The Game Is On

“The game is afoot.” When Sherlock Holmes speaks these words, the case has structure. A crime has been committed. The evidence is fixed. The victim does not change while Holmes examines the scene. In his mind palace, he connects the mud on the shoe to the quarry on the edge of town, the tobacco ash to a single tobacconist in Covent Garden, and deduces the killer before Scotland Yard has finished interviewing witnesses.

But Holmes was never just the deductive thinker of popular imagination. He tested poisons on himself. He fired bullets into walls to study impact patterns. He spent hours in his chemistry lab running experiments before a case even began. His real method was a loop: deduce a hypothesis from what you know, advance by testing it against reality, then induce new understanding from what you observe. Anticipate. Advance. Assess. Holmes did not solve cases by thinking harder. He solved them by looping faster.

The Loop

Enterprises love the deductive part. Analyze the market. Study the competition. Build a strategy from first principles. Deduce the right answer. The best enterprises also use Induction: observing patterns in data, customer behavior, and past performance to draw general conclusions. Both methods are essential. Both are insufficient on their own.

Deduction gives you certainty, but only if your premises hold. In a complex environment, they rarely hold for long. Induction gives you patterns, but as David Hume warned, past patterns do not guarantee future results. The market you observed yesterday is not the market you face tomorrow. The answer is Empiricism: the loop. Empirical Control combines both into a continuous cycle. The loop is not optional. It is the only method that works when the crime scene keeps moving.

But running the loop is not enough. Holmes examined every detail of a crime scene. A thread on the victim’s coat. A scratch on the window frame. A dog that did not bark. He never optimized for a single clue. He optimized for the whole picture. Overall Optimization demands the same discipline. When departments optimize for their own metrics, when Teams optimize for velocity, when individuals optimize for performance reviews, the enterprise works against itself. See the whole scene, not just the clue in front of you.

Here is where the metaphor breaks. A crime scene is static. Your enterprise is not a dead body on the floor of a London flat. It is a living system that grows, adapts, reorganizes, and creates products that did not exist when you started your investigation. A living system does not need a detective. It needs a doctrine that guides its own evolution. This is Evolution Focus. What worked in Genesis will fail in Commodity. What stabilizes a mature product will suffocate an emerging one. Your strategy must evolve with your enterprise.

Three doctrines. One system. Together they form a self-improving engine: learn through empirical loops, ensure improvements serve the whole, and direct energy where evolution demands it. This book gave you the Rules, the Playbook, and the Interplay. But a book is not a strategy. It is a hypothesis. Test it against your reality.

The Limit

Now consider a tempting idea. What if we skip the loop entirely? AI can model weather systems, simulate climate trajectories, and predict protein structures. Surely it can model your enterprise, your market, your competitors. Feed enough data into a powerful enough model, and the future becomes visible.

It cannot. And the reason is fundamental.

In his book Sapiens, Yuval Noah Harari distinguishes two kinds of systems. A weather system is what he calls level-one: it does not react to predictions about itself. Water drops have no self-awareness. Build a better model, and your forecasts improve. This is why AI excels at climate simulation, materials science, and drug discovery. The molecules do not read the forecast.

Markets, organizations, and competitors are different. They are level-two systems: systems that react to predictions about themselves. In the language of Cynefin and the Stacey Matrix, these are complex systems: cause and effect can only be understood in retrospect, never reliably predicted in advance. You map your competitor’s value chain and predict they will enter your market segment next quarter. They are mapping yours. Both of you act on predictions about each other. The landscape shifts before either prediction materializes.

You deploy AI agents to scan markets, analyze supply chains, and spot patterns no human could see at that speed. For a while, you are ahead. But your competitor will deploy the same agents. Not today, perhaps. But soon. And if they do not, a new competitor will, one that was born with AI in its operating model and never knew the old way of working.

But this does not mean prediction is useless. A better map is still a better map. AI-Enhanced Teams give you sharper sight: they process more signals, test more hypotheses, and update the map faster than any team could before. A deeper understanding of where your products sit on the evolution path gives you a real edge. You see the move before your competitor does. You score the goal. The point is that the edge is temporary. Your competitor will build the same models, read the same signals, and close the gap. In football, being one goal ahead means nothing if you stop playing.

This is why Empirical Control is not a one-time analysis. It is a continuous loop. The enterprise that maps better, experiments faster, and reassesses more honestly will stay ahead. Not because it can predict the future, but because it learns faster than the future changes. The goal is not perfect prediction. The goal is to be one loop ahead.

The future is not written. It is played.

The Leap

Life works against entropy. It takes energy and creates order: cells, organisms, societies, enterprises. Every structure we build requires energy to maintain. Stop investing, and it dissolves. This is not a metaphor. It is thermodynamics.

Every technology humanity has adopted demanded more energy, not less. Steam engines consumed more coal, not less. Electrification consumed more power, not less. Computing, the internet, and now AI follow the same pattern. Humans consume energy. The tools we build consume energy. Together, we push the boundary of complexity outward with every generation.

This creates a real contradiction. The climate demands that we reduce energy consumption. AI demands that we increase it. Growth demands that we increase it. Human ambition demands that we increase it. Strong strategic leadership does not pretend this contradiction will resolve itself. It navigates the tension, knowingly, empirically, one Tournament at a time.

But this is not just thermodynamics. Behind every product shipped, every Arena launched, every Goal pursued, there are people who chose to build something together. The energy that sustains an enterprise is not only capital and electricity. It is conviction. The work matters. The Team matters. Creating value for others is worth the effort. Enterprises that lose this conviction dissolve, no matter how well-funded they are. Enterprises that hold onto it can navigate any disruption, because the people inside them choose to keep building.

And leadership is not about making this decision for others. Leadership is leading others in making better decisions. You cannot not communicate, and in an enterprise, you cannot not lead. Every Team member leads, whether they intend to or not. When a Team acts with discipline and transparency, it becomes a model for other Teams. When an Arena evolves its products and work system empirically, it convinces the Enterprise. When an Enterprise outevolves its competitors, it demonstrates to the world that there is a better way to organize.

AME3 is the Adaptive Metaframework for Empirical Enterprise Evolution. Not a prescription. A metaframework for building your own unique operating model: one that evolves with your enterprise, your markets, and the forces shaping them. The Rules give you the structure. The Strategy gives you the doctrines. The methods and frameworks you choose give you the tactics. What you build from these is yours.

Andy and I continue this work at ame3.ai. New articles, new methods, new insights from enterprises applying these ideas. The framework evolves because the enterprises using it evolve. Join us. Challenge us. Share what you learn.

The question was never “Are we doing AI?” or “Are we agile?” The question is: how well does your enterprise navigate evolution compared to your competitors?

The game is on. Outevolve.