CORE IDEA: By moving from activity leverage to cognitive leverage, we can finally treat that time as strategic capital.

AI is giving us time back. A report that once occupied an afternoon can now be drafted in minutes. Research that once required several people can be accelerated by one person with the right tools. Analyses, presentations and proposals can be produced at a speed that would have seemed improbable only a few years ago. Business Intelligence can now operate at greater speed and scale—and that is an exceptional development.

AI is making people, organisations and society as a whole more productive. As it becomes increasingly integral to how we work, we are completing tasks faster and making better use of information. That efficiency is giving us more time—but it also raises a question: must every hour we recover be used to produce more?

Until recently, I probably would have accepted that without much hesitation. Now I wonder whether automatically turning every recovered hour into additional output is the least imaginative way to use what AI is giving us.

If a task that once took four hours can be completed in 30 minutes, must the remaining three and a half hours immediately be filled? What if the time itself is part of the return?

The empty hull

Many years ago, I raced 30-foot sailing boats as a hobby. I was fortunate to sail alongside very experienced people, including skippers who had competed at the Olympic level in Britain. One remark a skipper made has always stayed with me. He told me, “The best racing boats have an almost empty hull.”

At first, I assumed the explanation was simply weight. A lighter boat moves faster. But the reasoning was subtler. Everything placed inside the boat became something that had to be secured, maintained, or considered. In difficult conditions, an unnecessary object was not merely additional weight—it was another demand on the crew’s attention.

An almost empty hull allowed us to concentrate on what matters: the sails, the wind, the course, and what the boat was telling us. The empty space was not wasted. It was part of what made the boat manageable. Time racing was optimized to manage the parts that actually needed managing—that’s what the skipper really meant.

I often think about this when looking at organisations. Companies are exceptionally good at adding things. Most of these additions probably made sense when they were introduced. The difficulty is that organisations are generally much better at adding than subtracting. Gradually, the hull fills.

AI could help us empty it—eliminating repetitive work, streamlining processes, and slashing the time spent hunting for information; or, it could help us fill the hull even faster. When free time is automatically swallowed by more output just to justify headcount, efficiency turns into a trap. At that point, the paradox of the “Japanese salaryman” stops being a joke and becomes a warning.

The productivity reflex

Imagine that a team uses AI to reduce the preparation of a weekly report from four hours to 30 minutes. What happens next? The report might become longer. It may be produced more frequently. Someone may request additional charts or correlations because these are now inexpensive to generate.

Before long, preparing and reviewing the report may once again consume four hours. The only difference is that it contains more information. The time has not been recovered. It has been absorbed.

This is what I think of as the productivity reflex. Whenever technology creates spare capacity, organisations tend to find “something else” to place inside it.

That reaction is understandable. Additional output is visible. Recovered time is not. A new report can be circulated and presented. An hour spent thinking is more difficult to display on a performance dashboard.

Ask why three times

Whenever I get bogged down in the mechanics of a problem—particularly when trying to figure out how to spend my time—I lean on a simple mental exercise: I ask “why” at least three times in a row.

There is nothing magical about the number three. What matters is how the nature of the answers evolves.

Take a very typical corporate request I've seen before:

• Why do we need a new dashboard? Because management needs better visibility into customer retention.
• Why does management need that visibility? Because different teams respond inconsistently when key accounts show signs of leaving.
• And why do they report inconsistently? Because each team uses its own metrics to track customer retention.

In my experience, the problem shifted from building a tool to managing alignment. The real task here wasn't data—it was getting everyone to measure the objective using the same ruler.

By the third “why,” I stopped chasing surface-level symptoms and hit the root principle driving the work.

That realization completely changes what happens next. Instead of blindly building a solution, I can ask what decisions actually need to improve, what actions should be taken, and what the minimum viable information is to trigger them.

For me, this exercise isn't a framework for evaluating corporate governance. It's just a quick reality check on whether an activity—meaningful or trivial—actually deserves my time.

The bureaucracy tax

Bureaucracy is often discussed as though it were inherently bad. I don't think that is a useful way to look at it. Organisations need governance. They need controls, accountability, reporting, and defined decision rights. A bank, hospital, or enterprise cannot operate purely through improvisation. These structures impose a cost, but that cost is often necessary and entirely justified. I think of bureaucracy as a tax paid for governance.

The question is never whether the tax should exist. It is whether the organisation is receiving sufficient value in return.

Purposeful bureaucracy protects customers, manages risks, and makes responsibilities clear. But it becomes ruinously expensive when activities lose their connection to those core purposes—and nowhere is that more evident than when an organisation turns its capabilities inward.

I have always had considerable respect for good marketing people. Some of the smartest professionals I have encountered operate close to the psychology of markets, understanding perception, motivation, and the power of presentation. However, problems often arise when companies point that inward, encouraging people to market themselves to themselves.

