Musk warns there's still 10-20% chance of AI turning dangerous
What's the story
Elon Musk, the visionary behind SpaceX and Tesla, has long been worried about the risks posed by artificial intelligence (AI).
In a recent interview at the Future Investment Initiative in Riyadh, Saudi Arabia, he hinted that there is a 10-20% chance that AI could present major challenges by the end of this decade.
While Musk is optimistic about its benefits, he warned against downplaying these potential threats.
Future projections
Musk predicts AI's exponential growth
Previously, Musk had predicted at the Abundance Summit in March, that AI is progressing "as fast as 100 times per year" and by the end of this decade, it might be "as capable as 8 billion humans."
He repeated the prediction during his recent interview, adding that he is comfortable saying that it's improving 10 times a year.
The rate of progress means that in four years, AI could be 10,000 times more advanced than it is today, per Musk.
Capability forecast
AI could match human capabilities soon, says Musk
Musk also hinted that in the next year or two, AI could be capable of doing any task a human can do.
He also predicted that in three years, it could reach a stage where it could do "what all humans can do, combined."
The prediction highlights his faith in the rapid evolution and capabilities of AI.
Technological expansion
xAI to boost processing power
In other news, Musk also revealed that his AI start-up, xAI, will soon be getting a major boost in processing power.
The company plans to add 200,000 of NVIDIA's H100 and H200 chips to its training cluster in the near future.
This will further enhance Colossus, xAI's supercluster that already houses over 100,000 of NVIDIA's H100 GPUs.
AI dominance
Musk envisions a future dominated by AI and robots
Musk shared his vision of a future where "every country will have AIs, or multiple AIs," and that "there will be a lot of robots — more robots than people."
This forecast is in line with his earlier statements about the rapid advancement of AI.
It also emphasizes his belief in the transformative potential of this technology on a global scale.