SoftBank's Masayoshi Son discusses AI's potential for self-designing models

Jun 5, 2026

AI Summary

Masayoshi Son, CEO of SoftBank, stated that OpenAI's next model is being designed by an AI, indicating a shift towards 'super intelligence.' He emphasized the rapid advancements in AI and suggested that within two years, AI could surpass human intelligence in many subjects.

Masayoshi Son, CEO of SoftBank, mentioned that OpenAI's next model is being created by another AI model, suggesting advancements towards super intelligence.
Son spoke with OpenAI CEO Sam Altman and engineers, who indicated that AI models may soon be capable of designing their successors.
He expressed concern that human engineers may not be able to keep pace with AI advancements, predicting that AI could be exponentially smarter than humans.
OpenAI has previously stated that its GPT-5.3-Codex model was involved in its own development, showcasing early signs of AI self-improvement.
Son believes that AI will outperform humans in 70% to 80% of subjects within the next few years, potentially being ten times smarter in those areas.
He described the AI revolution as significantly larger than the dot-com boom, with implications for various sectors including robotics and autonomous driving.
Anthropic recently warned about the risks of recursive self-improvement in AI, suggesting that it could lead to humans losing control over advanced AI systems.
The need for governance and oversight in AI development is becoming increasingly urgent as these technologies evolve.

artificial intelligencesuper intelligenceforecastsoftbankmasayoshi son

SoftBank's Masayoshi Son discusses AI's potential for self-designing models

Related Stories

Meta launches Muse Spark 1.1 AI model, claiming advancements over competitors

Improving Coding Evaluations by Distinguishing Key Metrics

CEO proposes using video game data to enhance AI training for general intelligence

Startup General Intuition aims to use gaming data to advance artificial general intelligence