AI · Culture · Responsibility
Artificial intelligence and critical thinking: what to delegate
AI does not automatically make us more skilled or less human. It amplifies how we think and work. Alongside what it can do, we should ask what we want to keep knowing how to do ourselves.
AI amplifies
Method, culture and intention shape the quality of what we get.
Verification is human
A confident answer can still be weak in its facts and sources.
Responsibility cannot be automated
The system suggests; a person understands, chooses and answers for the result.
Resistance to AI is often a question of identity
When a technology enters a profession, people fear more than learning a new tool. They may fear losing the value of their experience, control of the process or trust built over time.
Dismissing this resistance as backward thinking is a mistake. Useful training starts there: it recognises what needs protecting and shows where AI can ease work while preserving its meaning.
Delegating a task does not mean delegating judgement
We can give a model a first draft, classification, transcription or search for alternatives. Responsibility for the result cannot automatically follow. The model does not truly know the consequences, human context or cost of an error.
For every use, it helps to distinguish what can be automated, what needs supervision and what must remain a human decision. The greater the effect on rights, reputation, learning or relationships, the stronger the verification must be.
Plausible does not mean true
Generative models produce convincing language even when their data is incomplete. Fluent prose can hide an error better than an uncertain answer.
Critical thinking means returning to primary sources, distinguishing facts from interpretations, checking dates and context, and stating uncertainty. Asking a model to check itself may help find alternatives, but it is not independent verification.
- Check names, numbers, quotations and references.
- Compare genuinely independent sources.
- Disclose when content is generated or substantially altered.
The skill of the future is building the relationship
Using AI well means designing a dialogue: giving context, assessing alternatives, correcting, stopping and recognising when the tool is unsuitable. It means going beyond the first output and choosing which developments to follow.
Technology changes quickly; the ability to frame problems, understand consequences and take responsibility for choices remains. This is the focus of my training for teachers, professionals, creatives and organisations.