According to an Apr. 18 post by Fox News, Elon Musk told the media outlet’s anchor Tucker Carlson that he’s developing a ChatGPT rival in the form of “TruthGPT,” a large language model (LLM) that Musk says will be trained to explore the mysteries of the universe.
“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,” said Musk.
This truth-seeking AI, as per Musk, will also push back against what he perceives as “left wing” bias in the industry. Musk told Carlson that ChatGPT “is programmed by left-wing experts, which train the chatbots to lie.” He also appeared to address concerns over his entering the crowded LLM market just weeks after signing a petition calling for a pause on related research in order to evaluate safety concerns. Carlson, for his part, also added that “the deeper problem is not simply that it will become autonomous and turn us all into slaves, but that it will control our understanding of reality and do it in a really dishonest way. It could be programmed to lie to us for political effect.”
Per the Fox News article, Musk said “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 also appeared to address concerns over his entering the crowded LLM market — a move he signalled with the purchase of a reported 10,000 GPUs — just weeks after signing a petition calling for a pause on related research in order to evaluate safety concerns.
“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 is a developing story and will be updated accordingly.
