TB: A New Kind of Science (Wolfram)
Core Thesis
Complex, unpredictable behavior can emerge from simple programs and simple initial conditions. Wolfram's central claim: this computational perspective reveals new principles about nature, randomness, and the limits of knowledge that traditional mathematics misses. (John's annotations are often skeptical of the book's overreach, but the core ideas on emergence, universality, and incompleteness are worth preserving.)
Key Takeaways
Complexity from simple rules
- You can get complicated behavior from simple programs and simple initial conditions — complexity doesn't require complex rules.
- The threshold of complexity required to produce interesting behavior is typically extremely low.
- Adding more dimensions does not ultimately have much effect on whether significant complexity occurs.
Three sources of randomness
- Randomness explicitly introduced into the underlying rules.
- Random initial conditions with deterministic rules for subsequent evolution.
- Randomness produced intrinsically by simple programs through iteration — without external randomness at all.
- When intrinsic randomness dominates, some level of repeatability in "random" behavior is realistic to expect.
Universality and initial conditions
- Universality is far more widespread than previously assumed — not a rare property of specially constructed systems.
- Once past the threshold of universality, the set of computations that can be performed is always exactly the same — more complex rules add nothing fundamental.
- Nothing fundamental is gained by rules more complicated than those for the universal cellular automaton; more complex rules can always be emulated by initial conditions.
Natural selection and complexity
- Natural selection can only efficiently work on simple features of an organism; it cannot easily reverse-engineer programs that lead to specific complex behavior.
- Human artifacts are biased by the fact that they must be simple enough that we know what they'll do; nature operates under no such constraint.
Incompleteness
- Gödel's incompleteness: any axiom system will have statements that are true but unprovable; there must be integer equations with no solutions where this fact cannot be proved from normal arithmetic axioms.
- The moment a question about whether a string can be reached is undecidable, there must be either incompleteness or inconsistency.
Memory and thought
- Memory underlies almost every aspect of human thinking; generalization, analogy, and intuition are closely related to retrieving data on the basis of similarity.
Mental Models
- Complex Systems: Features from Path, Not Design — behavior emerges from initial conditions plus simple rules; the path through state space is what matters
- Inversion — to understand complexity, study the simplest systems and observe emergence, rather than designing complexity in