Structured Summary: Impact of GenAI on Critical Thinking
A group of Microsoft researchers conducted a survey-based study of knowledge workers, which suggest that those who are more self-confident and self-reflective, show more evidence of critical thinking, while those who are more confident of generative AI's capabilities enact less critical thinking.
Specifically, higher confidence in GenAI is associated with less critical thinking, while higher self-confidence is associated with more critical thinking.
Also, the nature of critical thinking when using genAI shifts towards information verification and project-management aspects.
I ran this paper with the 2nd version of CustomGPT and the GEM, and the output from both are linked below:
In both cases I asked for a detailed explanation of table 4. The output from Gemini (2.5 Flash) was confusing and incorrect, while that from the CustomGPT seemed easier to understand and fairly accurate. I have not used Gemini 2.5 Pro for analyzing this paper. I liked how Gemini provided a more detailed response to each question. GPT responses were more high-level and concise. But based on the performance on the table interpretation, this round goes to ChatGPT as well.