People that use AI likely overestimate their abilities

When asked to evaluate how good we are at something, we tend to get that estimation completely wrong. It's a universal human tendency, with the effect seen most strongly in those with lower levels of ability. Called the Dunning-Kruger effect after the psychologists who first studied it, this phenomenon means people who aren't very good at a given task are overconfident, while people with high ability tend to underestimate their skills. It's often revealed by cognitive tests that contain problems to assess attention, decision-making, judgment and language.

- WORDS DREW TURNER

But now, scientists at Finland's Aalto University, together with collaborators in Germany and Canada, have found that using artificial intelligence (AI) all but removes the Dunning-Kruger effect - in fact, it almost reverses it. Their research showed that when using common chatbots to solve problems, everyone tended to put too much faith in the quality of the answers, with the most experienced AI users doing so the most.

As we become more AI-literate thanks to the proliferation of large language models, the researchers expected participants to not only be better at interacting with AI systems, but also better at judging their performance in using them. “Instead, our findings reveal a significant inability to assess one's performance accurately when using AI equally across our sample,” said Robin Welsch, an Aalto University computer scientist who coauthored the report.