Generating ideas is no longer the hardest part of finding solutions. AI models can produce dozens of suggestions within seconds, all equally fluent and confident. The focus is therefore shifting towards what only a human with relevant expertise can assess – which ideas make sense and are worth testing. This training shows how AI can support the generation, sorting and selection of ideas for implementation and testing.
- Mastering a process for using AI to move from a challenge through generating ideas to making a decision
- Expanding the range of possibilities in participants’ own work, including options that become feasible with AI
- Strengthening the ability to recognise when generated ideas start to become too similar and knowing how to address this
- Increasing the contribution of participants’ own expertise to the final result
- Improving the selection process when choosing from a large number of ideas based on predefined criteria
- Supporting practical application: each participant leaves with an action plan for testing one specific idea
The Opportunity Space and How AI Changes It
- The number of ideas generated without AI and with AI
- The difference between having a large number of ideas and having genuinely different ideas
- Why suggestions generated by different people using the same tool tend to become similar, and how the first ready-made solution can influence further exploration
What Has Become Possible in My Work Since This Tool Exists
- Breaking down your own work into individual steps as a starting point for exploration
- The difference between asking “What could be different?” and “How could we do it?” – and why these questions should not be addressed simultaneously
- Human domain expertise – what can only be assessed by someone who actually does the work
Generating and Sorting Ideas with AI
- Sequence and roles: why your own list of ideas should come before the AI-generated list, and where a co-creator is more valuable than an editor
- Domain expertise – undocumented knowledge, constraints and approaches that have already been tried and did not work
- Using multiple independent conversations for the same task to keep ideas diverse
- Confidentiality boundaries: what aspects of your own know-how should not be entered into an AI model
- How to sort ideas and create categories from a large number of suggestions
From an Idea to a Prototype
- Why an idea is easier to evaluate through a prototype or example than through a description
- The rule of three to four examples side by side instead of one fully developed solution, and getting a colleague’s perspective on them
- Using AI to analyse reactions from customers, colleagues or managers and turn them into questions for improving ideas
Selecting One Idea from Many
- Why selection becomes a weaker point than generation when ideas can be produced with almost no effort
- Defining criteria before making the selection, rather than afterwards
- Defining criteria and applying them with the help of AI
- Evaluating ideas that are written with equal fluency and persuasiveness, with AI taking on the role of a critic
From the Selected Idea to the First Practical Step
- An action plan: what I will test, on what, by when and how I will know whether it worked
- When an idea goes beyond an individual’s decision-making authority – budget, a new system or someone else’s approval
- Taking the idea forward: who the proposal belongs to, what to communicate and what supporting information to provide