Artificial intelligence: Musk says AI will surpass people – 04/09/2024 – Tech

New AI (artificial intelligence) models are expected to surpass human intelligence by the end of next year, according to Elon Musk. But this is provided that the supply of electricity and equipment meets the demands of technology.

“My assumption is that we will have AI smarter than any human probably by the end of next year,” said the billionaire, who runs Tesla, X and SpaceX.

And in the next five years, AI’s capabilities will surpass those of all people, Musk predicted this Monday (8) during an interview broadcast on X with Nicolai Tangen, the executive director of Norges Bank Investment Management.

Musk has been optimistic about the development of so-called artificial general intelligence, which could make individuals smarter on any subject.

But the forecast is ahead of the timelines he and others had previously predicted. Last year, he predicted that “complete” artificial intelligence would be achieved by 2029.

Some of Musk’s boldest predictions, such as launching self-driving cars and landing a rocket on Mars, have yet to be fulfilled.

Several advances in AI over the past 18 months, including the launch of sophisticated video generation tools and chatbots, have pushed the AI ​​frontier faster than expected.

Demis Hassabis, co-founder of Google’s DeepMind, predicted earlier this year that artificial general intelligence could be achieved by 2030.

The pace of development, however, has been slowed by a bottleneck in the supply of microchips, especially those produced by Nvidia, essential for training and running AI models.

Those restrictions were easing, Musk said, but new models are now testing other data center equipment and the power grid.

“Last year it was chip-constrained…people couldn’t get enough Nvidia chips. This year it’s moving to a voltage transformer supply. In a year or two, [a restrição é] just the supply of electricity,” he said.

Musk’s eagerness to enter the AI ​​race goes against the prominent stance he took last year, calling for a pause in the technology’s development.

He warned of the supposed risks to society and humanity posed by super-powerful AI tools and called for a pause in training any system that was more powerful than OpenAI’s GPT-4, the market-leading model.

On Monday, Musk said his own AI startup, xAI, was training a second version of its Grok model. For him, the technology must be better than GPT-4. This should be completed by May and will be followed by a new model with orders of magnitude more powerful, he added.

Musk has dedicated more time and resources to xAI over the past year. It is trying to raise billions of dollars from investors in the U.S., Middle East and Hong Kong to compete with OpenAI, in a funding round that would value the company at $18 billion, according to people with knowledge of that process.

Musk has played a central, and often controversial, role in developing new AI tools over the past decade.

He co-founded OpenAI in 2015, but left the company in 2018 after disagreements with chief executive and co-founder Sam Altman over the direction of research.

Musk sued OpenAI and Altman for breach of contract in March, claiming they had compromised the startup’s stated mission of building AI for the benefit of humanity — a claim OpenAI vehemently denied.