AI pilots rarely fail because of the technology. They fail because of what people can and cannot do, dare and know. An organization can have excellent data and solid infrastructure and still get stuck, because nobody really knows what AI does and does not do, when mistrust is justified and when it is not, and who has which role when a model delivers something that a human must assess. This page is about that dimension: people and skills, one of the seven dimensions mapped by the hybridresourcing maturity measurement.
People and skills covers three layers that are often treated separately even though they are connected. The first layer is knowledge: do employees understand what an AI system does, where the limits lie and which errors are typical for the type of application they use. The second layer is skill: can people actually assess an AI outcome, adjust it or hold it up against their own judgment. The third layer is attitude: is there room to try AI, report mistakes and make adjustments, or does hesitancy or blind trust dominate. An organization that only invests in training but does not address attitude often remains stuck at the same level, even after years of courses.
Because this dimension depends heavily on individual people, hybridresourcing does not measure it with a single score but with a plot round: several involved parties score separately, and the spread between their answers becomes visible. A large spread is itself already an outcome. It shows that the organization appears on paper to have one level, while the work floor experiences something different from management.
At baseline level, there is no shared picture of what AI is or does. Individual people experiment, but there is no common language and no structure for sharing experience. At foundation level, basic knowledge is present, usually within a limited group, and the rest of the organization follows at a distance. Activation level means that a substantial part of the involved employees knows what AI means for their own work and can deal with it practically. Insight level adds that people can critically assess AI outcomes and know when they should have doubts. At intelligence level, that skill is spread throughout the organization, anchored in how people work and not dependent on a few early adopters.
Where an organization stands on this scale is never a matter of intuition. It depends on how many people actually work with AI, how diverse their experiences are and whether those experiences come together somewhere. That is exactly why the plot round yields more than a survey with a single respondent per team.
A number of signals point to a low level. If questions about AI always end up with the same few people, knowledge is not spread. If nobody dares to reject an AI outcome because they do not know the logic behind it, the skill to assess is missing. If AI projects are started without involving the people who have to work with them, the support needed to move something beyond the pilot phase is missing.
A higher level is recognized by the opposite: discussions about AI outcomes that are substantive rather than principled, employees who themselves propose where AI can and cannot help, and an organization in which doubt about an AI result is seen as a normal part of the work, not as a sign that the project has failed.
The path to a higher level rarely begins with more technology. It begins with making visible what people already know and where the gaps lie, and with organizing moments where experience is shared instead of remaining isolated in separate experiments. What that costs in concrete terms, in time and attention, differs strongly per organization and per starting level; those wanting an indication of that can find it on the page that calculates what it costs to raise an organization by one level.
People and skills, incidentally, is not separate from the other dimensions. Anyone who wants to assess AI must also understand how that assessment is recorded and used, and that touches on performance management when it comes to AI. Anyone who wants to dare report mistakes must know that privacy and security are in order, something connected to what it costs to raise privacy and security by one level. And anyone who wants to build trust in AI outcomes will sooner or later run into questions that belong to what it costs to raise ethics by one level.
This page describes whether the people in your organization can carry AI: understand, assess and adjust it. It does not describe which part of the work itself can be taken over by AI. That is a different question, with a different instrument. The work scan of FTE TO AI calculates per task which part of the work qualifies for AI, given the nature of the task and the context in which it is carried out. That outcome only becomes meaningful once the carrying side, including the people who have to work with the outcomes, has already reached a certain level. Those who want to know where the organization currently stands can sign up for the waiting list of the maturity measurement, which is still under construction.
Vraag maar wat er moet staan voordat AI in uw organisatie kan landen.
Answers come from this site’s knowledge base. Not tailored advice, and not a scan of your company.