The Paradox of Progress

The responsibility of leadership is to create the conditions in which people can achieve extraordinary outcomes without sacrificing their capacity to sustain them. Hold two competing truths. Do not pretend one of them is optional.

There is a peculiar contradiction at the heart of modern leadership.

We have never had more powerful tools. AI accelerates analysis, ideas, problem-solving and execution. Weeks become days; hours become minutes.

Yet I increasingly ask a question technology alone cannot answer.

Just because we can move faster, does it mean our people can?

We are expected to drive transformation, deliver results, challenge convention and raise the bar, and to create environments where people feel safe, can think clearly, challenge decisions and grow.

These can feel contradictory. Perhaps the problem is not the contradiction, but our assumption that good leadership must eliminate it.

The first paradox: accelerating change while protecting human capacity

Transformation creates its own momentum. One initiative reveals another. A pilot becomes a programme; a programme a portfolio. Soon the organisation pursues more than people can deliver.

Especially with AI: technology increases what we can produce; it does not automatically increase our ability to absorb change, decide well, or manage operational consequences.

There is a difference between the speed at which technology generates an answer and the speed at which an organisation can responsibly act on it.

Every initiative competes for finite attention, experience, judgment and energy. Capacity is not just headcount. It is cognitive space, emotional energy, and organisational breathing room to learn.

When we overload teams, output may not fall immediately. People compensate: longer hours, deferred conversations, shortcuts, quiet absorption of pressure. Consequences arrive later: quality, slower decisions, creativity, loss of experienced people.

A team that delivers through extraordinary effort is not necessarily a high-performing system. It may be consuming future capacity to meet today's commitments.

The answer is not lower ambition. It is more discipline about what transformation we pursue.

Every new priority forces a conversation about existing commitments. Every acceleration needs to name what it displaces. Every tech investment must weigh efficiency against the change demanded of the people who will use it.

Sometimes the most valuable leadership decision is not approving the next exciting initiative. It is deciding what we will stop so the work that matters can succeed.

The second paradox: empowering people while remaining accountable

Complex organisations cannot route every important decision through senior leaders. Speed requires decisions close to the work. People closest to operational problems often understand constraints better than those several levels removed.

If I want to empower teams, I must give them authority to use that knowledge. But I remain accountable for outcomes. Accountability does not disappear when authority is delegated.

If I approve every consequential decision personally, I undermine the empowerment I claim to value.

Empowerment is less the distribution of tasks than the deliberate distribution of decision-making authority.

Clarity first: outcome; constraints; decisions the team owns; escalation triggers; support when things go wrong.

Without that, empowerment becomes responsibility without authority, resources or support.

The behaviours that got us here can constrain us later. Reassurance of being involved in everything can feel productive; every unnecessary decision I retain denies someone else the chance to develop judgment. Intervene too early and I teach that authority is conditional on my agreement.

Measure of leadership: not how many decisions I make, but how effectively the organisation decides without my constant involvement.

Be deliberate about decisions that genuinely need me, and accept responsibility for the environment in which all other decisions are made.

The third paradox: creating safety while demanding performance

Caring for people and demanding high performance are not opposites.

In complex transformation, speaking openly about uncertainty, challenging assumptions and raising concerns is essential to ambitious outcomes.

People must be able to say a deadline is unrealistic, a technology is not ready, or a risk is underestimated, and know that raising a concern starts a conversation about responsibility, not the end of one.

Psychological safety cannot mean optional commitments. Accountability cannot mean punishment for inconvenient truths or honest mistakes.

Disagree openly before a decision; understand why it was made; then own execution.

Mistakes: learn. Repeated negligence: address.

The person closest to the problem must feel confident enough to stop an unsafe change, regardless of who advocates for it.

The test is how I respond when someone tells me something I do not want to hear under delivery pressure.

Safety and performance should reinforce each other.

The fourth paradox: using AI without surrendering human intelligence

Faster output is not automatically better outcomes.

Human judgment will be a defining capability of successful AI-enabled organisations.

AI can pattern, model, optimise and automate. It cannot assume human accountability for consequences.

When an AI-enabled decision affects an employee, customer or critical operation, someone must understand context, question assumptions and own the outcome.

If AI does the analysis, recommendation and proposed decision, where does the next generation learn to recognise when the recommendation is wrong?

Not an argument against automation. An argument for designing human development into automation.

Use AI to free time for judgment, creativity, collaboration and complex problem-solving, not only to increase work volume.

Design technology that helps people think better, not just produce faster.

Invest as deliberately in people's capabilities as in the technology meant to transform their work.

Measure success by what people become capable of, not only by how much work we automate.

The responsibility to choose

We cannot pursue every opportunity, eliminate every risk, give complete autonomy while guaranteeing every outcome, or raise demands without considering cost.

Leadership makes tensions visible, invites challenge, and creates enough clarity to act.

Sometimes accelerate. Sometimes stop what consumes disproportionate energy. Sometimes let a team decide differently from how I would. Sometimes intervene because consequences exceed their authority or experience.

No perfect formula. Quality of leadership shows in how consistently we make these choices when both demands are legitimate and answers are uncomfortable.

What I want my leadership to stand for

Ambitious without reckless. Empowered without abandoned. Safe enough to challenge; accountable enough to deliver.

Challenge people to achieve what they may not yet believe possible, and challenge the organisation when ambitions exceed capacity.

Technology that amplifies human potential rather than only raising expectations.

Leaders who make difficult decisions without losing sight of the people affected.

The faster the world moves, the more deliberate we must become about how we lead.

Urgency without impatience. Decisive without inflexible. Accountable while giving confidence and authority to act.

People delivering transformation are not merely resources required to achieve it. They are the reason transformation matters.

True measure of progress: not only what we deliver today, but the capability, confidence and capacity we leave people with to deliver tomorrow.

Further reading

• Both/And Thinking, Wendy Smith and Marianne Lewis (2022)
• The Fearless Organization, Amy Edmondson (2018)
• Microsoft Research, Rethinking AI in Knowledge Work