AI can increase productivity. It can’t increase human capacity

We may be using AI to save hours of work, only to fill those hours with more work.

This is something I have been thinking about as organisations accelerate their investment in AI.

The productivity opportunity is undeniable. Generative AI can draft in seconds, analyse large volumes of information, automate administrative work and give leaders access to capabilities that previously required significantly more time and resources.

A 2025 field experiment involving more than 7,000 knowledge workers across 66 companies found that workers who actively used generative AI spent roughly two fewer hours each week on email and reduced work outside normal hours.

Yet look at what is happening alongside these gains. In Singapore, 58 per cent of leaders said productivity needed to increase, while 81 per cent of the workforce reported lacking sufficient time or energy to do their work. Globally, Microsoft workplace data found that during core working hours, employees can be interrupted by meetings, emails or messages as frequently as every two minutes.

Something does not quite add up.

We are giving people tools that allow them to work faster, while many of those same people are telling us they are running out of time and energy.

This is why I believe we need to separate two conversations that are often treated as one.

Productivity is not the same as human capacity.

We are increasing output, but are we increasing capacity?

Productivity asks how much we can produce.

Human capacity asks how much complexity, pressure, change and decision making a person can sustainably hold while continuing to think, lead and perform well.

AI can significantly expand the first. Leadership needs to take responsibility for the second.

Consider what happens when an executive saves five hours each week through AI. If those five hours are immediately replaced by additional meetings, decisions, targets and deliverables, have we improved the way that executive works?

Or have we simply made it possible for them to carry more?

This distinction matters because human beings have cognitive, emotional and attentional limits. Under sustained pressure, our judgement changes. We can become more reactive. We listen differently. Strategic thinking becomes harder because our attention is consumed by what is immediately in front of us.

The concern I have is that organisations could become significantly more productive through AI while the people inside them become increasingly depleted.

That is not sustainable transformation.

Human performance needs to become part of AI strategy

Human performance beyond productivity deserves its own attention.

Can a leader access their best judgement when the stakes are high? Can they maintain clarity when information is coming from every direction? Can they regulate their own response to pressure while creating confidence and stability for the people around them?

These questions become more important as AI accelerates the speed at which leaders receive information and are expected to act on it.

Also Read: AI and inflation are driving Singapore’s workplace anxiety. Employers are spending on the wrong fix

Senior leaders are already navigating transformation, geopolitical uncertainty, changing workforce expectations, stakeholder pressure and difficult questions about where human judgement should sit alongside machine intelligence.

Giving those leaders more information does not automatically help them lead better.

A tired executive can still answer another email and attend another meeting. The more important question is whether that executive still has the capacity to notice what others have missed, challenge an assumption, listen properly to resistance from the organisation or make a difficult decision with clarity.

This is why I believe human performance belongs inside the AI investment conversation, alongside technology, productivity and return on investment.

Reinvest the capacity AI gives back to people

We need to stop measuring the value of AI only by how much additional work people can produce.

The 2025 field experiment gives us an interesting example. AI users reduced the time they spent on email, but there was no significant reduction in meetings.

Technology changed part of the individual’s workload, but it did not automatically change the organisation around them.

That requires leadership.

If AI gives an executive team back 10 or 20 per cent of its time, organisations have a choice about what happens to that capacity. Some of it can certainly support greater productivity, but some should be reinvested into strategic thinking, customer conversations, developing people, experimentation and simply having enough space to think before another important decision is made.

For me, this is a much more interesting measure of AI return on investment.

It is not only about getting more output from people. It is about making better use of their human capability.

Also Read: Singapore’s next AI industry may be the people who check the agents

Build human capacity alongside AI capability

The second opportunity is to deliberately develop the human skills required to perform in an AI enabled environment.

We need leaders who know how to manage their attention, not simply their calendars. Leaders who can remain grounded under pressure, separate urgency from importance and question an AI recommendation rather than accepting it because the analysis looks convincing.

We also need leaders who can create enough psychological safety for their teams to experiment, make mistakes, challenge assumptions and admit when they do not know the answer.

Most importantly, we need to move away from the idea that peak performance means operating at maximum intensity indefinitely.

Sustainable peak performance is the ability to access greater capacity when the situation requires it and to know how to recover that capacity afterwards.

The strongest leader in an AI enabled organisation may therefore not be the person capable of carrying the greatest workload. It may be the person who can retain clarity, judgement, trust, energy and connection as the speed and complexity around them increases.

What will we do with the capacity AI gives us?

AI gives organisations extraordinary access to intelligence and efficiency. I am optimistic about what that can create.

But as intelligence becomes more accessible, our deeply human capabilities become even more important: judgement, discernment, creativity, courage, trust, presence and connection.

The organisations that thrive will understand that AI capability and human performance need to grow together.

As the founder of XfinitySG, a leadership and human performance practice, I believe this is where the next phase of the AI conversation needs to go. We have spent considerable time asking how quickly organisations can adopt AI and how much productivity it can unlock.

Now we need to ask what happens to the humans working inside those faster organisations.

Technological capability can scale rapidly. Human capacity needs to be intentionally developed, protected and expanded.

So, as AI gives your organisation the ability to do more, perhaps the more important question for leaders is:

What will you do with the human capacity AI gives back to you?

—

The post AI can increase productivity. It can’t increase human capacity appeared first on e27.