There is a lot of talk around AI about which jobs will disappear.
That is the wrong focus.
Work does not disappear all at once. Tasks disappear first. And not all tasks — primarily those that are technical, repetitive, describable, measurable, or sufficiently automatable.
- AI writes.
- AI analyses.
- AI produces summaries.
- AI finds patterns.
- AI builds initial decision trees.
- AI prepares proposals.
- AI processes data.
- AI maps out processes.
- AI creates content.
- AI codes.
- AI replies to customers.
AI is doing more and more of what we used to consider human work.
But here is the important distinction.
AI does not take away all human value. AI takes away the technical part of the work.
And as the technical part shifts more and more to the machine, a question we have been postponing for too long surfaces: what is the real value of the human?
In my view, in the AI era four central roles remain for the human:
- setting direction;
- distinguishing right from wrong;
- communication with other people;
- carrying accountability.
Everything else, sooner or later, moves more and more toward technology. Not tomorrow, not in everything, not at the same speed everywhere. But the direction is clear.
Technical work becomes ever cheaper, faster and more accessible. Human value shifts to where the machine cannot truly decide on the human's behalf.
1. Setting direction
AI can help answer the question "how?". But the human has to ask "where to?" and "why?".
That is a big difference.
A machine can suggest ten strategies. A machine can calculate the risks. A machine can show which path is more efficient. A machine can propose the next step. A machine can simulate possible outcomes.
But a machine does not know what kind of future a person wants.
It does not know what kind of company we want to build. It does not know what culture we want to keep. It does not know at what price we are willing to grow. It does not know what we are not willing to give up. It does not know what success actually means for us.
These are not technical questions. They are questions of direction. And setting direction remains the work of the human.
If a leader has no direction, AI does not solve that problem. On the contrary — AI makes directionless movement even faster. The organisation may become more efficient, while moving in the wrong direction. Doing more, faster and cheaper — but not what is actually needed.
In the AI era, the most dangerous company is not the slow one. It is the fast one without a clear direction.
Because as speed grows, the consequences of wrong decisions grow too. That is why the central question of leadership is no longer how to do everything faster, but how to choose what to do in the first place.
2. Distinguishing right from wrong
The second thing a machine does not do for the human is distinguish right from wrong.
AI can make a recommendation. AI can offer a solution. AI can show what is likely to be more profitable, faster or more efficient. But is it right?
That depends on context.
- What is right in a crisis may not be right in a growth phase.
- What is right for an investor may not be right for an employee.
- What is right in the short term may be wrong in the long term.
- What is right in the numbers may be wrong for people.
- What is right legally may be wrong culturally.
- What is right in a process may be wrong in a relationship.
AI can bring context to the table. The human has to understand the context. AI can help see more information. The human has to decide what that information means. AI can offer options. The human has to decide which option is right.
This is exactly where the role of the human becomes especially important. Human work is no longer only about gathering or processing information. The machine does that better and better. Human work is to decide what to do when every option is in some sense right and in some sense wrong.
Real leadership begins where the spreadsheet no longer decides on its own.
If everything were just mathematics, no leader would be needed. If everything were just efficiency, no values would be needed. If everything were just process, no people would be needed.
But life is not only process. Companies are made of people. People live inside meanings, expectations, fears, ambitions, loyalty, trust and conflict. There it is not enough for a machine to say what is optimal. Someone has to decide what is right. And that someone is the human.
3. Communication
The third remaining human role is communication. Not just transmitting words. AI is very good at that.
Communication in the AI era means something deeper: the ability to make direction, decisions and meaning understandable to other people in a way that makes them want to come along. That is something a machine cannot fully do on the human's behalf.
A machine can write a speech. A machine can draft an email. A machine can build a presentation. A machine can phrase a strategy. A machine can prepare a difficult conversation. But a machine cannot be credible for you.
People do not follow text. People follow a person.
They watch whether you yourself believe what you are saying. They sense whether you actually understand. They listen for whether your words and your behaviour match. They notice whether you are speaking with them or speaking over them. They feel whether there is accountability behind you, or just a clever turn of phrase.
That is why humans do not disappear from communication. People do not want to do important things only with machines. People want to do important things with other people.
