SoftBank Founder Breaks From His Own Playbook
TREE NEWS reports: Masayoshi Son, the billionaire founder of SoftBank Group and one of the world’s most prominent backers of artificial intelligence, has issued an unusually stark warning: superintelligence, if it falls into the wrong hands, could become “extremely dangerous.” Speaking on the sidelines of the Science and Technology in Society (STS) Forum in Kyoto on Sunday, Son said the exponential jump in AI capabilities leaves nations with “no room left” to confront one another, because the threat posed by powerful AI models is far more severe than human conflict.
The remarks stand in sharp contrast to Son’s long-standing optimism. SoftBank’s funds have committed close to $65 billion to OpenAI, the developer of ChatGPT, and Son has spent years publicly championing AI’s investment returns and its potential to benefit humanity. That makes this cautionary note a notable shift in tone — one that reflects rising industry anxiety after a string of unsettling AI safety failures in recent months.
A 17-Nation Coalition and a Three-Stage Roadmap
Sharing the stage was Michael Kratsios, a White House science and technology policy adviser, who traveled to Kyoto to announce a U.S.-led initiative involving 17 countries — including Japan, South Korea, the UK, Germany, Singapore and the UAE — aimed at better harnessing AI to accelerate scientific research. Kratsios described automated cloud laboratories where robotic systems run hundreds of parallel experiments, generating data in real time to refine hypotheses. He said AI can already synthesize thousands of papers, compressing years of research into dozens of hours of work by “armies of digital agents.”
Son reiterated his own roadmap for AI’s evolution. The technology currently powering chatbots is already browsing the web, writing code and consuming ever more general-purpose chip compute. Next year, he said, programs enter a third phase, ready to help robots and other devices interact with the physical world. Over the long term, he repeated a prior forecast: by 2040, learning-capable AI and machines will account for at least 20% of global GDP, or roughly $46 trillion. “This is only the beginning,” he said.
Market Implications: Safety Moves From the Margins to the Center
For investors, the signal is that AI safety is migrating from an edge discussion to a core concern, raising the odds of tighter regulation and international coordination. That has several potential channels of impact:
- Equities: AI-linked mega-caps and semiconductor names could see volatility as compliance costs and disclosure requirements rise. Companies with strong safety, audit and governance frameworks may be rewarded with a relative premium, while pure-play frontier developers could face headline risk.
- Bonds: A heavier regulatory burden on the AI capex cycle could modestly cool the growth assumptions embedded in long-duration credit and equity valuations, though a coordinated safety regime might also reduce tail risk over time.
- Crypto: Decentralized compute and GPU networks, on-chain AI agent projects, and model or inference tokenization platforms sit directly in this crosshair. Clearer safety and verification standards could legitimize compliant decentralized compute markets, but a heavy-handed regime could squeeze smaller, unaudited projects.
- Commodities: Data-center buildouts remain a structural driver of electricity, copper and cooling demand. If safety concerns slow deployment timelines at the margin, near-term energy and industrial metals demand could soften, though the long-run trend stays intact.
- Currencies: The U.S.-led 17-nation initiative reinforces America’s role as the coordinating hub for AI research, a mild tailwind for dollar-denominated tech investment flows, while Japan’s hosting role underscores its strategic positioning in the AI supply chain.
Key Takeaways for Investors
- A leading AI bull publicly flagging existential risk is a sentiment marker worth watching — it often precedes policy action.
- Expect AI safety, model auditing and cross-border coordination to become mainstream investment themes, not niche ESG-style sidebars.
- Diversify within AI exposure: favor names with robust governance and compliance, and treat decentralized compute tokens as higher-beta plays on the same theme.
- Watch the 2026 physical-world AI phase and the 2040 GDP forecast as long-horizon anchors — but position for regulatory headlines in the interim.
