When AI systems take over data gathering, report preparation and part of routine coordination, companies may need fewer management layers. The same change, however, removes activities through which younger employees learned to understand operations and prepared for decision-making roles. Savings from a flatter hierarchy therefore need to go hand in hand with a new system for developing managerial judgment.
Much of today’s corporate hierarchy arose at a time when information moved slowly. Someone had to obtain it, check it, summarize it, pass it to a superior and then deliver the decision back to the people who carried it out. AI is changing this information economy.
Using Palantir’s systems as an example, Forbes describes a model in which software connects organizational data, flags deviations, looks for relationships and prepares proposed actions. If a system can do in a few seconds work that previously required several people and several organizational levels, a legitimate question arises: how much traditional coordination does the company still need?
The article also points to an important side effect. Preparing reports, analyses or materials for meetings was not only production work. Younger employees learned through it which numbers mattered, how parts of the company were connected and which questions experienced managers asked. Automation can therefore remove not only work but also part of this informal apprenticeship.
Forbes cites several signals from the labor market. Research suggests a decline in hiring of younger workers in occupations exposed to AI, and a Revelio Labs analysis recorded a substantial drop from the 2022 peak in the number of vacancies for middle and senior management. At the same time, the article explicitly warns that these changes cannot simply be attributed to AI: previous overhiring, interest rates and ordinary cost-cutting also play a role.
For companies, the mechanism matters more than the number of eliminated positions. If you remove an activity that also served as training, you have to create another one. Forbes proposes a shift from “learning by producing materials” to “learning by making decisions”: a junior manager does not build the entire model manually but must dissect the system’s recommendation, challenge its assumptions, look for missing relationships and decide when not to follow it.
This also changes the way people are developed. It is not enough to give an employee access to a ready-made AI recommendation. They need to see what the system based it on, what options it considered and why a more experienced manager ultimately chose a particular alternative.
When streamlining the organization, therefore, track three practical indicators alongside savings: how much time managers spend moving information rather than making decisions, how many people an important decision passes through, and how long it takes from the emergence of a relevant signal to intervention by a person with the necessary authority. This is where you will see which managerial work can be removed and which has to be replaced by higher-quality decision-making.
KEY TERMS
- Hierarchy flattening: A reduction in the number of management layers between senior leadership and the operational part of the organization.
- Decision apprenticeship: Development of judgment through analysis of recommendations, alternatives, assumptions and the actual outcomes of decisions.
- Information delay: The time between the emergence of relevant information and the moment it reaches a person able to act.
- Decision distance: The number of organizational steps through which information and approval pass between a problem and the action taken.