Not because humans are perfect. On the contrary — humans are imperfect. But it is precisely in that imperfection that personality, character, presence and trust live.
In business, in the end, people do not buy only a solution. They buy trust. They buy certainty. They buy the feeling that the other side understands. They buy a person who stands behind a promise.
AI can help sell. But the human has to build trust. AI can help lead. But the human has to hold the relationship. AI can help communicate. But the human has to be present.
When technical work becomes cheap, trust becomes expensive. And trust forms between people.
4. Accountability
Fourth, accountability remains. This is the most uncomfortable part.
If AI makes a recommendation and a person decides based on it, the person is still accountable.
- AI is not accountable when the wrong person is hired.
- AI is not accountable when someone is let go unfairly.
- AI is not accountable when a team breaks.
- AI is not accountable when a client is promised something the company cannot deliver.
- AI is not accountable when culture is destroyed in the name of efficiency.
Accountability stays with the human. It cannot be automated.
And that is exactly why the leader of the future is not simply a person who uses AI tools. The leader of the future is a person who understands what they are accountable for when they let AI do the work.
AI amplifies.
- If there is clarity in the organisation, AI amplifies clarity.
- If there is confusion, AI amplifies confusion.
- If leadership is fair, AI helps scale fairness.
- If leadership is unfair, AI simply formalises unfairness faster and more cleanly.
AI does not turn bad leadership into good leadership. AI makes the consequences of leadership visible faster.
Accountability is the line where the machine ends and the human begins.
Everything else becomes technical
Putting these four things together leads to a rather sharp conclusion.
Anything that can be described, measured, repeated and automated well enough moves, sooner or later, to the machine.
This does not mean the human does nothing anymore. It means human value can no longer be based only on "getting things done". Getting things done becomes more and more a technical capability.
Human value moves up a level:
- where to go;
- what to consider right;
- how to bring other people along;
- what to be accountable for.
This is a very big change, because much of today's working world is built on the opposite assumption. We evaluate people by what they do. We write job descriptions in terms of tasks. We hire people based on past experience. We measure jobs by activities. We build organisations around roles made of tasks.
But when tasks start to change, roles start to fall apart.
People do not disappear first — roles do
AI does not replace people all at once. Before that something more practical happens: today's roles become hollow or different from the inside.
If AI takes 30% of the tasks out of a role, what is left of that role? If AI takes 60%, is that person still in the same role? If only communication, decision-making, accountability and direction-holding remain, does that person fit the new role?
This is the most important question for leaders and HR.
Not: "which people do we have to replace?" But: "which part of a person's value remains once the technical work moves away?"
Some people's value grows. Some people's value decreases. Some people's role has to be redesigned. For some, it turns out they were in the wrong role all along. For some, it turns out their greatest value was never in the technical work but in direction, communication, decisions or accountability.
You cannot see that from a CV alone. You cannot see it from a job title alone. You cannot always see it from today's performance metric either. It has to be mapped deliberately.
The Sparkly perspective on this shift
From Sparkly's perspective, this is not only an HR question. The question is how to protect being human in the working world at a time when more and more of work is becoming technical and automatable.
This does not mean fighting AI. That would be pointless. AI is not going to stop. Automation is not going to stop. The pressure for efficiency is not going to disappear. The transformation of tasks is not going to be avoided.
The question is whether leaders and HR can understand early enough where the real human value proposition lies.
Sparkly's job is to help see:
- where work and the person fit today;
- where they no longer fit;
- which part of a role is technical;
- which part of a role is genuinely human;
- which person has the capacity to move into a new role;
- where the organisation needs new role design;
- where the leader is still deciding by the logic of the old working world.
Because if, in the AI era, a company starts evaluating people only by how much of their current work can be automated, it can make very bad decisions. It can lose the people who hold the direction. It can lose the people others trust. It can lose the people who can communicate hard decisions in a human way. It can lose the people whose real value was never in the task, but in the impact on other people.
And those people are not easy to buy back.
Philosophical summary
When AI does the technical work, value is not taken away from the human. What is taken away is the human's ability to hide their value behind activity.
What remains is a much more honest question:
- can you set direction?
- can you tell what is right and what is wrong?
