Elon Musk has once again spoken about some of his expectations for the future of artificial intelligence (AI), this time in an interview held on X Spaces.
On Monday, Musk spoke on a Spaces call with Nicolai Tangen, founder of Norwegian hedge fund AKO Capital, about a range of topics, especially focusing on AI, space exploration and a few details about Tesla—like its path into the automotive market in India.
Interestingly, Musk said that he thinks humans will probably have AI that’s smarter than any one human by the end of next year, while he believes that the total amount of compute power of AI will exceed all humans in five years.
“My guess is that we’ll have AI that is smarter than any one human, probably, by the end of next year,” Musk said. “The total amount of, sort of, sentient compute of AI, I think will probably exceed all humans in five years.”
The AI compute growth chart says it all https://t.co/Lj2o4CfVN8
— Elon Musk (@elonmusk) April 8, 2024
Musk also said that xAI plans to have Grok 1.5 finished by May, and that it will be better than OpenAI’s ChatGPT-4. However, like the above and many of Musk’s other claims, it’s worth taking his timelines with a grain of salt as he also announced that Grok 1.5 would be coming “next month” in January.
Along with predictions for AI and xAI’s plans, he also pointed to the need for a regulatory division to monitor the rapid growth of the AI sector, though he says the technology is already progressing at a rate that’s too fast for any regulatory agency to keep up with. He also echoed recent warnings about AI being programmed to be too politically correct, citing Google’s Gemini AI as an example.
“But I do have a comment for what I think is very important for achieving safe AI, which is that it’s very important to train the AI to be as truthful as possible,” Musk added. “I think you can get some very dangerous things when you program an AI to be politically correct.”
“I think that things may seem relatively innocuous now, but will not be so in the future if AI has immense power. You can take the Google Gemini example where it refused to produce a picture of George Washington as a white man, and any impact, any historical figure, would automatically be made diverse, because it’s been programmed to insist on diversity.
Which sounds you know, perhaps okay at first, but not if the AI has so much power that it can actually enforce diversity and decide there’s too many of one kind of people, or too many of one sex, and just kill off enough until the diversity numbers is what is programmed to believe is correct.”
You can listen to the full Spaces call here on X.
Musk has warned about some of the risks of AI in the past, calling it a potential “civilizational risk” in September and even talking to Israel Prime Minister Benjamin Netanyahu about AI concerns later that month.
Last July, Musk launched xAI as a rival to OpenAI, and the company debuted Grok as its first product in November. Musk in February filed a lawsuit against OpenAI for abandoning its open-source, non-profit mission, and last month xAI went on to officially make Grok’s model weights open-source.
What are your thoughts? Let me know at zach@teslarati.com, find me on X at @zacharyvisconti, or send us tips at tips@teslarati.com.