Case study
How to Think: a curriculum for thinking alongside AI
Thinking taught as a practice rather than a reading list: every lesson is one idea with an exercise, a named failure mode, and a step that carries it into the reader’s week.
The problem
AI makes it easy to stop thinking: hand over the question, accept the answer. What is missing is not information about thinking, which is everywhere, but practice in it: noticing what you believe and why, separating signal from noise, deciding what to hand to a machine and what to keep.
Books about thinking are read and agreed with, and nothing changes. The gap between agreeing with an idea and practising it is where the work actually is.
What was built
Every lesson has the same structure: one primitive (the idea in a sentence), a concrete example, an exercise to do now, the failure mode that shows you have only agreed rather than practised, and an integration step that carries it into the following week.
Lessons are nodes in a graph, linked by what each one requires, enables and supports, and sequenced into guided paths. A reader enters through one of three doors (using AI in their own work, getting an organisation ready for it, or the foundational thinking skills underneath both), or through a guide for a specific problem such as brain fog or too much noise.
Each primitive a reader finishes goes into a personal recall deck that resurfaces it later, so the ideas are retrieved rather than re-read. Progress lives in the reader’s own browser: no account, nothing to sign up for.
What changed
A live curriculum anyone can enter from the problem they have today and leave with something to do this week.
There is no learning outcome to report yet, so this case study claims none.
Inside your team
Enablement inside a company fails the same way: training is consumed and agreed with, and behaviour does not change. Designing each unit around an exercise, a failure mode and an integration step is how a team actually adopts a new way of working, including working with AI.