Artificial intelligence has brought this unsettling reality to light. The technology is not the issue. A grading system that was already measuring the wrong things long before ChatGPT existed is the issue with the technology being used to game. Grades have been the focal point of American education for more than a century. Not interest. not skill. Not the slow, annoying, and truly satisfying process of truly grasping something. only grades. That system was not disrupted by AI. It just made it impossible to ignore its emptiness.

Think about what transpired in 2024 at the California State University system. In order to position CSU as the first “AI-Empowered” university system in the country and provide free AI tools to almost 500,000 students, administrators announced a $17 million partnership with OpenAI. The language in the press releases was glowing and enthusiastic. Faculty at several campuses, including San Francisco State, received layoff notices in the same month.

Twenty-three academic programs at Sonoma State were designated for elimination, including physics, economics, and philosophy. For a chatbot, millions. Pink slips for the instructors. It’s difficult not to interpret that as a declaration of the institution’s true values.

The line of reasoning has been well-known. First, faculty members worry about plagiarism. The same professors then start rebranding themselves as “AI-ready educators,” abruptly holding workshops on AI literacy, almost overnight. A resigned embrace replaces the initial fear. Make money off of it if you can’t beat it. The more awkward question—why are students pursuing AI in the first place—is overlooked in that turnabout.

Motivation researchers say the answer is depressingly logical. Using a tool that provides performance without understanding is perfectly reasonable behavior in a system that prioritizes performance over comprehension. Pupils are not defective. They’re reacting rationally to ineffective incentives.

It is clear from decades of psychology research how people truly learn. Struggle is necessary for real learning. It necessitates making errors, sitting in bewilderment, and learning the lesson the hard way. The particular friction of not quite knowing something, reaching for it anyhow, and arriving somewhere new is what researchers refer to as “productive challenge.” AI is designed to instantly remove that friction. Giving that struggle to a machine is about the same as hiring someone else to perform physical therapy for a student who is still developing fundamental skills. The muscles don’t grow. All you get is a printed report stating that they did.

Some educators believe that tougher regulations will address this. Software for detection and bans. blue-book tests. It’s a predictable reaction that will most likely fail. Schools attempted to regulate smartphones, calculators, and the internet. The atmosphere in the classroom became more suspicious and hostile as students seemed obedient and discovered solutions. Curious learners have never come from policing. It’s possible that this strategy will only exacerbate the mistrust that is already permeating campuses.

Students are more engaged when they believe that what they are doing is truly important to them, according to research that has been showing this for years, long before any of this. when they actually have a choice. when errors are not viewed as signs of failure but rather as a necessary part of the process. when educators are aware of their names.

These are not impersonal ideals; specific teaching strategies that promote these conditions have been documented by researchers. None of them neatly fit into a gradebook, which is the issue. Even though the evidence has been there for decades, the gradebook has been in charge for so long that it seems radical to suggest we reconsider.

It’s worthwhile to sit with this larger cultural trend. Our ability to navigate was gradually destroyed by GPS, not improved. We didn’t become more socially adept thanks to social media; rather, it provided us with a completely different approach. It’s possible that the hopeful theory of AI in education—that it will enable students to think more broadly, delve deeper, and learn more quickly—is equally idealistic. What we actually know about the adoption of digital tools by younger generations points to the opposite: they are being used as bypasses rather than as launching pads. That does not imply that the result is certain. However, it does mean that it is not a strategy to just give students AI and hope for the best. It’s a press release disguised as an abdication.

Underlying all of this is a question that has nothing to do with AI. It concerns the true purpose of schools. If credentials—performance proxies that indicate preparedness to institutions and employers—are the answer, then AI is a truly effective solution, and the entire discussion is merely about controlling appearances.

However, the current system was already failing long before a chatbot could compose a five-paragraph essay in four seconds if the solution involves something like actual human capacity—the ability to think, adapt, reason, and create. The defect was not caused by artificial intelligence. It simply made it impossible to act as though the defect didn’t exist.

The post Grades vs. Learning: How Artificial Intelligence is Exposing the Fatal Flaws of Modern Education appeared first on Creative Learning Guild.