A new study paints a concerning picture of how artificial intelligence impacts human cognition, suggesting that while AI offers immediate help, it may seriously undermine our long-term thinking abilities and motivation. The research, conducted by a team from the US and UK, is titled AI assistance reduces persistence and hurts independent performance, a name that succinctly summarizes its troubling findings. The paper states that AI assistance improves immediate performance, but it comes at a heavy cognitive cost. Researchers discovered that even brief exposure to AI tools can create a rapid dependency. In their experiments, just ten minutes of using an AI for help made participants reliant on the technology, leading to significantly worse performance and increased rates of giving up once the AI was taken away. The study focused on reasoning-intensive tasks like solving problems, which are common in fields such as writing, coding, and brainstorming. In one experiment, 350 Americans were asked to solve fraction-based equations. Half were given access to a specialized chatbot built on a model like OpenAI’s GPT-5, while the other half worked independently. Midway through the task, the AI assistance was abruptly cut off for the first group. The result was a steep decline in correct answers from the AI-assisted group, alongside many instances of people simply quitting. This pattern, where both performance and perseverance dropped, was confirmed in a larger follow-up experiment with 670 participants. A final test using reading comprehension questions instead of math yielded the same outcome. Rachit Dubey, an assistant professor at the University of California and coauthor of the study, explained the core issue. He noted that once the AI is taken away from people, it is not that people are just giving wrong answers. They are also not willing to try without AI. People’s persistence drops. Dubey extended a specific warning about education, suggesting that rapid deployment of AI in schools could risk creating a generation of learners and people who will not know what they are capable of, and then that will really dilute human innovation and creativity. The researchers compare the effect to the classic boiling frog parable, where a frog does not jump out of water that is slowly heated to a boil. They argue that sustained AI use erodes the motivation and persistence that drive long-term learning. These negative effects accumulate subtly over time, and by the time they are visible, they will be difficult to reverse. It is important to note that this particular study has not yet undergone formal peer review. Furthermore, the research did identify one small positive nuance. Participants who used the AI tool strategically for hints and clarification fared much better after it was removed than those who used it to simply generate answers. This suggests that how one uses AI may influence the cognitive cost. This study joins a growing body of research examining AI’s impact on the mind. Other investigations have found that reliance on AI can increase mental fatigue among full-time workers, a phenomenon some are calling AI brain fry. Paradoxically, some data indicates that employees using AI tools end up working harder and longer hours, perhaps due to increased expectations or the cognitive overhead of managing the technology. The effects appear especially stark in educational settings. Separate studies have linked AI use in schools to poorer social and intellectual development in children, and found that students who rely on chatbots tend to perform worse on tests. The collective findings point to a critical juncture. As AI becomes deeply integrated into work and learning, the research urges a cautious approach, highlighting the need to preserve the human capacity for deep, persistent, and independent thought.
