Using AI, Like ChatGPT, Damages Critical Thinking: Study

0
59
Using AI, Like ChatGPT, Damages Critical Thinking: Study

What if the most urgent danger of AI is not the ability to replace jobs because more than every fifth US worker fears, but his potential to cause a cognitive decline?

Researchers from Microsoft and Carnegie Mellon University published a new study last month, in which it is claimed to be the first to examine the effects of AI tools on critical thinking.

The researchers found that the self -confident people in AI's ability to do a task, the less critical skills that they used. People who are confident have a critical thinking, instead of doing it themselves and strengthening their cognitive skills.

Relatives: Would you pay 200 US dollars for Chatgpt? Openai's new argumentation model has a high price.

“Inappropriately, technologies will lead to the deterioration of cognitive skills that should be preserved” they should withdraw the routine opportunities to practice their judgment and strengthen their cognitive muscles, which means that they are switched on and unprepared when the exceptions arise. “

The researchers surveyed 319 knowledge workers or employees who deal with data or information to determine how secure they were in the skills of AI and how much critical thinking they used when using AI to fulfill tasks. Critical thinking has been defined as one of six categories: knowledge (memory ideas), understanding (understanding ideas), application (ideas for working in the real world), analysis (contrast and connection with ideas), synthesis (combination of ideas) , and evaluation (assessment ideas).

The knowledge workers surveyed used AI like Chatgpt at least once a week and gave 936 examples of how they used AI at work, from the search for facts to the summary of a text. They mainly used critical thinking in order to determine clear input requests, to refine the input requests and to check the AI ​​answers based on external sources.

Six of the seven researchers listed are connected to Microsoft Research, the research daughter of Microsoft created in 1991. Microsoft has profound interests in AI. The investment in Chatgpt -Maker Openaai is almost 14 billion US AI data centers in the financial year in June.

The researchers warn that AI can make jobs more efficient, but can also reduce critical commitment, especially for routine or lower tasks, in which users are simply dependent on AI, which are concerned about the long-term trust and a reduced independent Problem solving. “

In other words, AI has hidden costs: It could lead to workers losing muscle memory for routine tasks.

Relatives: Deepseek Ki costs less than 6 million US dollars for development. Here is why Meta and Microsoft spend billions.