When it comes to large language models, two camps emerge: those who keep a critical eye and those who outsource their thinking.

Researchers at the University of Pennsylvania have termed this latter group as ‘cognitive surrenderers’, where individuals accept AI answers without scrutiny, much like relying on an infallible oracle. But can we really trust these algorithms with our reasoning?

The study suggests that people are more likely to surrender their cognitive powers when time is tight or if external incentives prompt them to do so. In such cases, the seamless and authoritative delivery of AI responses can make them irresistible.

By defaulting on critical thinking, users might be undermining their own analytical skills. The researchers argue that this shift could redefine how we approach decision-making in a world increasingly dominated by AI. Could this trend lead to a collective intellectual decline?