Masayoshi Son, CEO of SoftBank and one of the largest investors in artificial intelligence, has issued a significant warning on the future of AI, cautioning that superintelligence could become “extremely dangerous” if it falls into the wrong hands. He made the remarks at the Science and Technology in Society Forum held in Kyoto, Japan.
Speaking at the event, Son stressed that as AI capability grows unpredictably and rapidly, trust and cooperation among nations have become more essential than ever, stating that the world no longer has the luxury of humans fighting humans and must instead work together to control the far greater threat posed by powerful AI. Son, who has long emphasised AI’s enormous benefits for humanity, striking a notably more cautious tone this time is significant, reflecting a broader shift in sentiment amid recent concerns surrounding AI model safety. SoftBank is known to have invested around $65 billion in OpenAI.
At the same forum, White House Science and Technology Policy Advisor Michael Kratsios announced the formation of a US-led coalition aimed at using AI to accelerate scientific research. Seventeen countries, including Japan, South Korea, the UK, Germany, Singapore, and the UAE, are part of the initiative, though China was notably absent from the initial list. Kratsios said the coalition would, in the future, establish automated cloud laboratories capable of running hundreds of experiments simultaneously through robotic systems, enabling research that would otherwise take years to be completed within hours.
Son also reiterated his long-standing projection that by 2040, at least 20 percent of global GDP — roughly $46 trillion — would be generated through AI and machine learning. He outlined AI’s progress as unfolding across three stages: the current chatbot stage, followed by a browsing and coding stage, and then, from next year, a stage involving physical interaction with robots and machines, making clear that current developments represent only the beginning.