- can you communicate this to other people?
- can you carry accountability?
These are the core questions of future human work. Everything else becomes more and more technical.
And that is exactly why the AI era is not only a technological revolution. It is a revaluation of human worth.
We should not only ask which work AI takes away. We should ask: what is the human really needed for afterwards?
My answer is simple. The human is needed to give direction, to make value-based choices, to build trust with other people, and to carry accountability. That is the core of being human in the working world. And we need to start preparing for it now — not when it is too late.
Practical summary for leaders
1. Map tasks, not job titles
Do not start from the question "which job titles will AI replace?". Start from the question: which tasks in each role are technical and automatable?
Split every role into four:
- technical tasks;
- tasks that require decisions;
- tasks that require communication and trust;
- tasks that carry accountability.
Only then do you see what actually remains of the role.
2. Find the real value of the person
For every key person, ask: is their value in technical execution? Is their value in holding direction? Is their value in decision-making? Is their value in communication? Is their value in accountability? Do other people want to work with them? Does clarity or confusion form around them?
In the AI era, the most important part of a person becomes the part that cannot be easily automated.
3. Redesign roles before the crisis
Do not wait until AI has already hollowed out the role. Look ahead: what disappears from this role in 12 months; what changes within 24 months; what remains within 36 months; what kind of person is actually needed in that role then.
If today's role consists mostly of technical tasks, the role has to be rebuilt. If the person does not grow with the role, conflict arises.
4. Teach leaders to decide, not just to use tools
AI training often focuses on how to use tools. That is necessary, but not enough. Leaders also need to be taught: how to evaluate AI's recommendations; how to decide on the basis of insufficient information; how to take human context into account; how to communicate uncomfortable changes; how to take accountability for decisions that AI helped prepare.
The leader of the future is not a great prompter. The leader of the future is a great decision-maker.
5. Communicate honestly
People understand that work is changing. If a leader hides it, fear grows. If a leader only talks about efficiency, resistance grows. If a leader honestly says what is changing and why, at least the possibility of trust emerges.
Communication has to answer four questions: what is changing; why it is changing; what it means for people; how we will help people move into the new role.
The AI shift cannot be managed only as a technical project. It is a human change project.
Practical summary for HR
1. Do not look only at the job description
Job descriptions often describe yesterday's world. HR has to start looking at the actual content of a role: which tasks are automatable; which skills are becoming less important; which human capabilities are becoming more important; what kind of person the company actually needs in that role.
2. Measure fit for the future, not only experience from the past
A CV shows what a person has done before. In the AI era, the more important question is: can the person learn; can they adapt to change; can they take accountability; can they communicate; can they understand context; can they make good choices in situations where there is no clear answer.
Past experience does not guarantee fit for the future.
3. Make human value visible
HR's new role is not just managing people. HR's strategic role is to help leaders see the part of a person that Excel, CVs and job descriptions often do not show: trustworthiness; communication ability; decision-making ability; carrying accountability; impact on the team; fit with role and culture; capacity to move into a new working world.
This is where HR can move from an administrative function into a strategic function.
Practical summary for the company
First step
Take one team and map all of that team's tasks. For each task, mark: can AI do it; can AI accelerate it; does a human have to decide on it; does it require trust and communication; who carries accountability.
This produces the first honest picture.
Second step
Look at which roles will change the fastest. They are not necessarily the weakest roles. Often the fastest-changing roles are the ones with a lot of writing, analysis, coordination, information processing and repetitive decisions.
Third step
Look at the people behind the roles. Ask: who grows with the role; who gets stuck in the old role; who needs support; whose real value has been invisible so far; who must not be lost in the rush for efficiency.
Fourth step
Start redesigning before pressure forces you to. If you wait for the crisis, you lead from fear. If you start earlier, you can lead from clarity.
Closing
AI does not ask whether we are ready. It will change work anyway. The question is whether we manage to understand, before that, what remains as the real value of the human.
My answer is: direction, distinguishing right from wrong, communication, and accountability. Everything else becomes more and more technical.
And that is why, from Sparkly's perspective, the question of future HR and leadership is not only about how to organise work with people better. The question is how to protect and make visible the value of being human at a time when machines are doing more and more of the work.