发生了什么
Masayoshi Son, SoftBank Group’s chief, delivered a rare warning about at the Science and Technology in Society forum in Kyoto on Oct 4, 2026. Speaking with U.S. White House science and technology policy adviser Michael Kratsios, Son said the rapid rise of powerful AI models poses a “super dangerous” threat if left unchecked and called for coordinated international governance.
At the Kyoto forum, Son emphasized that the exponential surge in AI capabilities makes it imperative for nations to build trust and cooperate, noting that "we don’t have any more luxury of having human fight against human." He warned that superintelligence in the wrong hands could become "super dangerous."
Son highlighted his own track record, noting SoftBank’s near‑US$65 billion investment in OpenAI, and reiterated his long‑term forecast that AI‑enabled machines could contribute at least 20 % of global GDP by 2040 (about US$46 trillion).
Kratsios used the same stage to announce a U.S.‑led coalition of 17 countries—including Japan, South Korea, the UK, Germany, Singapore, and the UAE—aimed at accelerating scientific research with AI while establishing safety norms. China was not among the initial signatories.
为什么这很重要
Son’s remarks carry because he has invested roughly US$65 billion in OpenAI and is a prominent champion of AI’s economic promise. His shift to caution signals growing concern among top investors about the societal and security implications of ever‑more capable models. The call for a multilateral framework aligns with a broader U.S.‑led effort involving 17 countries to harness AI for scientific research while managing risks, highlighting a potential new wave of policy coordination that could shape regulation, funding, and standards worldwide.
Son’s shift from unbridled optimism to caution may influence other investors and corporate leaders to prioritize safety, potentially redirecting capital toward alignment research and risk mitigation tools.
The multilateral dialogue signals a move beyond fragmented national policies toward a coordinated global approach, which could affect export controls on advanced chips, model licensing regimes, and standards for AI‑driven automation in critical sectors.
If the 17‑country coalition produces actionable guidelines, it could set de‑facto standards that shape how AI is deployed in industry, academia, and government, impacting everything from autonomous robotics to large‑scale data‑center operations.
互动机制：它实际上是如何运作的
以交互方式探索这一发展背后的基础技术。
crm_get_transaction(id='4092').Why can ethical evaluation not be reduced to one model score?
接下来看什么
Watch for concrete proposals emerging from the U.S.‑led 17‑country initiative, especially any binding agreements on model licensing, export controls, or safety testing. Also monitor SoftBank’s future investment decisions and whether the firm will fund research or impose stricter safeguards on its portfolio companies.
Follow announcements from the U.S. State Department and participating ministries for any draft treaty language or safety framework documents.
Track SoftBank’s portfolio for new safety‑focused investments or the establishment of internal bodies.
Watch for reactions from other major AI stakeholders, such as OpenAI, Google DeepMind, and Chinese AI firms, which could indicate the level of global consensus or push‑back.