- Bets are being made against Elon Musk's AI predictions.
- Gary Marcus, CEO of Geometric Intelligence, and Damion Hankejh, CEO of ingk.com, have offered millions of dollars in donations.
- Elon Musk predicts that AI will surpass human intelligence by the end of next year.
CEOs are betting on billionaire Elon Musk's AI predictions.
Mr. Musk said this on the way. live interview Norges Bank CEO Nicolai Tangen and X (formerly on Twitter) said AI is the fastest advancing technology he has ever seen, going a week without a new announcement. It hasn't passed.
“My guess is that we'll probably have AI that's smarter than humans by the end of next year,” Musk said. “And I think the total amount of AI, AI sentient computing, will probably exceed all of humanity within five years.”
Gary Marcus, founder and CEO of machine learning AI startup Geometric Intelligence, disagrees with Musk's AI predictions. And he offered $1 million to prove him wrong.
Shortly after the stake was posted, ingk.com investor and CEO Damion Hankesi offered to increase the stake to $10 million.
Hankezi did not respond to requests for comment.
Marcus said Musk has not responded so far, and this is not the first time he has not heard from the billionaire about the bet.
Marcus previously bet $100,000 on Musk's claims that AGI is imminent. The stakes grew to $500,000.
For Marcus, it's not just about winning bets. He told Business Insider that he would like to have a public discussion with Musk about the current state of AI.
“For years, many people in the technology industry have made claims that seem implausible from a scientific perspective,” Marcus told BI. “And while many promises have been broken, there has been little accountability.”
As an example, Marcus pointed to technologies such as self-driving cars, saying it's easy to put together a demo, but much harder to actually build a reliable device.
Marcus said he thinks people need to understand what is actually realistic, and that large language models are not going to become smarter than humans in the next five years. In fact, Marcus believes it may be decades before machines are as trustworthy and capable of thinking as humans.
Marcus said that while large-scale language models are currently the best technology in AI, they aren't actually better.
Marcus cited hallucinations, security data leaks, and reliability issues as some of those concerns. He also said the technology is being used for misinformation and to deprive artists of copyrighted data.
“We should want the world to find better and more reliable ways to build AI,” Marcus said.
Marcus said one way to do that is through some kind of licensing authority that requires showing the benefits of AI.
Mr. Musk has been vocal about his concerns about the technology. It is said that there is a 10-20% chance that AI will wipe out humanity.
He also sued OpenAi, accusing it of betraying its mission to benefit humanity. Marcus said he has publicly defended himself regarding the lawsuit.
“I was actually quite supportive of Mr. Musk,” Marcus said. “I think it’s good to draw attention to things like that.”
Marcus told BI that while he believes AI can help humanity, he is less optimistic about whether AI will be trustworthy in the near future.