Study Warns Using AI For Just 10 Minutes May Reduce Your Ability To Think Independently

Artificial intelligence tools are quickly becoming part of everyday life. Need help writing an email? AI can do it. Stuck on homework or brainstorming ideas for a project? Chatbots like ChatGPT, Claude and Gemini are always ready with answers in seconds. In many ways, these tools save time and reduce effort. But according to a new research paper, that convenience may come with a hidden downside. A recent study titled AI Assistance Reduces Persistence and Hurts Independent Performance suggests that relying on AI, even briefly, could affect how people think and solve problems on their own.What The Study FoundThe research was based on experiments involving 1,222 participants across three large studies. Researchers tested people on maths and reading

comprehension tasks through an online platform. Some participants were allowed to use AI assistants that could directly solve questions. Others had to complete everything without AI support. Obviously, the AI group looked stronger. They solved questions faster and scored better during the assisted phase. But the real surprise came later.Once researchers removed AI access, the participants who had depended on chatbots struggled noticeably more compared to those who had worked independently from the beginning. In the maths-focused test, participants who never used AI solved nearly 73 per cent of problems correctly. Those who earlier relied on AI dropped to around 57 per cent accuracy. Reading comprehension showed a similar pattern. The non-AI group scored 89 per cent, while AI-assisted users managed only 76 per cent afterward.The Bigger Problem Was PersistenceResearchers say the concern goes beyond lower scores. People who relied heavily on AI became more likely to skip difficult questions or simply stop trying once challenges appeared. In simple words, they lost patience faster. “AI assistance improves immediate performance, but it comes at a heavy cognitive cost: after just ~10 minutes of AI-assisted problem-solving, people who lost access to the AI performed worse and gave up more frequently than those who never used it,” the paper reads. The researchers described this as a possible “boiling frog” effect. One or two AI interactions may not seem harmful, but repeated dependence could slowly reduce a person’s willingness to think deeply or struggle through problems independently.Meet The AI Robot That Can Cook Meals, Solve Rubik’s Cubes And Play PianoNot All AI Usage Is HarmfulInterestingly, the study also found an important difference in how people used AI. Participants who asked AI tools for direct answers showed the biggest decline in independent performance. However, people who mainly used AI for hints, explanations, or guidance did not experience the same level of negative effects. That detail matters. According to the researchers, AI systems designed more like tutors or coaches may actually help learning, while tools that instantly complete tasks could encourage mental shortcuts over time.So, the issue may not be AI itself, but how comfortably humans start leaning on it.