AI · Transformation · Work
Buying licences is one decision. Changing how a company works is another. Between them sit failed tests, different rates of progress, processes that no longer make sense and managers who also have to learn.
AI can add value very quickly to a specific task. The problem begins when we mistake that first saving for a transformation that has already happened.
The licence does not change the work
Providing access and asking people to “use AI” leaves the team with a tool and little judgement about where to apply it. Some people will experiment. Others will wait until a safe and recognised way of doing so exists.
Adoption begins more effectively with problems: which part of the work consumes time without adding enough value, which response arrives too late or which decision could draw on more information?
Adoption grows when the team is allowed to test
The people doing the work see opportunities that do not appear in a central plan. They need room to experiment, share what works and admit what does not. They also need to know which risks are acceptable and which are not.
That combination — clear direction and freedom to test — allows a local improvement to become learning for other teams. Without it, good practice stays isolated or control prevents exploration.
The manager also has to learn in public
A manager who asks the team to use AI but does not enter the work signals that transformation belongs to other people. They do not need to master every tool. They do need to test, ask questions and show that their own judgement is still developing.
That includes acknowledging a failed test and distinguishing it from a poor decision. If every error is punished, the team will hide the learning. If every experiment is celebrated without looking at the outcome, there will be activity without direction.
Scale comes after value has been found
The useful question is not how many people opened the tool. It is where a workflow, a decision or a customer proposition changed. That gives us something to observe, improve and extend.
Transformation is not distributing technology. It is creating the conditions in which teams discover a better way to work and the organisation knows how to recognise it.