Why IT Students Should Struggle More Before Turning to AI
Research from Carnegie Mellon, MIT, Oxford, and UCLA found that even brief AI use measurably reduces students' ability to persist and perform without it — a warning sign for any IT program that hasn't built 'struggle first' requirements into its curriculum.
The Research Case for Struggling on Purpose
If you're studying for a CompTIA Network+ or working through a cybersecurity lab, you've probably used AI to get unstuck. That's not the problem. The problem is what happens when you skip the struggle entirely — and new research makes the cost of that shortcut uncomfortably clear. A joint study from Carnegie Mellon, Oxford, MIT, and UCLA found that even brief AI use reduces persistence and tanks performance on unassisted tasks. Participants got better results with AI in the short term, but performed significantly worse and gave up more easily the moment that support was removed. For IT students, that's not a minor trade-off — that's the difference between passing a certification and actually being able to do the job.
“"Even brief use of AI reduces persistence and impairment of unassisted performance. While AI assistance improved short-term performance, participants performed significantly worse without AI support and were more likely to give up." *— Carnegie Mellon / Oxford / MIT / UCLA joint study, via eCampus News*”
A separate Microsoft-Carnegie Mellon study adds another layer: people who had high confidence in AI-generated outputs showed reduced critical thinking. The inverse was also true — people who trusted their own abilities were more likely to question and push back on what AI told them. In IT fields, that instinct to question matters constantly. You're not just looking for an answer to paste into a terminal; you're diagnosing systems, evaluating risk, and making calls when there's no clean solution in front of you.
What 'Friction Maxxing' Actually Means for Your Coursework
"Friction maxxing" sounds like a TikTok trend, but the concept behind it is grounded in how technical expertise actually develops. The idea is straightforward: you deliberately work through hard problems without AI assistance before you reach for help. In IT education, that might mean spending 20 minutes troubleshooting a misconfig
This issue draws on reporting and data from eCampus News.