The Learning Gap: AI Is Automating Work. Who Is Building Experience?

What if the work we're automating is the very work that teaches people how to become experts?

It's an uncomfortable question. And after ASPIRE's Man vs Machine conference in Kraków, it feels particularly relevant.

We're getting better at making work faster. But are we getting equally good at developing the people who will lead that work tomorrow?

Welcome back to The Shortlist from Belvedere.

In Issue #20, we asked whether organisations should automate everything technology is capable of doing. This week, we're looking at an important consequence of those decisions: what happens to learning when the work itself begins to disappear?

The task nobody wanted may have been the training

Think about your first professional role.

There were probably tasks you couldn't wait to stop doing. Preparing reports. Checking data. Researching information. Fixing relatively simple problems.

At the time, much of it felt repetitive.

But something else was happening.

You were learning how the business operated. You were discovering what good work looked like, recognising patterns, making mistakes and understanding why experienced colleagues made certain decisions.

You weren't simply completing tasks.

You were developing judgement.

Now imagine removing much of that work before the next generation gets the opportunity to experience it.

We've saved time. But what have we lost?

A task can have relatively little immediate commercial value while providing considerable developmental value.

And that distinction is becoming increasingly important as AI changes the nature of entry-level work.

A conversation Kraków cannot afford to ignore

Last week's ASPIRE Man vs Machine conference brought together leaders from across the technology and global business services ecosystem to consider how people, businesses and the region itself should respond to AI.

The discussions also exposed a challenge that extends well beyond technology.

How do we build the next generation of capability when the work through which people traditionally gained experience is changing?

This question was reflected particularly well in the conference programme.

Eoghan O'Dwyer and Jacek Drabik brought the next generation into focus through The Vanishing Ladder: Aspiration, Ambition and the Next Generation, alongside student voices.

Gerard Vries , Edésio Soares Santana , and their fellow contributors explored the challenge of upskilling.

Meanwhile, the War Room's discussion on universities as innovation partners reflected another important shift: businesses and educational institutions have an opportunity to develop new forms of collaboration.

These are interconnected questions.

Employers cannot simply demand more experienced talent. Universities cannot prepare students for every emerging technology in isolation. And junior professionals cannot develop expertise without meaningful opportunities to practise.

The answer requires a different approach to developing capability.

Our workforce intelligence reveals another side of the problem

At ASPIRE, Belvedere launched the South Poland Workforce Capability Index 2026, powered by Horsefly Analytics and interpreted by Belvedere Intelligence.

The findings offer an interesting perspective on the region's specialist talent market.

Across three measured catchments, the research identified 8,554 GenAI candidate records. In the measured Kraków and Wrocław catchments, 74% and 70% of candidates had eight or more years of experience, respectively.

Only 4% and 5% were classified as early-career.

These figures relate specifically to the report's measured GenAI role family. They should not be interpreted as the proportion of graduates or junior employees across South Poland, nor do they establish that AI has caused the imbalance.

Nevertheless, they raise a commercially important question.

If the visible specialist market is already so heavily weighted towards experienced professionals, where will the next generation of expertise come from?

Organisations can compete for established talent. They can widen searches across Kraków, Wrocław, Katowice and neighbouring markets.

But repeatedly hiring from the same experienced talent pools is not a substitute for developing new capability.

At some point, organisations must become contributors to the talent pipeline, not just consumers.

The efficiency paradox

The promise of AI is attractive: remove repetitive work and give people more time to concentrate on valuable activities.

That can produce genuine benefits.

But consider the junior financial analyst who no longer prepares the report, the developer who no longer works through simpler programming problems, or the recruiter whose sourcing activity is increasingly automated.

What replaces the learning those activities previously provided?

An AI-generated result can be correct without the person reviewing it understanding why.

A dashboard can highlight a problem without teaching somebody how the underlying business operates.

A candidate shortlist can be produced without developing a recruiter's understanding of the market.

That creates a distinction leaders cannot afford to overlook.

Access to an answer is not the same as possessing expertise.

And being freed from a task does not automatically equip someone to take on more complex responsibilities.

AI can also become part of the solution

This isn't an argument for preserving inefficient work simply because people once learned from it.

That would be a strange way to prepare for the future.

The more interesting opportunity is to use AI to redesign development itself.

Imagine a junior analyst learning through an AI-generated report. Rather than simply accepting the output, they're asked to identify the assumptions, challenge the conclusions, and explain the commercial implications.

Or a junior developer working alongside AI, but required to understand the underlying architecture, test alternative solutions, and explain why one approach was chosen.

Or a recruiter using AI to identify potential candidates, while developing their own judgement about market availability, motivations, suitability and the client's actual hiring challenge.

