Google DeepMind Exec Says AI Investment Bets on Recursive Self-Improvement

DeepMind CSO Jasjeet Sekhon asserts that trillion-dollar AI capital flows are driven by the thesis of recursive self-improvement, replacing AGI as the core industry narrative.

Woofun AI reports that Jasjeet Sekhon, Chief Strategy Officer at Google DeepMind, identified recursive self-improvement (RSI) as the central thesis behind massive AI sector investments. Speaking at the Agentic AI Summit at the University of California, Berkeley, Sekhon stated that current capital influx requires validation through AI systems capable of autonomous optimization and next-generation system creation. He noted that RSI is supplanting Artificial General Intelligence as the primary industry focus, with major firms including OpenAI, Google, and Anthropic investing heavily in infrastructure to facilitate a transition from scaling to self-enhancement. Despite existing capabilities in code generation and automated optimization, achieving true autonomous recursive improvement remains contested due to significant technical, safety, and controllability hurdles.