A new academic study says most of the cognitive skills computing workers rely on today will matter less within a few years - except for one.
Researchers ran a mixed-methods study surveying computing students and industry experts about which cognitive skills matter in the past, present, and future as AI spreads through software work. Participants rated skills like memorization, analysis, and problem decomposition, and largely predicted their importance would decline as AI absorbs more of that cognitive labor. Critical thinking was the exception: students and experts alike said it will stay just as important, or more so, even as AI handles more of the routine thinking. The researchers also interviewed participants about why they expect this shift and how computing students should prepare for it.
That matters beyond academia. It's a signal for how computer science programs and hiring managers should recalibrate what they teach and test for. If the people closest to daily coding work already sense that rote technical skills are a depreciating asset, curricula built around memorizing syntax and following fixed procedures may be training students for a job market that's already changing shape.
The finding echoes a pattern showing up across white-collar work since generative AI tools arrived: routine execution is getting cheaper, and judgment is getting more expensive.