The technology assists.

The learning comes from questioning, applying and understanding it.

Research into AI-assisted customer service has already demonstrated that technology can help less-experienced workers benefit from knowledge accumulated by more-experienced colleagues.

That is an important distinction.

AI doesn't inevitably remove the learning curve.

But organisations need to decide whether their AI strategy will accelerate learning or merely accelerate output.

For employers: redesign the learning pathway alongside the work

Before automating an activity, understand what that activity currently teaches.

If a junior employee no longer prepares the report, how will they learn to recognise an error?

If AI handles initial research, how will employees learn to investigate unfamiliar problems?

If routine customer interactions disappear, where will future relationship managers develop their understanding of customers?

This requires practical changes to workforce design.

Give employees opportunities to examine AI outputs rather than merely approve them. Introduce supervised decision-making, simulations, structured mentoring, cross-functional assignments, and real responsibility that increases with demonstrated capability.

Most importantly, involve experienced employees in designing those pathways.

Their knowledge should become something the organisation deliberately transfers, rather than something it risks losing when they leave.

A useful starting point is to select one role already being changed by AI and examine its development pathway from entry level to specialist.

Identify what people previously learned from their tasks, what AI now performs, and what new experiences can build the same or better capabilities.

Then measure something more meaningful than hours saved.

Measure what your people can understand, challenge, and eventually own.

For candidates: learn the thinking, not just the technology

The expectations placed on professionals are changing.

AI proficiency increasingly matters. But there's a difference between knowing how to produce an answer and understanding whether it's the right answer.

Be curious about the work behind the output.

Ask experienced colleagues to explain their decisions. Volunteer for problems that require investigation. Build enough domain knowledge to recognise when technology produces something misleading.

And when describing your experience, explain how you developed your judgement.

What did you challenge? What did you improve? What did you learn from a mistake? What can you now take responsibility for?

Employers should create opportunities for this development. Candidates can also be active participants in seeking them out.

The objective isn't to compete with AI at completing every task.

It's to develop the expertise required to use technology responsibly and create outcomes that matter.

The next chapter for Poland and CEE

Harvinder Rattan :

"

At ASPIRE, I also participated in the World Stage discussion, Five Cities or One Region: The Business Case.

For me, an important strategic question arising from that discussion is how South Poland can use its connected ecosystem to develop capability rather than having individual cities and employers compete for the same existing talent.

Our workforce intelligence suggests that the measured GenAI talent markets already contain substantial senior expertise, but comparatively limited early-career representation.

That presents an opportunity for greater cooperation between employers, universities and regional institutions.

The same principle matters beyond South Poland, including Hungary and the wider CEE market.

As organisations expand higher-value operations, workforce planning needs to account for the time and investment required to develop the people who will eventually lead them.

It isn't enough to understand where capability sits today.

We must also understand how tomorrow's capability will be created.

"

The Belvedere view

At Belvedere, we believe recruitment should begin long before a vacancy is advertised.

Workforce intelligence helps organisations understand where talent sits, which skills are available, where competition is concentrated, and what existing talent pools can realistically support.

But market intelligence should also reveal where hiring alone isn't the answer.

Our new research provides a starting point for those conversations across South Poland. The same approach can be tailored to an organisation's particular roles, locations, competitors and hiring plans.

This connects our work across Workforce Intelligence, Market Entry, Capability Building and Regional Scale.

Because a genuinely effective workforce strategy must answer two questions:

Where will we find the expertise we need today?

And how will we develop the expertise we'll need tomorrow?

Ignoring the second question risks creating another capability problem while solving the first.

The question we leave you with

AI will continue changing the work.

That isn't the difficult part to imagine.

The more challenging question is whether our approach to learning, development and leadership will change alongside it.

If AI takes over the work your future experts once learned from, what are you putting in its place?

We would welcome your thoughts.

Continue the conversation: our South Poland Workforce Capability Index

Following ASPIRE Poland , we're making our South Poland Workforce Capability Index 2026 available to readers.

Developed with Horsefly Analytics, it explores talent depth, demand pressure, skills, salary expectations, and capability risk across the measured South Poland markets.

For your personal copy, visit www.blvdrglobal.com/aspire.

And if you're planning to hire, expand or build capability, we can develop a tailored Workforce Intelligence Briefing for your organisation, examining your target locations, specific roles, competitor talent pools, salary pressures and hiring risks.

If this edition resonated, follow Belvedere and share it with someone thinking about talent, leadership or workforce transformation.

And tell us which workforce challenge you'd like us to explore next. We genuinely read every message.

The Belvedere Team

Growing globally. Rooted in Poland, Hungary & the UK.

www.blvdrglobal.