Why are simple products beautiful? Because they signal evolvability: the ability to change and improve without breaking.
The ability to evolve is the most important property of any system. This applies to companies, products, governments, and species. It determines how fast a startup can kill features and pivot to serve a new fast-growing demand, a state can delete bad regulation, oust a bad leader, etc.
In product design, this implies that every additional feature can become another constraint on rapid improvement.
I first realized this when reading The Cost of Bad Rulers: Extending Popper’s Criterion by Sam Kuypers. For context, Karl Popper’s criterion of a good democratic system was that it can get rid of bad rulers and rules without violence. Sam pointed out that even non-violent removal can be costly and proposed that
Good political systems minimise the cost of replacing bad rulers and decision-makers
David Deutsch first laid out the theory that knowledge evolves by evolution where conjecture and criticism are analogous to variation and selection. Companies are potentially the most impactful type of organism that originated in the last millennium (the first company was formed in the 1600s) and evolution is the theory to understand their existence.
Evolution requires variation but also selection. In evolution of knowledge, selection happens through criticism. In engineering, it often happens through deletion. The world’s richest man, Elon Musk is religious about this:
The most common error of a smart engineer is to optimize a thing that should not exist... First, make your requirements less dumb... Then try very hard to delete the part or process. If you’re not occasionally adding things back in, you’re not deleting enough.
Deletion does not merely remove waste. It improves evolvability. Every non-crucial part of a process becomes evolutionary drag: another dependency to understand, another approval to get, another assumption to revisit.
In biology, the genotype-phenotype map describes how changes in underlying genetic code (genotype) become observable traits (phenotype). Companies have the same problem: how do changes in structure become useful behavior?
An unproductive company can mutate, but every mutation affects the whole organism. For example, every improvement needs approval from four people and gets lost in compliance instead of creating customer value. A productive company is modular enough that small changes produce local, measurable improvements. A single engineer or pod owns solving a single problem.
For a state, the question is how costly it is to remove a bad regulation. The cost is not just the salaries of regulators, lobbyists, and lawyers. It is also the economic damage caused while the regulation exists: businesses not started, treatments not researched, drugs not developed, diseases not cured, love not experienced, planets not explored.
In my earlier post “Transaction Cost and Risk Elimination,” I argued that modular organization design reduces coordination costs and risk. The evolvability analogy explains why: modular systems allow parts to be criticized, improved, or replaced without redesigning the whole system.
These are two different ways to improve the same mapping between change and useful behavior. Deletion removes noise from the system. Modularity makes the remaining system easier to evolve.
Sam Kuypers, “The Cost of Bad Rulers,” for the Popper criterion extension.
David Deutsch, “Q&A: David Deutsch,” for the idea that beauty is objective.
Geoffrey West, Scale, for the analogy between organisms, cities, economies, and companies.
My earlier posts,
Transaction Cost and Risk Elimination: Lessons from Thiel, Rabois, Bezos, and Khosla
“Open-endedness in Society,” on curiosity, novelty search, and unexpected stepping stones.
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.