网曝吃瓜

We shouldn’t make the mistake of thinking that the use of GenAI for 鈥渃heating鈥� and the use of GenAI for 鈥渓earning鈥� are tied together.

More and more, I am convinced that these are two big separate questions:

1. How can we help learners use GenAI effectively?
   (With a lot of room for what 鈥渆ffective鈥� means.)
2. How can we stop learners from using GenAI to avoid learning or to cheat?

There is evidence in the form of student behavior and student surveys, along with solid reasoning about educational incentives and long-standing problems in academia that are leading me here.

Conflating these questions is one way we talk past each other in discussions about GenAI. If some people in the conversation think these two questions are on the same axis, or tied directly together, the conversation can stall and leave everyone unsatisfied.

For the last year the has been important in helping me understand this moment in higher education. If it suits you, I recommend essentially all the episodes!

If you only have a few minutes, here are two snippets from a recent episode. Both are worth your time.

1. (Timestamp 8:52.)
2. (Timestamp 27:25.)

Here is the full episode:

Over the summer there was a widely circulated story about an economics course at Brown University, with a now-famous chart showing student scores on both a take-home exam and a subsequent in-class exam. From Inside Higher Ed: .

[See the original piece for more details. This is an incomplete screenshot of the plot in that post.]

What you might not have seen was a follow-up piece by Leo Schumann: .

Here are two key snippets, which I hope will motivate you to read the whole piece.

But 鈥渄id students cheat鈥� is the least interesting question this chart raises. This is one course, in one semester, under one assessment format. What we鈥檙e looking at isn鈥檛 a portrait of a generation. It鈥檚 a single, unusually clean picture of what happens when a decades-old incentive structure meets a technology that removes all the friction. And on the question of what that incentive structure does to students, we have data.

And then there鈥檚 the detail I鈥檇 put next to the Brown chart in every faculty meeting in the country.

Some of the students in our focus groups asked their professors to bring back blue-book exams. Not because they wanted to be policed. Because the temptation to use AI, knowing their peers were using it, was so hard to resist that they wanted it removed. Students are asking us to take the temptation away. They aren鈥檛 defending a right to cheat. They鈥檙e telling us, in the plainest language they can find, that the game we built is one that almost no rational person can refuse to play. Indeed, Student 22 from the chart quickly achieved near-mythical status on social media on account of their exceptionality, their apparent refusal to join their classmates in playing the game.

If we hear the students asking for blue-books and take that as their verdict on GenAI overall, we have once again merged the two questions.

We have to address good use of GenAI (effective/ethical/professional) separately from the prevention of bad use (ineffective/unethical/professionally detrimental). For some assignments, learning objectives, courses, or programs the main need is to understand how GenAI is being used in the field, and bring that to our students. For others, the main need is to find truly valid assessments that evaluate the student on relevant skills and abilities.

And, for many contexts at 网曝吃瓜, both will need attention and effort. Explicitly dividing this labor among colleagues and leaders will be key.

The chart image comes from Emma Whitford, 鈥�,鈥� Inside Higher Ed, July 8, 2026.

Generative AI disclosure:
I used Claude to get feedback on drafts of this piece and to format it for WordPress. After writing this piece I used generative AI to write a first draft of the short 鈥渢easer blurb鈥� that went out by email. Want to know more? Send me an email and we can chat.