Masayoshi Son rarely warns of the extreme dangers of super intelligence and calls for international cooperation

[Financial Channel/Comprehensive Report] Masayoshi Son, the founder of SoftBank Group who has long supported artificial intelligence (AI), issued a rare warning that as AI capabilities continue to grow exponentially, super intelligence may become "extremely dangerous" if it falls into the "wrong hands." He emphasized that there is no room for countries to continue to confront each other, and they must strengthen cooperation, build mutual trust, and jointly deal with the risks brought by powerful AI models.

Foreign media reported that Masayoshi Son recently delivered a speech at a side event at the "Science and Technology in Society" Forum (STS Forum) in Kyoto, Japan, saying that AI capabilities are rapidly leaping forward and it is urgent for all countries to work together to master this technology. He warned that the threats posed by powerful AI models may be far more serious than internal conflicts among humans, so international cooperation must be carried out.

Please read on...

This statement is in sharp contrast to Son’s long-term optimistic stance on AI. The 69-year-old is an important investor in the global AI field. The cumulative investment of SoftBank's funds in OpenAI, the developer of ChatGPT, has reached nearly 65 billion US dollars. He also continues to emphasize that AI has the potential to bring investment returns and benefit mankind.

Son’s rare warning also comes at a time when AI models have been experiencing disturbing security vulnerabilities in recent months. In his speech, he made it clear that once super intelligence falls into the wrong hands, it "could become super dangerous" and all countries must work together to deal with it.

White House science and technology policy adviser Michael Kratsios, who was on the same stage as Masayoshi Son, announced in Kyoto an initiative led by the United States and involving 17 countries, with the goal of using AI to accelerate scientific research. Representatives from Japan, South Korea, the United Kingdom, Germany, Singapore and the United Arab Emirates also participated in related activities.

Kratsios pointed out that AI can be applied to automated cloud laboratories operating on an industrial scale, where robotic systems can simultaneously conduct hundreds of parallel experiments, generate data in real time, and then feed back analysis and hypothesis optimization. He also said that AI can now integrate thousands of papers to help researchers shorten the time required for research work.

As for the next stage of AI development, Son is expected to enter a new stage next year. At present, AI can browse the web, write codes, and use more and more general-purpose chip computing power; in the next stage, it will further assist robots and other devices to start interacting with the physical world.

In terms of longer-term forecasts, Son reiterated that by 2040, AI and machines with learning capabilities may account for at least 20% of global GDP, equivalent to US$46 trillion. "This is just the beginning," he said.

Grasp the economic pulse with one hand. Click here to subscribe to Free Finance Youtube channel