The Second Law of Thermodynamics Is a Selection Effect

Why does entropy increase? The standard answer invokes the Past Hypothesis: entropy was low at the Big Bang, and the rest follows statistically. But why was it low? Penrose puts the odds at 1 in 10^(10^123). Albert treats the Past Hypothesis as an unexplained axiom. This paper argues the question dissolves under algorithmic probability. Low-entropy states have low Kolmogorov complexity: "uniform gas at temperature T in region R" is a short description. Mid-entropy states peak in complexity: irregular configurations with no compressible pattern. The complexity curve is non-monotonic. Observers find themselves in entropy-increasing histories for the same reason they find themselves in universes with simple laws: the Simplicity Dominance Principle. The Past Hypothesis is a theorem, not a postulate.