What is coherence, and why it governs your AI decisions

Coherence is the alignment between your thinking mind and your settled body: the clear, regulated state from which you read a situation accurately and decide well. It governs your AI decisions because every AI call is a judgement call, and judgement runs on the state of the person deciding. Buy the best model in the market and a depleted, scattered leader will still steer it into the wrong problem. The technology has caught up. The question now is the quality of the hand on the wheel.

Ask any leader I work with about their best decisions. They rarely describe the data. They describe a feeling of clarity. They knew. The room went quiet inside them and the right move was obvious. Then ask about their worst calls, and a pattern appears. Tired. Rushed. Pulled in nine directions. Same person, same intelligence, two very different states, two very different outcomes. That gap has a name, and the name is coherence.

What is coherence, and how is it different from intelligence?

Intelligence is the horsepower. Coherence is whether the engine is tuned. You can have a brilliant leader running incoherent, and the brilliance leaks out as second-guessing, reactivity and decisions that contradict last week's decisions. Coherence is the state where head and heart point the same way, where the analytical mind and the body's quieter signal agree rather than argue. In that state the leader sees more, reacts less, and chooses from clarity rather than pressure.

This matters more now than ever, because the centre of gravity in business has moved. It helps to name the arc. Phase One, the Age of Effort: work hard, get a little more, linear growth. Phase Two, the Age of Scale: build once, sell to millions, exponential growth. Phase Three, the Age of Acceleration: output decoupled from human effort almost entirely, the phase AI unlocks. Phase One was muscle. Phase Two was machine. Phase Three is mind. The work that remains uniquely human is judgement: what to build, what to ignore, what a result actually means, where to point a capability that can do almost anything. The bottleneck is no longer the technology. It is the human operating system (the way a leader thinks, feels and decides) choosing what the technology serves.

Why does coherence govern AI decisions specifically?

Because AI value is decided by people, not algorithms. BCG's 2025 research is precise about this: in a successful AI transformation, roughly 70% of the value comes from people and process, about 20% from data and technology, and only around 10% from the algorithms themselves. The models are largely commodity now. The differentiator is the quality of the decisions made around them, and humans make those decisions in a particular state.

Fresh survey work sharpens the point. In a 2025 study of 1,400 US employees, reported in Harvard Business Review (the management journal published by Harvard Business School) and tied to Columbia Business School, 76% of executives believed their people were enthusiastic about adopting AI; only 31% of individual contributors actually were, a 45-point gap. Leaders are reading the room wrong. The lag sits at the top, in the layer doing the deciding. That is a coherence question before it is a capability question.

What does the science actually say about state and decisions?

Two bodies of evidence are worth separating, because one is settled and one is emerging, and a careful reader deserves both labelled clearly.

The settled one is decision fatigue. Across decades of work, the finding holds: as the load of decisions accumulates through a day, judgement degrades, shifting from deliberate, analytical thinking toward fast, status-quo choices. The well-known study of Israeli parole judges (which tracked the share of favourable rulings across the working day) found rulings became measurably less considered as cognitive resources drained, recovering after a break. The depleted mind reaches for the easy answer. For a leader making capital-grade AI calls at 5pm after a day of meetings, that carries a direct and expensive cost.

The emerging one is the physiology of the regulated state. Peer-reviewed autonomic research links higher vagally-mediated heart-rate variability, a marker of a well-regulated nervous system, to stronger executive function and better decisions under risk and uncertainty, through stronger prefrontal control. A 2022 meta-analysis in the journal Cortex (Magnon and colleagues) pooled the studies and found a small but reliable positive association between this measure and executive function (r = 0.19). A separate systematic review the same year reached the same direction for decision-making specifically. The more expansive claims about heart-brain coherence remain exploratory, and I hold them lightly. The boardroom version, though, stands firm: a regulated, coherent leader makes better calls than a depleted one. The data simply names what every experienced leader already feels.

How do you build more coherence as a leader?

You treat your own state as infrastructure, because that is what it is. Coherence is trainable, and the practical work is simpler than the science. Build it in this order:

1. Protect the decisions that matter most. Schedule your highest-stakes AI calls for when you are fresh, and let the routine ones happen later. You are matching the weight of the decision to the quality of the state available to make it.
2. Regulate before you decide. Two minutes of slow, even breathing settles the nervous system and lifts the very physiology the research links to clearer judgement. A coherent five minutes beats a frantic hour.
3. Build recovery into the rhythm. Capacity is the asset. Real rest, real breaks, and clear boundaries keep the prefrontal system working for you rather than draining toward the easy answer by mid-afternoon.
4. Separate the signal from the noise. When head and heart disagree, pause and listen to both. The clarity you are after lives where the two align, and it rewards waiting a beat to find it.
5. Upgrade yourself before the next tool. Build the leader's capacity to stay clear under load first, then buy the next capability. Most leaders are trying to install new software on broken hardware. The leaders who win the next decade are the ones who upgrade themselves first.

Here is the frame to carry out of this. Your AI strategy does not run on the model. It runs on you, on the state from which you read the field and choose. Tune that, and every decision downstream gets cleaner. The technology is ready and waiting. The opportunity now is to meet it as the clearest version of yourself.

About the author

Thomas Green

British technology futurist, AI keynote speaker and advisor. Thirty years across enterprise technology and AI strategy, helping leaders navigate the future of work. The futurist who died.