CNBC reports SoftBank CEO Masayoshi Son told CNBC that an AI "model is designing" OpenAI's next model and that "it's going to be exponentially smarter than all of us," calling that development a sign of "super intelligence." CNBC reports Son characterized his prior ten-year forecast for artificial superintelligence as "conservative." The article also notes that Anthropic has warned the pace of AI development should be slowed, according to CNBC. An OpenAI spokesperson declined to comment on unreleased models but, CNBC reports, highlighted areas where the company is already using AI in model development. Seeking Alpha ran a similar summary of Son's remarks.
What happened
CNBC reports that SoftBank CEO Masayoshi Son told CNBC an AI "model is designing" a future OpenAI model, and quoted Son saying, "So once that happens, [the] model generates [the] next model ... and it's going to be exponentially smarter than all of us. That's a super intelligence." CNBC reports Son described his earlier ten-year forecast for artificial superintelligence as "conservative." CNBC also reports that Anthropic has warned that the pace of AI development may need to be slowed. Per CNBC, an OpenAI spokesperson declined to comment on unreleased models but highlighted areas where the company already uses AI in parts of model development. Seeking Alpha published a similar summary of Son's remarks.
Editorial analysis - technical context
Industry-pattern observations: automating aspects of model design is an active research and engineering area, using techniques such as neural architecture search, AutoML pipelines, automated hyperparameter tuning, and model distillation. These techniques can reduce human intervention in routine design steps, though full end-to-end autonomous design of frontier models would require combining architecture search, training-scaling decisions, data curation pipelines, and safety-aligned evaluation, a complex engineering stack beyond any single published component to date.
Industry context
Observed patterns in similar announcements: public statements from influential investors or executives about rapid automation of model design tend to accelerate regulatory and safety discourse. CNBC's report places Son's remarks alongside Anthropic's call for slowing development, reflecting a broader debate among investors, builders, and safety-focused labs about pace and oversight.
What to watch
Indicators observers can follow include: announcements or papers from major labs describing automated architecture- or objective-search applied to large-scale models; reproducible benchmarks showing systematically improved architectures discovered by models; public disclosures on evaluation frameworks that test emergent capabilities and alignment at scale; and any comments or clarifications from OpenAI on the scope and mechanics of automation in their development process. Also monitor filings, conference talks, or preprints that document concrete tooling or reproducible experiments in automated model design.
Takeaway for practitioners
Industry context: claims that models are autonomously designing next-generation models highlight two practical concerns for ML teams: the increasing role of automation in iterating architectures and the need for robust, reproducible evaluation pipelines. These are trends practitioners should track, but current public reporting does not provide technical details that would let practitioners reproduce or validate the specific capabilities Son described.
Scoring Rationale #
The story is notable because an influential investor publicly described autonomous model-design as occurring at a leading lab, which can shift discourse on model development pace and safety. The claim lacks technical detail and independent verification, so its practical immediacy for practitioners is high-interest but not yet actionable.
Practice interview problems based on real data
1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.