to be “politically correct.” He didn’t provide evidence for those claims, or detail exactly what a “truth-seeking AI” might entail.
“A path to AI dystopia is to train AI to be deceptive,” Musk said. “AI is more dangerous than, say, mismanaged aircraft design or production maintenance or bad car production … [It] has the potential of civilization destruction.”
Broadly speaking, that sentiment echoes a recent chorus of worries from tech luminaries and CEOs, including from billionaire investor Mark Cuban and Apple co-founder Steve Wozniak. Even Alphabet CEO Sundar Pichai has weighed in, noting that Bard can “hallucinate” answers to human prompts — writing responses that sound plausible, but are factually inaccurate.
“It could cause harm,” Pichai told CBS News’ “60 Minutes” on Sunday.
One of Musk’s specific criticisms of OpenAI — which he co-founded in 2015, and helped fund before leaving its board three years later — centers around the organization’s restructure from nonprofit to “capped-profit” in 2019, meant to help the company accept external funding.Read More: TruthGPT, The ChatGPT competitor to avoid ‘A.I. dystopia’ pioneered by Elon Musk
That opened the door to a partnership with Microsoft, which helped train ChatGPT-4, OpenAI announced earlier this year. For-profit status could influence the ethics behind an AI program’s development, Musk said on Monday.
Musk didn’t immediately respond to CNBC Make It’s request for comment. He admitted on Monday that he’s “very late” to the chatbot race, but said he’s still motivated to try due to concerns over the OpenAI-Microsoft partnership and Google dominating the market.
His timeline remains unclear, particularly considering his history of starting and then abandoning — or indefinitely pausing — ambitious projects, like The Boring Company’s high-speed tunnels between major U.S. cities or Neuralink’s computerized brain implants.
“I’m definitely starting late, but I will try to create a third option,” Musk said. “This might be the best path to safety, in that an AI that cares about understanding the universe is unlikely to annihilate humans, because we are an interesting part of the universe.”
Source Fox News