Scientists map AI building itself after the last AI built by humans.
They publishedCould AI ever build itself? Scientists map the road to the 'last AI built by humans.'
First reportLive Science · 30 Oct, 10:00Z
Latest reportLive Science · 30 Oct, 10:00Z
Span0 hours · two reports
Base — every candidate starts here5
Public-interest topic (politics, world, science, health…)+1
Public-interest terms in the text (courts, tariffs, ceasefire…)+1.25
High-reliability source boost+0.75
Significance8
Deterministic and identical for every story — no engagement signals, no personalization of the score.
Why it matters
What
Researchers propose five levels, while early tests remain limited and require substantial computing and human oversight.
Why
Full autonomy depends on systems handling training, hardware and testing without people.
Watch
Progress with less human oversight.
Where the coverage comes from
Perspective mix: 1 center.
Lean and credibility are curated v1 mappings from public ratings (AllSides / MBFC); unknown outlets stay unrated, never guessed. How outlets are rated →
How it unfolded — every article
Summary
- Researchers study recursive self-improvement (RSI) to reduce AI development costs.
- Paper outlines a 1-5 level framework for AI self-improvement capabilities.
- Early experiments show limited RSI forms, like A-Evolve-Training improving a 30B LLM.
- Experts question feasibility, noting need for training, hardware, and testing autonomy.
- Authors warn substantial computing and human oversight are still required for RSI.