Sakana AI hires Jürgen Schmidhuber, inventor of deep learning, world models, and your next ChatGPT update
What happened
Tokyo-based AI startup Sakana AI appointed Jürgen Schmidhuber as its Chief Scientific Advisor. Schmidhuber is credited with pioneering deep learning and world models in AI since the 1990s. He will lead Sakana’s new Recursive Self-Improvement (RSI) Lab, focusing on AI systems that can autonomously improve their own algorithms and capabilities over time. His foundational theories have influenced Sakana’s existing projects, like the Darwin Gödel Machine, which use recursive approaches to AI development.
Why it matters
Schmidhuber’s involvement signals a push toward AI architectures that can self-iterate and enhance themselves without constant human redesign. This capacity for recursive self-improvement could speed up progress beyond current static model updates. For operators and builders, it means future AI systems may require less manual intervention and evolve more continuously. This could accelerate innovation rhythm but also raise complexity and unpredictability in AI behavior.
Sakana positioning itself around recursive self-improvement challenges prevailing incremental advances and could apply pressure on competitors relying mostly on scaling or fine-tuning large models. Meanwhile, Schmidhuber’s influence may shape new AI designs that become embedded in future ChatGPT-like updates, altering how conversational AI and autonomous agents evolve technically and structurally.
What to watch next
Focus on Sakana’s RSI Lab outputs and announcements for tangible demonstrations of recursive AI in action. Watch if commercial or open source tools emerge from this effort that reduce manual retraining or speed up model iteration cycles. Also monitor how larger AI platforms respond to this approach. Will they integrate similar self-improving concepts or double down on traditional scaling paths?
Investors and founders should track whether recursive self-improvement delivers measurable efficiency or capability gains. That could raise the bar for AI startups and require more technical rigor in model design and maintenance. The hiring of a foundational figure like Schmidhuber pressures the AI sector to reconsider how next-generation systems should evolve beyond incremental tuning.
AI Quick Briefs Editorial Desk