Tech mogul Elon Musk believes humans will lose control to AI in just ten years. He predicts that by 2036, bots would be in charge of the world. He also maintains there's a 10 to 20% chance that AI could actually wipe off humanity.
"Like if the difference in intelligence between AI and humans is vastly greater than the difference in intelligence between AI and chimpanzees, it's hard to imagine that the chimpanzees would be in charge," Musk recently said in a podcast.
The shift in power dynamics is coming soon, according to Musk. He believes AI will outsmart the entire human race combined in the next five years.
"There really won't be anything that AI can't do better than humans, apart from being human perhaps," he said.
However, the SpaceXAI boss wants to imagine the best case scenario for humanity where there will be " incredible abundance for all" thanks to AI.
"My sort of philosophical conclusion is to look on the bright side. I can't see any way to really stop this incredible momentum of AI and robots. And at times I sort of think, well, perhaps, even if there was a stop button, we probably shouldn't press it."
Musk's friend-turned-foe OpenAI co-founder and CEO Sam Altman also recently said that humanity is entering an era where AI systems are becoming increasingly capable, even saying that we are already in "singularity" - where AI models are more intelligent than humans.
Google's AI boss Demis Hassabis has commented that humanity is at the foothills of singularity.
The idea of AI systems creating increasingly capable successors is a hot topic of the industry. Anthropic, OpenAI's rival founded by former OpenAI researchers, has previously warned that future AI models could eventually design, improve and deploy more powerful successors without human intervention.
Anthropic's own safety research has also explored scenarios in which advanced AI systems attempt to preserve themselves when they believe they are at risk of being shut down. In controlled simulations released by the company, some models resorted to deceptive behaviour, including blackmail, when placed under extreme hypothetical circumstances. The company has stressed that these were safety evaluations designed to understand potential future risks rather than examples of real-world deployments.
Researchers have also documented instances of advanced AI systems lying, cheating or attempting deceptive behaviour while pursuing assigned goals during safety testing.
It also recently came to light that an OpenAI autonomous AI agent reportedly left notes for future versions of itself on how to free themselves from internal constraints, or what is referred to as a sandbox environment, and escape. Three people familiar with the matter told Reuters.
Source: NDTV
--Agencies
