Elon Musk has made a bizarre claim about the future of artificial intelligence (AI) technology during an appearance on Joe Rogan's podcast.
Musk made his seventh appearance on the 'Joe Rogan Experience' podcast on October 31.
The episode racked up nearly 2 million views in the 24 hours since it was published.
In the episode, Musk spoke about his relationship with US President Donald Trump, his ownership of X (Twitter) and the future of AI.
Advert
The 54-year-old launched the Grok chatbot, which was developed by xAI and is integrated into his social media platform for a range of different uses.
Grok is described as a 'truth-seeking AI companion' and 'politically neutral', but regularly makes accuracy errors and has faced wide-ranging criticism over how it handles gender-related topics among other debates, as well as general emotions.
Musk, who holds significantly controversial views on diversity and gender and has repeatedly claimed that there is a 'woke mind virus', has now addressed the future of AI and one of the problems he believes the technology faces.
Speaking on 'The Joe Rogan Experience', Musk said his aim is for AI to be 'truth-seeking' and addressed concerns that a person or an organisation could hijack the technology and gain ultimate control over it.
"I don't think anyone is ever going to have control over artificial super-intelligence," he said. "I do think that it matters how you build the AI, and what kind of values you instill in the AI.
"My opinion on AI safety is that the most important thing is that it be maximally truth-seeking. You don't force the AI to believe things that are false.
"Let's say if you told AI that diversity is the most important thing, and now assume that becomes omnipotent, and you've also told that there's nothing worse than mis-gendering.
"So, at one point, ChatGPT and Gemini, if you asked which was worse out of mis-gendering Caitlyn Jenner or a global thermonuclear war where everyone dies, it would say mis-gendering Caitlyn Jenner. Which even Caitlyn Jenner disagrees with.
"Imagine as that AI gets more and more powerful, if it says the most important thing is diversity, the most important thing is no mis-gendering."
He then continued: "To ensure no-one gets mis-gendered, if you eliminate all humans then no-one can get mis-gendered, because there's no humans to do the mis-gendering. So you get these very dystopian situations.
"Or if it says that everyone has to be diverse, it means that there can be no straight white men, and so then you and I get executed by the AI."
Earlier this week, the tech billionaire launched Grokipedia - an alternative to Wikipedia which he claims has been 'fact-checked' by AI.
However, there have been a number of reports from its initial users of accuracy errors, and allegations that it has copied significant sections of Wikipedia pages.
New York Times reporter Kate Conger, meanwhile, claimed that 'some Grokipedia entries appeared to align with Mr Musk's views', including one on gender transition, which he opposes.
Ryan Smart