OpenAI CEO urges U.S. to prepare for AI ‘superintelligence’ risks and gains

OpenAI CEO Sam Altman is urging U.S. policymakers to proactively prepare for the risks and benefits of advanced artificial intelligence, particularly 'superintelligence.' He highlighted that AI is rapidly transitioning from theoretical concepts to practical economic applications, already assisting with tasks like coding and research that previously required teams. Altman foresees even more advanced AI systems aiding scientific breakthroughs and enabling individuals to perform the work of entire groups. He also pointed out the growing cybersecurity risks in the crypto industry, where AI tools are lowering the barrier for attackers to exploit software flaws, potentially leading to increased asset theft. Altman stressed the need for stronger defenses, including verified code and secure hardware. Beyond cybersecurity, he warned of AI's potential to accelerate harmful biological research and enable powerful cyberattacks, suggesting these threats could emerge within a year, necessitating urgent coordination. He also discussed the debate around OpenAI's potential nationalization, arguing that U.S. leadership in developing aligned superintelligence is crucial and that a government project might not be optimal. Altman believes AI will become a utility, like electricity, with personal AI assistants becoming common, and emphasized the importance of trustworthy individuals building these systems.

(Source：CoinDesk)