Why I still haven’t bought into true RSI

Nathan Lambert · 19 Sept 2026 · interconnects.ai ↗

3 korrents from this piece

Nathan Lambert did not write this page.

Every claim below was made in this piece, quoted word for word and numbered in the order the piece makes them, so you can read it there rather than take our word for it. The sentence above each quote is our reading of the claim, not their wording. Each quote was checked against a stored copy of the page at build time; where the two differ, the quote is the fact.

Our reading
AI scaling laws require exponentially more compute and resources to produce only linear gains in model intelligence.

Their wordsall of our scaling laws show that you need exponential compute and resources to make linear improvements in intelligence.
Our reading
Recursive self-improvement in AI will mainly make existing model capability cheaper rather than raise peak intelligence.

Their wordsRSI is poised to make modern LLMs vastly cheaper. Trends that have shown LLMs get exponentially cheaper at a given intelligence are likely to accelerate.
Our reading
The rising level of public concern about AI extinction risk is overblown relative to how progress will actually unfold.

Their wordslossy self-improvement remains my baseline on the trajectory of progress, and the increased discussion of extinction risk seems very misplaced