We publish these three essays because the question of how to think about heuristics — fast-and-frugal rules that work in genuinely complex systems — has become urgent. Every organisation is under pressure to formalise, instrument, and optimise. Simultaneously, artificial intelligence promises to absorb more and more of the cognitive work that humans have traditionally done by intuition, pattern recognition, and compressed experience. Neither impulse is wrong. But applying them blindly produces systematic failures that could be prevented with clearer thinking.
The work here draws on Gerd Gigerenzer's concept of ecological rationality: the fit between decision rule and environment matters more than absolute optimisation. We extend that observation into clinical medicine, market strategy, financial regulation, and organisational design, asking: what makes a heuristic robust? When does it decay? How does AI change what can be safely automated and what must remain human?
These are not rhetorical questions. They have architectural implications. Start with the seven examples if you want a practical sense of how heuristics outperform models. Read the full taxonomy if you need to understand why, or the companion piece if you want to know what to do when a once-reliable rule stops working.
For a first reading: Start with When Simple Rules Beat Complex Models (seven examples). It gives you the intuition without requiring the taxonomy.
To understand the architecture: Move to When Simple Rules Beat Complex Models — Taxonomy Edition. It explains the four types of heuristics and their failure modes.
To act on this thinking: Read Knowing When to Let Go alongside your own organisational context. It addresses how to preserve heuristic wisdom as environments change, and why automation isn't always the answer.