Are ‘soft skills’ about to become the hard skills of the AI era?

For years, young people have been given a fairly clear message about preparing for the future of work: study STEM, develop technical skills and choose a degree with an obvious route into employment.

It made sense. Technical skills were in demand, employers struggled to find enough of them and people who possessed them could command a premium.

But as AI becomes capable of taking on more technical and knowledge-based tasks, is it time to reconsider what will really differentiate people in the workplace?

And could that be particularly good news for humanities graduates who might currently be wondering where their skills fit in an increasingly AI-enabled world?

When producing the answer is no longer the difficult bit

An article by Kevin Wheeler recently raised an interesting argument about education in the age of AI.

AI is rapidly reducing the time and expertise needed to perform tasks such as coding, research, data analysis, translation, report writing and creating presentations.

That doesn't make technical expertise redundant. Far from it. We will continue to need people with deep specialist knowledge, including those developing and improving AI itself.

But it does potentially change the value equation.

If technology can produce code, analyse a dataset or generate a credible-looking report in seconds, the differentiating skill may increasingly be knowing what to ask it to do in the first place.

And then knowing whether the answer it gives you is actually any good.

That requires a rather different set of capabilities.

Judgement. Curiosity. Critical thinking. Communication. Context. Ethics. Understanding people. Recognising assumptions. Dealing with ambiguity.

We have traditionally bundled many of these together under the heading of "soft skills".

There is increasingly very little soft about them.

Humanities graduates: your skills are far from redundant

For someone currently studying history, philosophy, English, psychology or another humanities or social science subject, the conversation around AI could understandably be unsettling.

When so much attention is focused on technology, coding, data and AI skills, where does a humanities degree fit?

Potentially, rather well.

Think about what those subjects actually require students to practise.

A history student learns to interrogate sources, understand context, weigh conflicting evidence and recognise that apparently similar situations may have very different causes.

Philosophy develops logic, argument, the examination of assumptions and the ability to reason about questions where there may not be one objectively correct answer.

Literature requires interpretation, sensitivity to language, understanding of narrative and an appreciation of ambiguity.

Psychology and related disciplines develop understanding of motivation, cognition, relationships and human behaviour.

Those capabilities don't suddenly become obsolete because AI can produce an essay, summarise a document or analyse some data.

They may become more important precisely because it can.

In a world where producing an answer is becoming incredibly easy, working out whether it's a good answer could become much more valuable.

AI literacy still matters

None of this is an argument for avoiding technology. Quite the opposite.

Young people entering the workplace should understand how to use AI effectively. They need to know what these tools are capable of, where they can fail, how to question their outputs and how AI can help them work differently.

Basic numerical, data and digital literacy will matter too.

But perhaps we need to stop thinking about this as a choice between technical and human capabilities.

The strongest combination may be people who can think deeply, communicate clearly, understand evidence and people, exercise judgement and use AI extremely well.

After all, someone starting university today could still be working in the 2060s. Trying to predict which particular software, programming language or technical capability will be most valuable throughout that career is virtually impossible.

The ability to learn, question, adapt and reason has a rather longer shelf life.

What does this mean for employers?

There's another side to this conversation. If the capabilities employers need are changing, are our early-career recruitment processes changing quickly enough with them?

Graduate recruitment has often relied on easily defined proxies: degree subject, university, academic results, technical knowledge and previous experience. Those things can tell us something about a candidate. But they don't necessarily tell us how that person thinks.

• Can they recognise when information doesn't add up?

• Can they ask a better question?

• Can they consider competing explanations?

• Can they explain something complex to somebody without specialist knowledge?

• Can they make a reasoned decision when there isn't a perfect answer?

• Can they use AI to improve their work without outsourcing their judgement to it?

Those could become increasingly important questions for employers recruiting the next generation.

It also means employers may want to think carefully before narrowing opportunities unnecessarily by degree discipline. A humanities graduate who can think critically, understand people, communicate exceptionally well and become highly proficient with AI may bring exactly the combination an organisation needs.

Perhaps we need to rethink what makes someone "future ready"

The debate shouldn't become STEM versus humanities.

We need scientists, engineers, mathematicians, technologists and deep technical specialists. We also need historians, psychologists, writers, philosophers and people whose education has taught them to look at problems through a different lens.

Increasingly, we may need people who can combine elements of both. For students and graduates worried about what AI means for their prospects, there's an important message in that.

Your degree doesn't need to have taught you to compete with AI. The more interesting question is whether it has taught you to think, question, understand and judge while learning how to use AI alongside those capabilities.

And for employers, perhaps there's a corresponding challenge. If those are the capabilities that will matter in an AI-enabled workplace, are our recruitment processes actually designed to find them?

‍