A new study by scientists at Carnegie Mellon University, Massachusetts Institute of Technology (MIT), Oxford University and University of California, Los Angeles has issued a noteworthy warning about the impact of artificial intelligence (AI) on human thinking abilities.
According to research, just about 10 minutes of using AI to help solve problems can also reduce users' ability to reason independently and make them more likely to give up when there is no longer AI help.
In the context of AI tools such as ChatGPT, Claude or Gemini becoming increasingly popular, this study raises a big question about whether humans are gradually becoming overly dependent on AI.
The research team conducted three experiments with hundreds of participants through an online platform. Participants were asked to solve various types of problems such as basic math and reading comprehension.
Some people are provided with AI assistants capable of directly solving tasks. However, when the AI tool is suddenly removed from the test process, this group tends to quit faster or give more wrong answers than those who have never used AI support.
The results were published in a study entitled "AI Assistants reduce persistence and affect independent performance" on arXiv (operated by Cornell University and considered one of the world's largest research publications in the field of technology and science).
Scientists realize that the problem is not in the use of AI in general, but mainly appears when users let AI completely solve tasks for them.
Meanwhile, the use of AI to suggest, explain or support orientation does not cause a significant decline in thinking performance.
Michiel Bakker, assistant professor at MIT and one of the research authors, believes that AI clearly helps increase productivity in the short term, but humans need to be cautious with the use of this technology.
He emphasized that the goal is not to ban AI in learning or work, but to find ways to use AI so that it still maintains human learning and problem-solving capabilities.
According to Michiel Bakker, the willingness to continue facing difficulties is a very important factor for people to acquire new skills and develop thinking over time. What the research team is concerned about is that many people quickly lose motivation when there is no longer AI support.
MIT researchers also believe that future AI systems should be designed in the direction of "support and guidance" instead of just providing answers immediately.
This can help users both take advantage of the benefits of AI and maintain independent thinking capabilities.
The study also warns that the impact of AI on human cognition can occur silently and accumulate over time, especially as AI is increasingly appearing in learning, work and daily life.
