What if the most pressing danger of AI is not its ability to replace jobs, as more than one in five U.S. workers fear, but its potential to cause cognitive decline?
Researchers at Microsoft and Carnegie Mellon University published a new study last month that claims to be the first to examine the effects of AI tools on critical thinking.
The researchers found that the more confident human beings were in AI’s abilities to get a task done, the fewer critical-thinking skills they used. Humans confident in AI left critical thinking to ChatGPT instead of doing it themselves and strengthening their cognitive abilities.
Related: Would You Pay $200 for ChatGPT? OpenAI’s New Reasoning Model Has a Hefty Price Tag.
“Used improperly, technologies can and do result in the deterioration of cognitive faculties that ought to be preserved,” the researchers wrote, adding that “a key irony of automation is that by mechanizing routine tasks and leaving exception-handling to the human user, you deprive the user of the routine opportunities to practice their judgment and strengthen their cognitive musculature, leaving them atrophied and unprepared when the exceptions do arise.”
The researchers surveyed 319 knowledge workers, or workers who handle data or information, to find how confident they were in AI’s capabilities and how much critical thinking they employed when using AI to complete tasks. Critical thinking was defined as falling under one of six categories: knowledge (remembering ideas), comprehension (understanding ideas), application (putting ideas to work in the real world), analysis (contrasting and relating ideas), synthesis (combining ideas), and evaluation (judging ideas).
The surveyed knowledge workers used AI like ChatGPT at least once a week and gave 936 examples of how they used AI at work, ranging from looking up facts to summarizing a text. They mainly used critical thinking to set clear prompts, refine prompts, and verify AI responses against external sources.
Six out of the seven researchers listed are associated with Microsoft Research, the research subsidiary of Microsoft created in 1991. Microsoft has deep interests in AI, with its investment in ChatGPT-maker OpenAI totaling close to $14 billion and its plans to spend $80 billion on AI data centers in the fiscal year ending in June.
The researchers caution that while AI can make workplaces more efficient, it could “also reduce critical engagement, particularly in routine or lower-stakes tasks in which users simply rely on AI, raising concerns about long-term reliance and diminished independent problem-solving.”
In other words, AI has a hidden cost: It could lead workers to lose muscle memory for more routine tasks.
