ToKCast Ep 259: Reaction to Michael Levin Part 3: Bait and Switch, Podcastistan and Surprising(?) algorithms.
9 snips
Mar 11, 2026 
Michael Levin, biologist known for work on morphogenesis and bioelectricity, discusses his shift to studying sorting-algorithm experiments and the idea of polycomputing. He describes surprising behaviors in simple sorting systems and how local, cell-like implementations reveal emergent secondary computations. Conversation covers methodological choices, added nondeterminism, and debate over whether these findings reflect classic algorithms or modified automata.
AI Snips
Chapters
Books
Transcript
Episode notes
Levin Claims Minimal Sorts Hide Unexpected Behaviour
- Michael Levin argues simple deterministic sorting algorithms can hide unexpected behaviours beyond their explicit task.
- Levin claims observers can find “intrinsic motivations” or side quests when watching minimal six-line sorts closely, despite visible code.
Classic Bubble Sort Is Fully Top Down And Predictable
- Brett Hall demonstrates classical bubble sort step-by-step to show how top-down read-head control deterministically sorts an array.
- The example (5,3,1,4,2) traces pairwise comparisons and swaps to a final 1,2,3,4,5 to illustrate predictability.
Paper Changes Bubble Sort Into A Different Cell Model
- Brett Hall identifies a bait-and-switch: Levin frames his work as studying classical sorts but actually alters them into a cell-centric model.
- Levin's paper explicitly says it breaks top-down control and introduces damaged/frozen elements, changing the algorithm's locus of control.
