Researchers from Microsoft and Carnegie Mellon University recently conducted a study to explore how using generative AI—like chatbots or automated writing tools—at work impacts our ability to think critically.
The researchers pointed out a concern: if we don’t use these technologies correctly, they can actually harm our thinking abilities. They found that when people depend heavily on generative AI for their tasks, their focus shifts. Instead of using advanced critical thinking skills to create, analyze, and evaluate information, they start to spend most of their time just checking if the AI’s answers are acceptable. This means that if employees only pay attention when the AI fails to do its job well, they miss out on valuable chances to practice their own judgment skills.
To put it simply, relying too much on AI could make us worse at solving problems on our own. If the AI sometimes gives incorrect or unhelpful answers, we might struggle to handle those situations since we haven’t been practicing our critical thinking.
In their study, the researchers surveyed 319 people who used generative AI at least once a week in their jobs. They asked these individuals to share three ways they use AI, which generally fell into three types:
- Creation – like writing standard emails.
- Information – such as researching a topic or summarizing long articles.
- Advice – asking the AI for guidance or creating charts from data.
They wanted to know if people were actually exercising their critical thinking skills while doing these tasks, and whether using AI made them think more or less. Participants also provided their confidence levels in themselves, the AI, and their ability to judge the AI’s responses.
Interestingly, about 36% of those surveyed admitted they used critical thinking to avoid possible mistakes when using AI. For example, one woman said she used an AI tool to write a performance review but double-checked its work because she was worried it might lead to errors that could get her in trouble. Another person mentioned having to adjust AI-generated emails to better suit his boss’s preferred communication style, which values hierarchy and experience. Many participants mentioned that they often double-checked the AI’s work by performing additional searches on the internet using resources like YouTube and Wikipedia, which kind of defeats the purpose of using AI in the first place.
Read Also: Abia: Radisson Blu To Upgrade Enyimba Hotel
To effectively cope with the limitations of generative AI, workers need to grasp how these shortcomings arise. However, not all participants were fully aware of the constraints of AI technology.
The researchers emphasized, “Being aware of the potential downsides of AI can encourage critical thinking, but only if users recognize these risks.” The study also found that participants who trusted AI more—believing in its capabilities—actually put in less critical thinking effort compared to those who were more confident in their own skills.
While the researchers hesitated to claim that generative AI makes us less intelligent, the findings suggest that becoming too reliant on these tools could weaken our ability to tackle problems independently.
