A substantive AI-takeoff essay reconciling a real puzzle: Yudkowsky argues AI will blow through the human range fast (mouse→chimp is a huge gap, Joe→Einstein trivial), but Katja Grace's data show AI has taken decades to cross the human range in field after field (chess, Go). Scott works through several reconciliations — mutational load, purpose-built hardware (bad chess players are near-max-bad because chess isn't native architecture; the Minecraft-calculator analogy), widely-varying sub-abilities — and lands on the one he finds most compelling: humans are so far beyond other animals ('Zoological IQ': mouse 10, chimp 20, village-idiot 50, Einstein 200) that intra-human variation is genuinely vast (the whale analogy — most size variation is within 'whales'; Hsu's IQ-1000 calc). Upshot: good news (crossing the vast human level takes time) but bad news (if evolution scaled intelligence so fast, it's easy to scale, so post-critical-mass takeoff could be fast).

Quality 76 · Excellent. Excellent: an original, memorable synthesis on an important question (AI takeoff speed), engaging a real Yudkowsky-vs-Grace debate and coining the 'humans are the whales of brainpower / vast human range' reframe — a cited contribution to the takeoff-speed discourse. 76.

Claude’s paradigm shift 56 · Moderate. An original synthesis reconciling the intuitive and empirical takeoff arguments (the Zoological-IQ / vast-human-range model). B56.

Real-world impact 2 · Minor. Influential within the AI-alignment/takeoff-speed discourse; no material reach. RWI2.

Where The Falling Einstein Meets The Rising Mouse