[DigitalToday reporter Jinju Hong] Tesla CEO Elon Musk said artificial intelligence (AI) is likely to surpass the combined intelligence of humanity within the next 5 years and could reach a level that humans would find difficult to control within 10 years. He also proposed that major companies should jointly review safety issues as AI competition accelerates.

On July 27 local time, blockchain media outlet Decrypt reported that Musk said in a recent conversation with Economist editor-in-chief Zanny Minton Beddoes that AI is developing in a way that surpasses humans in almost all intellectual tasks.

Musk said the spread of AI is a change that can reshape social structures, beyond simple technological progress. He said, "The only thing AI may be worse at than humans is being human," and projected that the most likely long-term outcome is the opening of an "amazing age of abundance" in which anyone can get what they want. He also stressed that an abundant future does not necessarily mean a safe future and underscored the importance of risk management.

He also assessed that it would be unrealistic to stop development itself given the pace of advances in AI and robotics. Musk said, "I don't see a way to stop the enormous momentum of AI and robots," adding, "Even if there were a stop button, it might be better not to press it." He said it is a more realistic approach to build mechanisms to control risks rather than artificially blocking technological progress.

To that end, Musk proposed creating a consultative body in which major AI companies around the world meet regularly to discuss safety and security issues. He said there should be a system in which even rival companies conduct limited checks on each other before releasing next-generation AI models to identify risk factors, and in which governments can intervene if serious problems are found.

While stressing cooperation on AI safety, he continued to criticise OpenAI and Sam Altman. Musk said OpenAI had strayed from its original goal of developing AI for humanity based on a non-profit and open-source foundation and had turned into a closed, for-profit company. He said, "I'm not a fan of Sam Altman," and criticised OpenAI for losing its founding mission.

Musk's legal offensive has not produced results recently. In May, a U.S. California jury did not find liability on the defendants in a $150 billion lawsuit Musk filed against OpenAI, Sam Altman and co-founder Greg Brockman, among others. Musk had filed the suit claiming OpenAI abandoned the purpose of a non-profit organisation and shifted to a for-profit structure.

By contrast, he offered a relatively positive assessment of Anthropic. Musk claimed that co-founder Dario Amodei and key figures could no longer trust Sam Altman, left OpenAI and founded Anthropic. He said, "Dario is a man of principle," and said he has not sensed malice from Anthropic members so far. He added that goodwill alone is not enough and that continued vigilance on AI safety is needed.

Musk repeatedly stressed that cooperation must take priority when it comes to AI safety even if there is competition between companies and personal conflict. He said, "If we ultimately have to talk, we'll talk," adding, "We need to set aside personal differences for the world."

As competition in AI technology intensifies, attention is focusing on whether the global AI safety consultative body Musk proposed can lead to industry-level cooperation.