TECH NEWS

The last mile of innovation

Unlocking technology’s full potential means looking beyond its capabilities to understand how organisations work and what holds them back.

October 29, 2026

Osemhen Okenyi, Senior Innovation Manager – PWC LUXEMBOURG
Osemhen works as a senior manager within PwC’s Innovation Team, where they accelerate the development of new products and services to address clients’ evolving needs.

 

Business leaders are being offered ever more powerful tools to accelerate work. Yet the distance between what a technology can do and what an organisation can achieve with it remains large. Understanding that distance requires looking beyond the machine.

There is an old story from the early days of electrification. When factories first gained access to electric power, many owners did something perfectly logical: they replaced their steam engines with electric motors. 

Then they waited for the much-promised productivity revolution. Except, it didn’t come. Why?

Factories had been designed around steam: a central engine drove shafts and belts connected to every machine, dictating layout and production flow. Replacing it with an electric motor while leaving everything else untouched captured only a fraction of electricity’s potential.

The real transformation came later, when manufacturers stopped asking, “How do we replace the steam engine?” and started asking, “If electricity had always existed, how would we design the factory?”

Once machines had their own motors, production lines could be reorganised around the work rather than the power source. The technology arrived decades before the organisational transformation did. 

Sound familiar? 

Walk into any large organisation today and you will hear talk about AI: pilots, copilots, hackathons, agents, proofs of concept, strategies, and committees.

 

The technology is moving quickly. But are organisations changing just as fast?

The defining competition of the next decade may not be between organisations that have AI and those that don’t. Almost everyone will have AI, just like everyone now has electricity. The more interesting competition will be between organisations that can absorb technological change and those that cannot. Imagine you suddenly have a racing car capable of more than 300 kilometres per hour. You drive into Luxembourg at 8:15 on a Tuesday morning, but there’s an accident on the A1, roadworks, and a red light ahead.

Your racing car is an extraordinary piece of engineering, but the performance of your journey is determined by the system around it. This is increasingly what AI looks like inside organisations.
The model produces an answer in seconds, but it enters an organisation where a process may require six approvals, data sits in different systems, ownership is unclear, risk must review it and technology must integrate it. Nine months later, everyone is still discussing whether to scale the pilot.

The AI was fast; the organisation was not. As powerful technology becomes democratised, the same foundation models, cloud platforms and tools are available across industries. If everyone can buy a racing car, the competitive advantage is not the car. It is the road. 

This question is particularly interesting in Luxembourg, whose economic success has been built, in part, on its remarkable ability to reinvent itself.

From steel to financial services, and more recently towards digital assets, sustainable finance, space and technology, Luxembourg has repeatedly responded to structural change.

But AI does not simply create another industry alongside existing ones. It reaches horizontally across almost every knowledge-intensive activity. Consider a fund manager, bank, or insurer. Much of its economic machinery consists of people reading documents, reconciling data, applying rules, preparing reports, and making decisions; the activities AI is beginning to augment. This creates enormous opportunity, but also a trap.

Organisations can mistake experimentation for transformation. Running twenty AI pilots feels innovative, but how many change how the organisation actually works?

This is the experimentation-to-value gap. Technology has reduced the cost of experimentation dramatically; the cost of organisational change has not fallen as quickly.

Ideas were once scarce. Now organisational attention is the constraint. Leaders need to reframe the most important question. It can no longer be “Where could we use AI?” but “Which problems are important enough to redesign the organisation around?”

Organisations should think about adaptability as athletes think about fitness. You cannot become fit on marathon morning; fitness accumulates through repeated stress and recovery.

Organisations build adaptability similarly: they notice signals, experiment, learn, decide rather than wait for the market to decide, scale what works and stop what does not.

And then they rinse and repeat.

Signal → Experiment → Evidence → Decision → Scale → Value.

Organisations that move exceptionally well through this cycle develop something valuable: institutional muscle for change. That muscle consists of mundane things: clear decision rights, accessible data, movable budgets, proportionate governance, cross-functional teams, leaders willing to stop initiatives and incentives that reward adoption rather than invention. None sounds as exciting as AI, and that’s the point. Previous technological revolutions delivered their greatest productivity gains not from installing technology, but from reorganising work around it.

There is another consequence. For years, innovation teams have been asked, “What should we build?” Well, AI makes building dramatically easier. So now the more valuable question is, “What should we bother building at all?”

An organisation capable of doing 100 things does not benefit from doing all 100. Strategy is partly the discipline of deciding which 95 not to do. As intelligence becomes cheaper, judgement becomes more valuable.

More than a century ago, the factory owners who first installed electric motors were not foolish. They were doing what organisations naturally do when confronted with new technology: fitting the future into the architecture of the past.

Today, there’s a risk of doing the same thing with AI. Putting it into an existing process may make one task faster but redesigning the process may remove the task entirely.

Adding an AI assistant may improve an employee’s productivity. Redesigning how decisions, information and expertise flow through the organisation may change its economics.
Those are different ambitions. The challenge for leaders in Luxembourg and elsewhere is not the Sisyphean task of keeping up with every new model, agent, or breakthrough. It is building organisations that can decide which changes matter and absorb them faster than competitors. 

Because in a world where increasingly powerful technology is available to almost everyone, technology itself becomes less distinctive. Racing cars will keep getting faster. More people will be able to own one.

The competitive advantage may belong to whoever redesigns the road.

 

Osemhen Okenyi
Senior Innovation Manager
PWC LUXEMBOURG
Osemhen works as a senior manager within PwC’s Innovation Team, where they accelerate the development of new products and services to address clients’ evolving needs.

 

Want to know more?
PwC Experience Center Luxembourg

 

    Watch video

    Share This Story, Choose Your Platform!

    In the same category