Long before artificial intelligence, people were already using tools to support their thinking. We write notes to remember information, use calculators to perform complex calculations, and rely on maps or navigation systems to find our way. Cognitive scientists describe this tendency as cognitive offloading: the use of external resources to reduce the mental effort required to complete a task (Risko and Gilbert, 2016).

Recent advances in generative artificial intelligence have expanded this pattern into new territory. Tasks that once required significant time, training, or specialised knowledge can now be completed with increasingly sophisticated tools. Reports, presentations, analyses, and recommendations can be generated in minutes, often in a form that appears polished and professional.

For much of modern professional life, these outputs acted as useful signals of expertise. A detailed report or carefully argued recommendation often reflected a combination of knowledge, experience, and effort. As professional-quality outputs become easier to generate, that relationship becomes less straightforward.

So, if expertise can be convincingly simulated, what signals of expertise are we actually responding to?

To answer that question, it helps to ask what expertise actually is. In everyday conversation, expertise is often treated as a synonym for knowledge. Yet research defines it as a practice that develops through repeated engagement with a domain, where experience and feedback gradually shape judgment (Kahneman and Klein, 2009).

This is one reason why expertise cannot be reduced to information alone. A person may possess facts, concepts, or technical language without necessarily knowing how to apply them in real-world situations. Expertise involves recognising patterns, interpreting context, and understanding when established rules may not apply.

In many professional settings, the greatest value of expertise lies not simply in producing answers, but in evaluating them. Recognising what is missing, what assumptions are being made, and where uncertainty remains requires a depth of judgment that extends beyond the information itself (Ericsson et al., 1993; Kahneman and Klein, 2009).

Research suggests that people often believe they understand a subject more deeply than they actually do. Psychologists Leonid Rozenblit and Frank Keil (2002) described this as the illusion of explanatory depth: the tendency to feel that we understand how something works until we are asked to explain it in detail. A clear explanation can create a strong sense of understanding, even when our own knowledge remains limited.

This tendency becomes particularly relevant as increasingly sophisticated tools support cognitive work. While cognitive offloading can reduce mental effort and improve efficiency, it can also make it easier to accept explanations without examining the reasoning behind them (Risko and Gilbert, 2016). So, at what point does cognitive offloading begin to include judgement itself?

Research on automation has similarly shown that people may place greater trust in recommendations generated by technology than they intend, particularly when those recommendations appear competent or authoritative (Parasuraman and Riley, 1997).

Taken together, these findings suggest that the challenge is no longer simply accessing information but evaluating it. The appearance of understanding can be surprisingly convincing, making careful evaluation an increasingly important part of professional judgment. If the way knowledge is produced is changing, then the role of expertise is changing with it. 

This does not diminish the value of new technologies. They continue a long history of tools that have made work faster, more accessible, and more efficient. What changes is not the need for expertise, but where expertise adds the greatest value from now onwards.

Increasingly, professional judgment involves asking different kinds of questions. What evidence supports this conclusion? What assumptions does it rely on? What information is missing? Under what circumstances might a different conclusion be reached? These questions move beyond the quality of an answer to the quality of the reasoning behind it.

Seen in this way, discernment is not about resisting technology or becoming more sceptical of every new tool. It is about recognising that as information becomes easier to generate, judgment becomes increasingly valuable. The ability to question, interpret, and place information in context may become one of the defining professional capabilities of the years ahead.

​

 

 

Reference List:

 

1. Ericsson, K.A., Krampe, R.T. and Tesch-Römer, C. (1993) ‘The role of deliberate practice in the acquisition of expert performance’, Psychological Review, 100(3), pp. 363–406. https://doi.org/10.1037/0033-295X.100.3.363.

2. Kahneman, D. and Klein, G. (2009) ‘Conditions for intuitive expertise: A failure to disagree’, American Psychologist, 64(6), pp. 515–526. https://doi.org/10.1037/a0016755

3. Parasuraman, R. and Riley, V. (1997) ‘Humans and automation: Use, misuse, disuse, abuse’, Human Factors, 39(2), pp. 230–253. https://doi.org/10.1518/001872097778543886

4. Risko, E.F. and Gilbert, S.J. (2016) ‘Cognitive offloading’, Trends in Cognitive Sciences, 20(9), pp. 676–688. https://doi.org/10.1016/j.tics.2016.07.002

5. Rozenblit, L. and Keil, F. (2002) ‘The misunderstood limits of folk science: An illusion of explanatory depth’, Cognitive Science, 26(5), pp. 521–562. https://doi.org/10.1207/S15516709COG2605_1