SoftBank CEO Masayoshi Son announced that OpenAI’s next model is being designed by AI, a development he describes as indicative of “super intelligence.” In an interview with CNBC, Son expressed that this trend signals a rapid advancement in AI capabilities, suggesting that engineers may soon lack the ability to design future models. He further updated his previous timeline, now predicting that artificial super intelligence could emerge within the next two years, a significant shift from his earlier ten-year forecast. This comes amid warnings from Anthropic about the potential risks associated with rapid AI development, advocating for a cautious approach to prevent humans from losing control over increasingly autonomous systems.

OpenAI: OpenAI is an artificial intelligence research and deployment company focused on developing advanced generative AI models and tools. In the news, Son reports that the company’s engineers confirmed an AI model is designing the next model, highlighting internal use of AI for development. An OpenAI spokesperson noted existing applications of AI in model creation without commenting on unreleased systems.
SoftBank: SoftBank is a major global technology investment firm with significant holdings in AI-related companies. As one of OpenAI’s largest shareholders, the firm is positioned centrally in Son’s comments about accelerating AI progress and super intelligence. Son, its CEO, drew on direct conversations with OpenAI leadership to discuss these developments.
Anthropic: Anthropic is the developer of the Claude AI chatbot and a leading AI safety-focused lab. The company recently released a blog post warning about risks from recursive self-improvement in AI systems and suggesting that AI labs coordinate to slow development. This contrasts with Son’s optimistic views on rapid AI progress mentioned in the news.
Sam Altman: Sam Altman is the CEO of OpenAI. Son referenced direct discussions with Altman and OpenAI engineers as the basis for claims that AI is now designing future models, indicating Altman’s involvement in the reported advancements.
Masayoshi Son: Masayoshi Son is the founder and CEO of SoftBank, a prominent investor in technology and AI infrastructure. In the reported interview, he described speaking with OpenAI’s Sam Altman and engineers about AI-led model design and revised his timeline for artificial super intelligence to the next two years. He also highlighted his personal daily use of ChatGPT and SoftBank’s broader AI investments including Arm.

AI Development: OpenAI has begun using early AI versions to debug training, manage deployment, and evaluate its own systems.
Safety Concerns: Anthropic has called for coordinated efforts among AI labs to slow development of recursive self-improvement capabilities due to risks of losing human control.
Super Intelligence Timeline: Son updated his outlook on artificial super intelligence, stating it is now expected within the next two years rather than longer.