Is AI Diminishing Our Intelligence?

# ai # research

A recent study from Microsoft and Carnegie Mellon University dives into how using generative AI at work can mess with our critical thinking skills. It's pretty clear from their findings that if we lean too heavily on AI, our brainpower might start to slack off. The paper points out, "Used improperly, technologies can and do result in the deterioration of cognitive faculties that ought to be preserved." What this means is, when we use AI at work, we tend to focus more on just checking if the AI's output is okay to use, rather than really digging deep with our own thinking. This can lead to a drop in using those high-level skills like creating, evaluating, and analyzing stuff. If we only jump in when the AI messes up, we miss out on regular chances to flex our judgment muscles. Over time, these muscles can get weak, leaving us pretty unprepared when the AI can't handle things. In simpler terms, if we let AI do too much thinking for us, we get rusty at solving problems on our own when the AI falls short. The study looked at 319 folks who use generative AI at work at least once a week. They were asked to give three examples of how they use AI, which mostly fell into three buckets: creation (like writing a standard email), information (like researching or summarizing), and advice (like getting tips or making charts). Then, they had to say whether they used critical thinking during these tasks and if using AI made them think more or less critically. They also shared how confident they felt about their own skills, the AI's abilities, and their knack for judging AI outputs. About 36% of the participants said they used critical thinking to avoid any negative consequences from AI use. For instance, one person used ChatGPT to draft a performance review but double-checked it, worried she might accidentally submit something that could get her in trouble. Another guy had to tweak AI-generated emails to his boss to avoid any cultural blunders. And often, people would double-check AI answers with general web searches, which kinda defeats the purpose of using AI in the first place. To make up for AI's weaknesses, workers need to get a handle on what those weaknesses are. But not everyone in the study knew the limits of AI. The paper notes, "Potential downstream harms of GenAI responses can motivate critical thinking, but only if the user is consciously aware of such harms." Interestingly, the study found that people who trusted AI more used less critical thinking than those who trusted their own abilities more. While the researchers stop short of saying that generative AI tools make you dumber, the study does suggest that relying too much on these tools can dull our ability to solve problems independently.

Related article

Comments (20)