Two main threats from AI – what OpenAI's head warned about
OpenAI CEO Sam Altman outlined two catastrophic scenarios for AI development: loss of control over future systems and excessive concentration of power in the hands of one person or corporation. The statement came in response to an essay by Anthropic CEO Dario Amodei, who proposed a plan to slow the development of the most powerful models. Altman supported the initiative, calling for unified safety standards.
OpenAI CEO Sam Altman publicly outlined two catastrophic scenarios for artificial intelligence development that humanity must avoid at all costs. The first is loss of control over future AI systems, when methods for aligning algorithms with human values fail to keep pace with technological progress. The second is excessive concentration of super-powerful AI in the hands of one person or corporation, which could lead to the imposition of a narrow worldview and the construction of a dystopian society. The statement followed the publication of a policy essay by Anthropic CEO Dario Amodei, 'We Must Pace the Frontier,' in which he proposed a comprehensive plan to slow the improvement of the most powerful models to protect society from unpredictable risks. Altman publicly supported his colleague's initiative on social media platform X, noting the need to regulate the pace of progress. He emphasized that OpenAI is already implementing preventive safety measures, including in-depth risk analysis before launching training of advanced systems. Altman clarified that 'slowing down' does not mean a complete halt to science—progress will continue but become more measured and controlled, and the justified cost for additional safety will benefit the entire world.