AI ate my homework

What should we do about it?

This is the second post in a three-part series on the future of the university. Consider a paid subscription to read all three posts in full.

I’m happy to bet anyone that I can write a PhD in a week.

Maybe not yet a PhD in applied maths or engineering, but almost certainly a PhD in the social sciences. Yes – and I know this will irk many of my colleagues – I’m pretty confident I could produce a dissertation that would pass an Economics examination in less than a week.

And it’s not just me. My students can now turn in passable essays having spent less than ten minutes writing them. If they really want to impress their professor, they might spend a few more minutes (and a couple of extra dollars) refining the text – and walk away with an A. Many university professors are in denial or, worse, oblivious to how rapidly the technology of Large Language Models is evolving.

But let’s be honest: almost all students are now using them. I think I can count on one hand the number of essays I graded this semester that did not have at least some touch of AI. Trying to stop it is pointless. As the economist Tyler Cowen argues in The Free Press ($), the only response is to assume that students now have ‘an invisible, very high-quality helper’. I use it all the time, so why shouldn’t they? In truth, they might already be better at it than I am, and than many of my colleagues. Says Cowen:

I have observed that college and university students, on average, know more about LLMs than do their professors. Much of that knowledge likely comes from cheating (or what we currently regard as cheating) with AI. But it is knowledge nonetheless. And it is knowledge driven by incentives, which can be a pretty powerful method of learning, whether we like it or not.

In early 2024, I wrote about the four horsemen of the apocalypse facing universities: financial, political, ideological and technological. I claimed then that despite universities’ acute financial troubles, technology is by far the most serious threat, even if they did not realise it yet:

Ask university administrators about this, and they are largely ignorant. No, a committee on whether ChatGPT is ‘plagiarism’ is not sufficient to deal with the looming revolution. And, sure, education is not just about imparting knowledge but also about fostering human values, ethical reasoning, emotional intelligence, and interpersonal skills. That sounds wonderful until a good undergraduate student with a laptop and some creativity produces a better research paper than most university professors.

We’ve reached that point faster than I expected. The new ‘deep research’ capabilities of ChatGPT, for example, are astounding. For R400 a month or, if you can afford it, R4000 for the elite model, you have access to a research assistant able to do high-quality research faster than you can write a good prompt. With tools like Cursor, even complex programming tasks now require little more than a vibe. OpenAI CEO Sam Altman recently asserted that GPT-5 will be smarter than all of us. Soon, AI agents will simply run things on their own.

More voices are acknowledging this reality. Here is professor of strategy Scott Latham in The Chronicle of Higher Education ($):

Professors need to dispense with the delusional belief that AI can’t do their job. Faculty members often claim that AI can’t do the advising, mentoring, and life coaching that humans offer, and that’s just not true. They incorrectly equate AI with a next-generation learning-management system, such as Blackboard or Canvas, or they point out AI’s current deficiencies. They’re living in a fantasy.

So what to do about this crisis?