One Platform. Unlimited Learning.

The more capable AI becomes, the more human skills matter

If AI can draft, summarise and predict in seconds, what place remains for human skills? A growing one. Knowing when to trust AI, when to verify it and when to take back control has become essential, and emotional intelligence, judgement and creativity remain distinctly human. This article explains why training digital professionals today means training them in AI, in judgement and in human relationships, and how Holberton Edu makes Life Skills a core part of every programme.

Holberton
November 16, 2026
4min Read

The more capable AI becomes, the more human skills matter

Artificial intelligence now drafts, summarises and predicts in a matter of seconds. In companies and educational institutions alike, one question has become unavoidable: if machines can produce so quickly and so well, what place remains for human skills? Our conviction is clear: the more powerful our tools become, the more the ability to understand, to engage in dialogue and to make decisions makes the difference. This is why, across all Holberton Edu programmes, Life Skills and human-centred AI are taught every year, on a par with technical subjects.

Producing more does not mean communicating better

AI has profoundly transformed the production of information. A report, a summary, a presentation or a first draft of code can now be obtained in moments, often with impeccable formal quality. This gain in efficiency is real, and there would be little point in denying it.

Yet information only has value if it is understood, discussed and carried by someone. A perfectly structured presentation will not convince if the person delivering it can neither explain it nor respond to objections. A flawless document is not, on its own, enough to build trust within a team. In other words, what AI makes easier to produce is not necessarily easier to communicate.

This is precisely where the challenge lies: automation is taking on a growing share of production, but it leaves untouched, and at times weakens, the relational and reflective dimension of work.

Real but uneven performance

This limitation does not stem solely from the nature of human interaction; it also stems from the way the tools themselves work. The capabilities of AI are undeniable, but they are uneven. Researchers at Harvard Business School and Boston Consulting Group have described a “jagged technological frontier”: the same tool may excel at one task and fail at another of comparable difficulty, without the user always noticing.

This creates a new requirement. Knowing how to use AI is no longer enough: one must also know when to rely on it, when to verify and when to take back control. Such discernment cannot be acquired through a tutorial or a mere introduction to the tools. It rests on a genuine command of the fundamentals, without which it is impossible to detect an error, an approximation or a weak line of reasoning.

What remains distinctly human

If AI cannot be used wisely without discernment, the skills on which that discernment rests still need to be identified. Four of them currently appear particularly difficult to entrust to a machine.

The first is emotional intelligence: perceiving what another person feels, adapting one’s message, building trust. These abilities are not innate; they develop through the experience of professional relationships.

The second, closely related to the first, is active listening. Language models predict probable sequences of words; they do not interpret a hesitation, a silence or what a request leaves unsaid. Understanding a need also requires grasping what has not been expressed.

The third is judgement in ambiguous situations. When data are incomplete or contradictory, it is for people to decide, taking into account the context, the stakes and the consequences.

The fourth, finally, is creativity. AI recombines what already exists; imagination, intuition and the ability to question an accepted assumption remain human.

Taken together, these skills describe a single profile: that of a professional who is able not only to execute, but also to understand, to engage in dialogue and to arbitrate.

Producing an answer is not the same as owning a decision

This capacity to arbitrate leads to an essential distinction. AI can help formulate an answer; the decision, and the responsibility that comes with it, remain human. Some moments in professional life cannot be delegated: difficult feedback to a colleague, the resolution of a conflict, a decision with ethical implications.

The risk for organisations would be to become ever more efficient at producing information, and ever less able to communicate it, discuss it and take responsibility for it. Employers are fully aware of this. According to the World Economic Forum’s Future of Jobs Report 2025, analytical thinking remains the skill considered essential by seven out of ten companies, ahead of resilience, flexibility and leadership. Technical skills related to AI are growing rapidly, but they do not replace these transversal skills: they add to them.

Training in AI, training in judgement

The question, therefore, is not whether to choose between mastering AI and developing human skills. These two requirements reinforce each other. Digital professionals must be able to use the latest tools, but also to evaluate, challenge and explain what those tools produce.

This principle shapes all our programmes and unfolds in three stages: knowing how to work without AI, knowing how to work with it, and knowing how to verify. Each subject first teaches the core practices of the profession, then the use of the AI tools specific to that profession, and finally the verification of what those tools produce. The very first module our students take is devoted to this: formulating a precise request to an AI, checking its answer against an independent source, declaring how it was used and correcting it where necessary.

This requirement is also reflected in the way students are assessed. Throughout the programme, Holberton Edu students are assessed on certain tasks without the use of artificial intelligence. The use of AI is then introduced progressively, so that students master the fundamentals before relying on tools. Every student thus learns to reason independently, an essential condition for judging what a machine produces.

Life Skills and human-centred AI: a common foundation across all our programmes

Training in judgement cannot, however, be limited to the use of tools. Emotional intelligence, listening, the ability to decide: these skills can be learned, provided they are given a genuine place in education. This is why, in all our Bachelor’s programmes, a cross-disciplinary foundation is devoted to them every year. It has its own hours, exercises and evidence of achievement, exactly like technical subjects.

Life Skills follow a three-year progression. Students first learn to know themselves better and to make a reasoned decision. They then practise giving and receiving useful feedback, and working through a disagreement or a setback. In their final year, they take responsibility for a group. Human-centred AI follows the same logic: from the first year onwards, every solution is examined in light of the people it affects, what it changes for them and the remedies it must provide.

This learning takes shape in real situations. Every year, our students carry out a team project for a partner organisation. They review their peers’ work and defend their choices before their stakeholders. They use AI as professionals do, but must always be able to explain and justify what they produce.

Human skills, more decisive than ever

AI does not make human skills obsolete; it increases their value. The more powerful our tools become, the more important it is to use them with discernment, to verify their results and to take responsibility for their consequences. Training digital professionals today therefore means training them in AI, in judgement and in human relationships: this is the purpose of the common foundation shared by all our programmes. AI can draft an answer; the decision, however, always commits the person who makes it.

‍

Also on our blog: “The human value of expertise in the age of AI” and “AI brings professional judgement into focus”.

References: Dell’Acqua et al., “Navigating the Jagged Technological Frontier”, Harvard Business School and Boston Consulting Group, 2023; World Economic Forum, Future of Jobs Report 2025; Harvard Business School Online, “4 Important Human Skills AI Can’t Replace”, 2026.

‍