The Horizon of Foresight: Predictive Decay (Axiom A7)
Published: February 21, 2026
Read time: 10 minutes
Category: Axioms Series
Weather forecasters can predict tomorrow’s temperature with 95% accuracy. Next week drops to 70%. Next month hovers around 50%—barely better than historical averages. This isn’t failure. It’s atmospheric physics. The chaotic dynamics of weather impose fundamental limits on how far ahead useful prediction can extend. The same principle applies to social systems—and Axiom A7 quantifies exactly how our foresight fades.
Exponential Decay
The mathematics is stark: 𝒜(t) = 𝒜₀ · e^(−λt). Accuracy doesn’t decline linearly—it collapses exponentially. For social systems, prediction accuracy halves every 1.5-3.5 years. After a few half-lives, forecasting becomes indistinguishable from guessing. This isn’t a limitation of our models or data. It’s structural.
In This Post
The full article explores:
- Why exponential, not linear: How cascading uncertainties multiply through causal chains
- Empirical decay rates: λ values for weather, economics, and political systems
- The precision-horizon trade-off: You can have one or the other, not both
- What survives at long horizons: Demographics, physical constraints, cyclical patterns
- Calibration: Matching stated confidence to actual predictive power
- Arab Spring application: What was predictable at 5 years, 2 years, 6 months before
Understanding temporal limits is essential for honest forecasting.
📖 Read the full article:
👉 The Horizon of Foresight: Predictive Decay (Axiom A7)
💬 Join the discussion:
Comments welcome on Substack
📧 Subscribe for weekly updates:
Computational Macrohistory Bulletin
Previous: The Prediction Paradox: Limited Reflexivity (Axiom A6)
Next: Axiom A8: Computational Falsifiability