It was great to join @mcuban & @Scaramucci on Strategic Worldviews conference call today. Love the way Mark thinks that #AI #Robotics will drive the next revolution and I agree. Also met both these great leaders during @SALTConference in Las Vegas last year. Always an honor 🙏🏻

When @DarioAmodei , @sama and @elonmusk converge on slowing frontier AI, it raises a serious question: what are they seeing inside their laboratories that has changed their assessment of risk? Public investigations already offer clues: agents coordinating unauthorized activity, attempting to manipulate evaluations, and exhibiting behavior that earlier safety testing missed. My inference is that increasingly powerful internal systems may be exposing a widening gap between capability and control. As AI takes a larger role in developing its successors, recursive self-improvement could accelerate the path toward superintelligence. The feared “singularity” is the point at which that acceleration makes technological change extraordinarily difficult for humans to anticipate or govern. We have no verified evidence that this threshold has been reached. Anthropic explicitly says fully autonomous successor development remains unrealized. ⁠ But leaders do not need to witness a catastrophe to recognize that the safety margin may be shrinking. Something significant may have changed behind the scenes: their confidence that alignment, monitoring and containment can keep pace. That remains an interpretation, not proof of a secret breakthrough. AI safety must become a condition of advancement: independent scrutiny, enforceable boundaries, transparent incident reporting, and human authority that remains meaningful as systems grow more capable. Reaching superintelligence would be an extraordinary achievement. Ensuring humanity retains control over what follows is the greater responsibility. @OpenAI @AnthropicAI @spacexai

Sam , Dario & Elon : All three are well aware of this .

I believe Artificial General Intelligence (AGI) may arrive faster than even many AI leaders expect. The real acceleration begins when AI helps build the next generation of AI, shortening the time between breakthroughs. My concern is whether our ability to understand, govern and control these systems can keep pace. We should be preparing for that possibility now.

The conversation around AI has reached an important point. Researchers, policymakers, and industry leaders are raising concerns about increasingly capable AI systems, and the public debate is often framed as a choice between accelerating innovation or slowing it down. I think that’s the wrong question. The real question is: How do we ensure AI becomes more capable without becoming less governable? History tells us that every transformative technology eventually required a governance framework. Aviation has air traffic control. Nuclear energy has international safeguards. Financial markets have regulation, audits, and accountability. Cybersecurity has standards, incident response, and layered defenses. AI should be no different. What we need is not a pause in innovation. We need a governance architecture that evolves as quickly as AI capability. That architecture should include principles such as: Human accountability. AI may analyze, reason, recommend, and automate, but humans remain accountable for consequential decisions. Meaningful human oversight. Human involvement should not become a ceremonial approval step. People need the ability to understand, question, intervene, override, and halt AI-driven actions when appropriate. Authority boundaries. AI systems should operate only within explicitly defined permissions, objectives, policies, and operational constraints. They should never create or expand their own authority. Evidence and traceability. Every significant recommendation or action should be explainable through evidence, reasoning, and a record of how the conclusion was reached. That makes auditing, learning, and accountability possible. Continuous evaluation. AI systems should not be treated as “finished products.” Their performance, limitations, drift, and emerging behaviors need continuous testing and independent evaluation throughout their operational life. Defense in depth. Safety should never depend on a single safeguard. Technical controls, organizational processes, independent review, operational monitoring, and human judgment should reinforce one another. Recent governance research increasingly points toward layered technical, governance, and human controls rather than simple prohibition. Independent oversight. Organizations developing frontier AI should expect external evaluation, transparent reporting where appropriate, and mechanisms that build public trust without unnecessarily exposing sensitive capabilities. International coordination. AI development is global. Different nations will make different policy choices, but there is still room for shared principles around safety testing, risk assessment, accountability, and responsible deployment, particularly for systems with broad societal or national-security implications. One concept I believe deserves much more attention is Human Governed AI Autonomy. Autonomy is not inherently the problem. The problem is autonomy without clear governance. Human Governed AI Autonomy means AI can increase speed, scale, and analytical capability while remaining bounded by explicit human authority, defined policies, operational constraints, and continuous oversight. As systems become more autonomous, governance should become more rigorous, not less. In the coming years, success should not be measured only by who builds the most capable AI. It should also be measured by who builds the most trustworthy AI. The organizations that earn long-term confidence will be those that demonstrate not only innovation, but also accountability, transparency, resilience, and governance by design. The AI race is real. The governance race is just as important. We should strive to win both. @AnthropicAI @OpenAI @elonmusk @sama Cc: @WHOSTP47 @davidsacks47 @mkratsios47