The wolf did not eat the deer

The impact may lie in how tasks are redistributed, not in how many jobs disappear. A cautious look at preparation for AI.

The wolf did not eat the deer

Series: What nature already knew about talent · 5/6 · Preparation for AI

In January 1995, after seventy years without wolves, fourteen specimens brought from Canada were released in Yellowstone National Park. The following year, seventeen more arrived. Today, there are just over a hundred.

What happened next has been told so many times that it has almost become a fable. The deer, which had lived for decades without predators and had multiplied to twenty thousand, stopped grazing peacefully along the riverbanks. The willows and aspens, which had been unable to grow for years because the deer ate them from the shoots, recovered. With the trees came the beavers, who built dams, which created wetlands, which brought birds, amphibians, and fish. Even the course of the rivers, some say, changed.

Ecologists today debate how much of that story is exactly true. The deer also decreased due to hunting, drought, and bears. The effects on the rivers are likely exaggerated. But the core of what happened is well documented and is what interests me: the return of the wolf did not eliminate the deer. It changed what the deer did. And that change in behavior, not the death of any animal, was what reordered the ecosystem.

What a predator does

Intuition says that a predator kills. And it does kill, of course. But the effect of a predator on a population does not primarily come from the animals it eats. It comes from the behavior it imposes on those it does not eat.

The deer in Yellowstone continued to be deer. They continued eating, reproducing, migrating. What changed was the distribution of their time: less time in the open valleys by the water, where they were vulnerable, and more in the slopes and forests. Less time grazing with their heads down and more time watching. The tasks that made up a deer's day were redistributed, and from that redistribution came everything else.

Ecologists call it the 'ecology of fear.' I would call it something else: the predator does not change the position, it changes the tasks.

AI does not eliminate positions

For three years, we have been hearing the wrong question about artificial intelligence in the workplace: which positions will it eliminate. And for three years, we have seen that the answer, in most organizations, is 'almost none yet,' which leads many people to conclude that the impact is less than what was said.

It is the conclusion of someone who looks at how many deer have died.

What is happening is the other thing. Within each position, the distribution of tasks is changing. The analyst position still exists, but the part of their week spent preparing data has decreased, and the part spent deciding what to ask and explaining the answer has grown. The sales position still exists, but it prospects differently and prepares meetings differently. The administrative manager, the support technician, the writer, the programmer: all still exist, and none do the same thing they did in 2022.

We have our own data that illustrates this well. In the analysis we conducted this quarter on half a million job offers, the two profiles that grew the most in the entire healthcare sector were not clinical. They were the medical writer and the medical transcriptionist, with growths of over a thousand percent and almost six hundred percent in ninety days. They are not new positions. They are the signal that the care pressure has redistributed tasks: the documentation work that clinicians used to do between patients is shifting to profiles that take it on, and to tools that assist it. The physician position is still there. What a physician does in their day has changed.

That is Yellowstone. The wolf has arrived, the deer are still alive, and the landscape is changing underneath without anyone having seen any die.

The question to ask

If AI changes tasks and not positions, the useful question is not 'which positions will it eliminate,' but 'what part of the work of each position in my organization is changing, and what does the person occupying it need to remain valuable in the part that remains.'

That question can be answered. Each profile is broken down into the specific tasks of their day-to-day. For each task, it is estimated how much of it can be assumed by technology today and how much still requires human judgment. It is aggregated by profile and the entire organization is examined. And three groups emerge, quite consistently.

Profiles where most tasks still require judgment, context, and relationship: AI affects them little in the short term, and the priority is to continue developing them.

Profiles where a significant part of the tasks can be assisted: here AI multiplies those who learn to use it and leaves behind those who do not, and this is where training has the most immediate return.

And profiles where a good part of the tasks is automatable. Here the word is not 'cut,' it is 'reconvert in time.' Because the person occupying that position has knowledge of the business, of the clients, of how things really work, and that knowledge is worth much more in another role than lost.

The deer that adapted

What I like most about Yellowstone is that the deer that survived were not the strongest or the fastest. They were the ones that changed their behavior first. The ones that learned to lift their heads, to move to the slopes, to distribute their time differently.

In organizations, the same will happen, and it will happen person by person. The difference from the park is that we can see the map beforehand: knowing which roles change the most, what new skills each one requires, and who is closer or farther from having them. With that, decisions can be made about whom to reinforce first, in what, and with what urgency. Without that, decisions are made when it is already too late, and then the only tool left is the one no one wants to use.

At The Wise Skill, we measure that: how each role changes with AI, task by task and with the same formula for comparable profiles, and what capabilities each person has today in the face of that change. And with a principle that I think is important: the map exists to reconvert in time, not to decide who gets replaced. No reading translates into an automatic decision about a person.

But you can ask the question tomorrow in the committee when someone brings up the topic of AI.

Not how many deer are going to die. What are the ones that are still alive doing differently.

If you are adopting AI in processes and someone in the committee asks to know its real impact on roles and on people, message me directly and I will show you how to read the map.

Final installment: the geese, or why the journey is prepared months before takeoff.