The introduction of AI is changing individual tasks before processes and job descriptions have had time to change. In many cases, the main obstacle is not the tool itself, but assumptions about how a task should be performed and what a person must do themselves. This training helps participants identify these assumptions in their own work, test them rather than defend them, and explore alternative ways of approaching the same task.
The training is based on participants’ real work tasks rather than hypothetical examples and follows the principle of “your own answer first, the tool second”. Otherwise, AI can become another anchor rather than a lever for change.
- Recognise personal assumptions about how work is performed, including assumptions that may no longer be valid in the age of AI.
- Learn a process for testing assumptions rather than defending them.
- Develop the ability to generate multiple possible approaches to the same task.
- Build confidence when working with AI tools and recognise when asking the tool for confirmation rather than information.
- Learn how to redistribute one specific task between yourself and AI.
- Support practical application – each participant leaves with a personal action plan card.
Autopilot: When It Helps and When It Holds You Back
- Established ways of working as an advantage – and the point at which they become a limitation
- Experiential exercise “Six Requests” – participants work through a series of tasks and then compare how they approached them
- Which of my ways of working have I not reconsidered for a year or more?
- When is it worth switching off autopilot, and when is autopilot actually the right choice?
Assumptions About Our Own Work
- An assumption as an explicit statement: “a person has to do this”, “this is how we have to do it here”
- Where an assumption comes from and what evidence supports it
- Assumptions that used to be valid but are no longer valid following the introduction of AI
- The difference between an assumption and a complaint
Mind Traps When Working with AI
- Anchoring effect: how the first number or first formulation can influence our own judgement
- Confirmation bias when evaluating our own judgement
- Leading questions: when do we ask the tool for confirmation rather than information?
- Why simply being aware of a bias is not enough – and what works better
Testing Assumptions and Exploring Alternative Scenarios
- Consider the opposite: what would have to be true for my assumption to be wrong?
- Asking a specific question instead of simply asking AI to “be objective”
- Using AI as a challenger of our assumptions rather than as a source of answers
- Changing perspective: how might someone with different options and resources approach the same task?
Applying the Learning in Practice
- Three directions for change: what I delegate to AI, who I discuss the task with differently, and how I think about the task
- Tool or agent: what do I keep as my responsibility and what do I delegate?
- Action plan: one concrete change rather than a list of intentions
- Premortem: what could derail the action plan before it even starts, and why asking this question usually makes the plan more robust
- Personal checkpoint: when and with whom I will review and verify the result