As AI Changes the Work, Which Skills Become More Valuable?

There’s a tendency to divide workplace skills into two camps: technical skills on one side and “soft skills” on the other.

AI is making that distinction increasingly unhelpful.

As AI takes on more routine cognitive tasks – producing first drafts, summarising information, analysing data and generating options – the question for organisations is not simply which technical skills people need to learn.

It is also which capabilities become more valuable when the work itself changes. And the answer may be less about learning to do more, and more about learning to think, judge and act effectively.

From producing answers to evaluating them

One of the less discussed consequences of generative AI is that it can make producing an answer relatively easy.

The harder question is whether the answer is useful, accurate, appropriate and relevant to the situation. This requires judgement.

It means being able to challenge an assumption, identify what is missing, recognise when information doesn't quite make sense and understand the consequences of a decision. These aren't new workplace skills. But their importance changes when the cost of producing information drops.

If everyone can produce a competent-looking report, presentation or proposal in minutes, the differentiator is less likely to be who can produce one – and more likely to be who can determine what is worth saying in the first place.

That has significant implications for L&D.

The changing value of “soft skills”

The phrase “soft skills” has always been slightly misleading. Many of the capabilities that sit under it have very hard consequences.

-          Poor communication can delay a project.

-          A manager who avoids difficult conversations can allow a performance problem to grow.

-          A team that cannot challenge ideas constructively can make poorer decisions.

-          A leader who cannot bring people with them through change can undermine an otherwise sound strategy.

These aren't peripheral workplace behaviours. They directly affect performance. And there is growing evidence that employers are placing greater value on a combination of technical and human capabilities as work changes.

The World Economic Forum's Future of Jobs Report 2025, for example, identifies analytical thinking as the most sought-after core skill among employers, while resilience, flexibility and agility, leadership and social influence, and creative thinking also feature prominently among the skills expected to remain important.

The interesting point is that these aren't competing with technological capability.

They sit alongside it.

What this means for managers

There is perhaps an even more immediate implication for organisations. AI may change the tasks managers spend time on. It may automate reporting, help prepare information and make some administrative work considerably easier. But managers still have to interpret what is happening and decide what to do about it.

-          They still have to distinguish between a capability problem and a motivation problem.

-          They still have to give feedback.

-          They still have to manage disagreement.

-          And they still have to make decisions when the information available to them is incomplete.

In other words, as some of the work becomes more automated, the quality of human judgement becomes more important.

This makes management development particularly interesting.

So what should L&D be developing?

Perhaps the answer isn't to move away from traditional “soft skills” in favour of AI skills.

Nor is it to continue delivering communication, leadership and interpersonal programmes exactly as we always have.

It is to rethink the combination.

An employee who knows how to use AI but cannot evaluate its output isn't necessarily more capable.

A manager who can automate a report but cannot act on what it tells them hasn't solved the management problem.

And a team with access to better technology but poor communication may simply make decisions faster — not better.

The opportunity for L&D is therefore to look at capability more holistically.

Can people use the technology?

But also:

Can they question it?

Can they interpret information?

Can they make sound decisions?

Can they communicate those decisions?

Can they influence others and navigate the inevitable disagreements that follow?

These are not “soft” questions. They are questions about organisational effectiveness.

The skills shift

AI will undoubtedly change which technical skills organisations need. Some skills will become less valuable; others will emerge. But that doesn't mean the human side of work becomes less important.

It may mean we need to become more deliberate about developing it.

The organisations that benefit most from AI won't simply be those that teach people how to use the latest tools. They'll be the ones that understand which parts of human capability need to evolve alongside the technology.

For HR and L&D teams, that may be the more interesting question to ask now:

Not simply, “What do our people need to know about AI?”

But, “What will our people need to be exceptionally good at when AI is doing more of the work?”

That is a very different learning conversation.