Projects need to look important. Departments need to demonstrate impact. Investments require numbers that suggest progress. When people find themselves with newly available capacity, they often feel compelled to prove their productivity by creating more visible output. This is where shallow vanity enters.

Another analysis is produced because it looks sophisticated. Another metric appears simply because it can be measured. Another AI project is launched because the technology is fashionable. The work creates visibility, but its relationship to actual purpose remains unclear.

Shallow vanity raises the bureaucracy tax considerably. When people use recovered time to create superficial work, the organisation eventually has to review, explain, distribute, and govern it. A vanity project rarely stays isolated; it generates meetings, metrics, slide decks, and reporting obligations.

AI makes such vanity dangerously easy to sustain. Almost anything can now be analyzed, correlated, and presented professionally. But a statistical correlation is not causation—it can be fascinating and managerially useless at the exact same time. A polished dashboard does not guarantee a better decision.

The question I increasingly ask myself is not, “Can we measure this?” It is, “What would we do differently if we knew?” If the honest answer is “probably nothing” the measurement is consuming time rather than creating value.

Without discipline, AI will simply give us time with one hand while vanity takes it back with the other. The result is a peculiar kind of progress: we don't get rid of bureaucracy; we just learn how to manufacture it faster.

Time as strategic capital

For much of my career, I thought about productivity conventionally. If technology allowed us to complete something faster, the sensible response was to use the saved time to do more. I am less certain now.

We have spent decades optimizing activity leverage—using technology to do more things, faster. What AI demands of us now is cognitive leverage: the courage to use saved time not to multiply our output, but to radically simplify our focus.

Some of the most valuable work I have witnessed did not initially look productive. It occurred when someone had enough mental space to notice that an assumption was wrong, investigate unusual customer behavior, or follow an idea without knowing where it might lead.

Research requires funding. Innovation requires talent. Both also require something that rarely appears on a balance sheet: uncommitted time.

An organisation in which every person, budget, and hour is 100% allocated may appear impressively efficient. In reality, it has zero capacity to absorb change or imagine its future. When I operate at full capacity, I find I have neither the time nor the resources to handle the unexpected. This is precisely why hustle culture has never worked for me.

This is why, after studying AI for years, I have started to think of free time not as unused capacity, but as strategic capital.

Like financial capital, time can be invested—or it can be wasted on vanity matters, as I noted earlier.

Treated as an asset, free time can be used to explore a new technology, challenge an entrenched process, talk directly with a customer, or pursue an unexpected question. Much of this exploration will produce nothing immediately useful. That is unavoidable. Curiosity does not arrive with a business case attached.

Occasionally, however, a question becomes an idea. An idea becomes an experiment. And an experiment reveals a product, a market, or a way of working that nobody anticipated.

The time AI returns to us could fund that exact kind of innovation—if we choose to spend it wisely rather than letting bureaucracy swallow it whole.

Navigation without occupation

Protecting time does not mean abandoning objectives or measurement. Uncommitted capacity still benefits from direction. This is one reason I continue to value simple, well-designed OKRs.

A useful objective should be expressed clearly, in a few words, what an organization, team, or individual is trying to achieve. Its key results should provide simple, observable signals showing whether things are moving in that direction.

They should be easy to understand, easy to communicate, and easy to measure.

This takes me back to sailing: a racing boat rarely reaches its destination in a straight line. The wind shifts. Currents intervene. The crew adjusts the sails and alters course. Sometimes, the boat must temporarily tack away from its destination just to make headway against the wind. That is not inefficiency. It is navigation.

Good OKRs serve the exact same purpose. They show whether our experiments, decisions, and adjustments are actually bringing us closer to our objective. They should not, however, attempt to account for every hour or turn every activity into a performance indicator. If they do, they risk becoming another form of bureaucracy—consuming the very mental space they were meant to protect.

Protecting the flame

AI is giving us time back. What happens to that time will become a management decision, whether organisations recognise it or not. We can use every recovered hour to generate more innovation, or more bureaucracy—that will be up to us.

Organizations aren't going to have the luxury of remaining neutral; if leadership doesn't intentionally choose to protect and invest that recovered time, the default corporate gravity will automatically swallow it up with noise and busywork.

AI may not invent the next great product or business model by itself. What it can do is return some of the time, attention, and mental space to the people, who are the ones who make new ideas grow.

I still remember that almost empty racing boat. We didn’t fill the hull simply because space was available. The emptiness helped us remain attentive, responsive, and prepared for changing conditions.

AI is beginning to give us some of that space back. In my view, that is a very good thing. We should be careful not to let any blind temptation for vanity or bureaucracy take it away again.