Artificial intelligence is changing what people can accomplish at work and how quickly. Tasks that once took hours can be completed in minutes. Information can be analysed, documents summarised, reports drafted and ideas generated almost instantaneously. It is both exciting and worrying what AI tools can help us all achieve.
So it is tempting, therefore, to assume that companies will need to invest less in developing their employees. After all, if technology can provide information on demand and take over an increasing number of (often tedious) workplace tasks, perhaps employees simply won't need the breadth of skills they once did.
I disagree with that view and believe the opposite may prove to be true.
As AI becomes embedded in everyday working life, the skills that distinguish highly effective employees are increasingly those that technology finds hardest to replicate. For example, judgement, communication, negotiation, leadership, empathy and the ability to understand complex situations involving real people.
These have always been some professions where success depends as much on people as it does on processes. But those same “soft skills” (or maybe more accurately “power skills”) are equally applicable across almost every industry. As an example, an IT company full of people with in-depth technical knowledge won’t thrive without people who can explain advanced technology in layman’s terms or negotiate deals in difficult environments or understand when and how to compromise.
In an AI-enabled workplace, Learning & Development (L&D), therefore, has a new and potentially more important role – that of developing the human capabilities that allow employees to use increasingly powerful technology effectively.
AI provides answers but people need to judge which are the right ones
One of the great advantages of generative AI such as ChatGPT, Claude, Gemini etc. is its ability to provide apparently knowledgeable answers almost instantly. But having an answer is not the same as making a good decision.
Business decisions rarely take place in isolation. Instead they involve organisational priorities, commercial pressures, personalities, politics, incomplete information and competing objectives. There may be several technically reasonable answers but no objectively correct one.
An AI system might analyse the available data and suggest the most efficient course of action. But it’s only an experienced employee who can recognise that the theoretically optimal solution will never gain the support of a key stakeholder or that implementing it would create problems elsewhere in the organisation. That ability to interpret information in context comes from an understanding of the business, experience, critical thinking and judgement.
Consequently, Learning and Development should be moving beyond simply transferring knowledge. Employees increasingly need opportunities to practise evaluating information, challenging assumptions and making decisions when there isn't an obvious right answer.
Face-to-face communication becomes more valuable
Technology has already reduced the amount of face-to-face communication in many workplaces. Still many people work remotely or have a hybrid arrangement so we are all well-used to meetings being held virtually on Zoom or Teams. Indeed I’ve heard many people refer to teams meetings as “face-to-face” – technically correct maybe but a world away from a group of people in the same room.
Either with or without the help of AI tools, we can generate emails, summarise meetings and produce written explanations without ever speaking to another person. That is very useful when communication is primarily about transferring information, but some conversations are about much more than transferring information. And if we don’t recognise that fact then there are potential dangers in employees failing to communicate fully on important tasks or projects.
Imagine telling a valued employee that their role is changing significantly because of a new project being planned. Or explaining to a client why an important deadline will be missed. The words used in those conversations are only part of what matters. Tone, timing, body language, and the ability to recognise and respond to another person's reaction all influence the outcome. A conversation may need to change direction because of a hesitation, facial expression or unexpected response.
AI can help someone prepare for that conversation. It might, for instance, suggest questions to ask or identify possible objections beforehand. But eventually somebody has to walk into the room - physically or virtually - and have the conversation.
Companies that allow employees to become overly dependent on digital communication risk weakening precisely the interpersonal skills that will become more valuable as many other activities are automated.
Difficult negotiation cannot be delegated to an algorithm
Negotiation provides an even clearer example of why human skills are needed in certain situations.
AI can be extremely useful before a negotiation. It can analyse information, identify possible scenarios, help establish a negotiating position and suggest responses to likely objections.
But during actual negotiations people may reveal new information – intentionally or unintentionally - so priorities can change during the discussion. One party may make what appears to be an unreasonable demand because there is an underlying concern that has not been addressed.
People who are good negotiators will listen for those signals and know when to press for more information or suggest a compromise, or even make a minor concession if that could lead to an important agreement elsewhere.
These negotiation skills are relevant far beyond senior management or specialist commercial roles. Employees negotiate constantly: over deadlines, workloads, budgets, resources, priorities and responsibilities.
As AI takes care of more of the preparatory and analytical work, being able to manage the human interaction itself becomes increasingly valuable. That is a skill that needs to be developed through practice, feedback and experience.
Making something complex easy to understand
Think we can all agree that AI is good at processing complexity, but businesses still need people who can make that complexity easily understandable to other people. This is especially true in technical environments.
A technical specialist may understand exactly why a new system needs additional investment, but the people deciding whether or not to increase the available budget do not need or want a technical explanation. They need to understand the business implications.
Effective communication techniques involve understanding not only the subject but also the audience.
AI can certainly help simplify information, but an employee still needs to understand the context well enough to determine whether the simplified explanation is accurate, appropriate and useful.
The ability to take something complicated and explain it clearly to a non-expert will become increasingly important when we are surrounded by ever-growing volumes of AI-generated information.
Employees need to learn how to challenge AI
There is another reason L&D matters in an AI-enabled workplace - employees need the confidence and knowledge to disagree with the technology.
An inexperienced employee presented with a polished, authoritative AI-generated answer may be inclined to accept it. Whereas, a more experienced professional is more likely to ask questions. This creates an interesting paradox in that the easier it becomes to access expertise via AI tools, the more important genuine expertise becomes.
People need sufficient knowledge of their field to recognise when an AI-generated response is misleading, incomplete or simply wrong. Removing professional development because employees can "ask AI" could therefore create companies in which people become increasingly dependent on systems whose outputs they are progressively less capable of evaluating.
For that reason, modern L&D should include AI literacy but only alongside continuing to develop professional expertise, especially in the area of communication, negotiation, leadership etc.
Human skills are developed by doing, not just by knowing
You can read about communication, negotiation, leadership and so on, but that doesn't necessarily make you a good communicator, negotiator or leader. Knowing the principles of, for instance, good communication doesn't guarantee that you can explain a difficult decision clearly when somebody is angry or anxious.
Human skills need opportunities for practice - opportunities to apply knowledge, work with others, solve problems and learn from experience.
L&D should complement AI, not compete with it
None of this means businesses should resist the use of AI tools. Indeed, employees can use AI to explore scenarios, challenge their thinking or obtain explanations tailored to their existing level of knowledge.
What it does mean is that it’s not an either-or situation. Businesses do not need to choose between developing people and investing in AI - both investments complement one another. Research from McKinsey shows that companies gain the greatest value from using AI for routine work when employees actively apply human judgement instead of relying on automation alone.
AI technology may become less of a differentiator as more and more businesses adopt it. What could instead differentiate businesses is the quality and expertise of the people using it. And that can only be assured by continuing to develop the people. AI is undoubtedly changing the skills people need at work, but that does not make Learning & Development less relevant. It just changes what that development needs to achieve.