Editorial written by Bloomberg Opinion Editorial Board.
For many college students these days, life is a breeze. Assignments that once demanded days of diligent research can be accomplished in minutes. Polished essays are available, on demand, for any topic under the sun. No need to trudge through Dickens or Demosthenes; all the relevant material can be instantly summarized after a single chatbot prompt.
Welcome to academia in the age of artificial intelligence. As several recent reports have shown, outsourcing one’s homework to AI has become routine. Perversely, students who still put in the hard work often look worse by comparison with their peers who don’t. Professors find it nearly impossible to distinguish computer-generated copy from the real thing — and, even weirder, have started using AI themselves to evaluate their students’ work.
It’s an untenable situation: computers grading papers written by computers, students and professors idly observing, and parents paying tens of thousands of dollars a year for the privilege. At a time when academia is under assault from many angles, this looks like a crisis in the making.
Incorporating AI into college curricula surely makes sense in many respects. Some evidence suggests it may improve engagement. Already it’s reshaping job descriptions across industries, and employers will increasingly expect graduates to be reasonably adept at using it. By and large this will be a good thing as productivity improves and innovation accelerates.
But much of the learning done in college isn’t vocational. Humanities, in particular, have a higher calling: to encourage critical thinking, form habits of mind, broaden intellectual horizons — to acquaint students with “the best that has been thought and said,” in Matthew Arnold’s phrase. Mastering Aristotle or Aquinas or Adam Smith requires more than a sentence-long prompt, and is far more rewarding.
Nor is this merely the dilettante’s concern. Synthesizing competing viewpoints and making a considered judgment; evaluating a work of literature and writing a critical response; understanding, by dint of hard work, the philosophical basis for modern values: Such skills not only make one more employable but also shape character, confer perspective and mold decent citizens. A working knowledge of civics and history doesn’t hurt.
For schools, the first step must be to get serious. Too many have hazy or ambiguous policies on AI; many seem to be hoping the problem will go away. They must clearly articulate when enlisting such tools is acceptable — ideally, under a professor’s guidance and with a clear pedagogical purpose — and what the consequences will be for misuse. There’s plenty of precedent: Honor codes, for instance, have been shown to reduce cheating, in particular when schools take them seriously, students know precisely what conduct is impermissible and violations are duly punished.
Another obvious step is more in-class assessment. Requiring students to take tests with paper and pencil should not only prevent cheating on exam day but also offer a semester-long incentive to master the material. Likewise oral exams. Schools should experiment with other creative and rigorous methods of evaluation with AI in mind. While all this will no doubt require more work from professors, they should see it as eminently in their self-interest.
Longer-term, technology may be part of the solution. As a Bloomberg Businessweek investigation found last year, tools for detecting AI-generated text are still imperfect: simultaneously easy to evade and prone to false positives. But as more schools crack down, the market should mature, the software improve and the temptation to cheat recede. Already, students are resorting to screen recordings and other methods of proving they’ve done the work; if that becomes customary, so much the better.
College kids have always cheated and always will. The point is to make it harder, impose consequences and — crucially — start shaping norms on campus for a new and very strange era. The future of the university may well depend on it.
