Study finds that using AI chatbots for just 10 minutes may harm people's problem-solving abilities and willingness to persist

Researchers at Carnegie Mellon, MIT, Oxford, and UCLA conducted three experiments with several hundred people each, asking participants to solve problems including simple fractions and reading comprehension. Some participants had access to an AI assistant capable of solving problems autonomously; when the AI was removed, these people were significantly more likely to give up or produce incorrect answers.

The study suggests that widespread AI use might increase immediate productivity while eroding foundational problem-solving skills. Michiel Bakker, an MIT assistant professor involved in the research, notes that AI systems giving direct answers may have different long-term effects than systems that scaffold, coach, or challenge users.

A person's willingness to persist with problem-solving is crucial to acquiring new skills and predicts capacity to learn over time, making the study's findings particularly concerning, according to Bakker.