Ryan Greenblatt, chief scientist at Redwood Research, has estimated the risk of an AI takeover at between 50 and 60 percent if current development trends continue, according to The Decoder. He attributes this heightened risk to competitive pressures between leading AI companies Anthropic and OpenAI, describing the situation as a race dynamic that intensifies the pace of innovation.
Greenblatt expressed hope for an international agreement to mitigate these risks, highlighting the need for global cooperation in managing AI development. Meanwhile, commentator Sam Harris noted that this level of risk contrasts sharply with historical precedents, referencing that Manhattan Project scientists might have halted progress at just a 10 percent risk level, underscoring the urgency of addressing AI safety.
For Japanese markets, where AI technologies are rapidly being integrated across FX, crypto, and equities trading, understanding these risks is crucial as the country balances innovation with regulatory oversight.