Your Book Review: Why Machines Will Never Rule the World
Read the original on Astral Codex Ten →
Summary
Anonymous 2023 book-review-contest finalist on Landgrebe & Smith's Why Machines Will Never Rule the World, the best technical book-length case against AGI. The book's a-priori argument: AGI requires emulating human intelligence; human intelligence is a complex system (non-ergodic, non-Markovian, irreducible); complex systems can't be mathematically modeled/emulated by a computer (which can only execute computable functions on logic systems); therefore AGI is impossible for reasons of mathematics. The reviewer structures it cleverly as a series of 'Ghosts' -- Hayek (what makes a system 'complex'), Turing (the brain-isn't-a-Turing-machine computability argument, echoing Dreyfus/Penrose), Moravec (biological neurons are vastly more complex than artificial ones; we can't even model C. elegans's 302 neurons), Churchill (machines already 'rule' via opaque algorithmic mediation). Comes away unconvinced -- the strong conclusions rest on 'strong but leaky propositions'; LLMs are taking down syntax/semantics; Church-Turing-Deutsch -- but credits the complexity/computability points, and offers two Straussian readings: that the complexity critique really argues against AI alignment (we build things we don't understand), and the meta-point that nukes/bioweapons exist now, so there are opportunity costs to AI treatises.
Why this score
Quality 74 · Strong. Strong, upper band. A sophisticated, critically-engaged review that fairly reconstructs a demanding technical argument, structures it memorably, and reaches a nuanced verdict with genuine value-add (the alignment-not-impossibility Straussian read). Held below Excellent as faithful critical transmission of someone else's argument.
Claude’s paradigm shift 50 · Moderate. Moderate. The book's argument and the Dreyfus/Penrose lineage are existing; the reviewer's fresh contributions are the 'Ghosts' framing and the Straussian readings.
Real-world impact 2 · Minor. Minor. Within the AGI-feasibility discourse; conceptual, no material reach. 2.