Tag
recursive-self-improvement
- EssayA Paradigm Shift in AI Safety
A one-page strategy brief introducing RESI (recursive environmental self-improvement) as a proposed, testable contribution to AI safety: let AI improve its methods inside a loop of observing, acting within limits and independent verification against physical outcomes, while human-set purpose and limits stay outside the system's authority to change. The living world serves as the external check—does greater capability actually deliver greater protection?
- EssaySuperintelligence, Grounded in Life
Superintelligence names a level of capability, not a purpose, so the deciding question is: superintelligent at what? This essay proposes Life-Grounded Superintelligence—advanced intelligence that learns from living systems, is developed to help protect them, is judged by independently verified consequences in the world rather than benchmarks, and stays under human authority—driven by RESI's two loops, in which verified environmental outcomes improve the system's methods while the mandate stays fixed.
- EssayThe Planet Is the Benchmark
Answering the call to halt recursive self-improvement, this essay argues that the peril of recursion is not its speed but its referee: a loop graded by its own benchmarks learns the exam instead of the world, while the planet is an examiner that cannot be flattered. It defines recursive environmental self-improvement as a governed loop that may improve its means but never redefine its ends, enlarge its jurisdiction, or lower its standard of proof, with physics judging truth, outcomes judging function and people judging legitimacy across nested reflex, deliberative and constitutional speeds.
- EssayRecursive Environmental Self-Improvement
AI has begun to improve the process that builds AI, so the deciding question is no longer whether intelligence will improve itself but what it will improve itself for. This essay defines recursive environmental self-improvement (RESI): an inner sense-infer-decide-act loop wrapped in an outer loop that improves the intelligence running it, verified against physical reality rather than self-approval, with separated Seeker, Builder, Experimenter, Verifier and Governor roles so the system can rewrite its mechanism but never its mission.