Jeff Sharlet

A little bit of a tangent, because I don’t much care about cheating. Were it up to me I’d slash the amount of cheating massively by getting rid of the pedagogically useless incentive to cheat: grades.

Earnest q: if the surveys haven’t been good, how can the evidence be “enough”? Do they actually have access to the test keys maintained by some Greek orgs? What incentive does one who has cheated have to be truthful? How can they trust confidentiality?

A colleague once called me a “neoliberal pimp” for helping my students publish & get journalism jobs. I work hard at that. At the same time, I make clear in my courses that creative writing isn’t vocational education, as valuable as such education is.

Keep an eye out for a theater

I interviewed the director, if you're interested. www.eyeforfilm.co.uk/feature/2026...

Oh very! Thanks

Natchez | 2025 Documentary From Suzannah Herbert After generations of showcasing its antebellum homes and hoop-skirted docents, Natchez, Mississippi, is now reckoning with romanticized past, an uncertain future and the debt it owes to the descendant...

There's a list of theatre screenings here:

www.natchezfilm.com

I just saw Natchez, a new doc I’ll recommend again once I figure out how to speak to the deep intelligence of its structure, the way it moves from a cloying world you know is actually horrific into a realer world in which the horror, courage, witness, and humanness are all made visible.

Do we have any evidence that there is more cheating now? How would one study the ubiquity of cheating? Do we decontextualize it from structural corruption?

Which is reductive. But not nearly as much as the AI debate would have us believe. This is why I agree with @kylesaunders.bsky.social that AI is an accelerant, not a cause. Gas on a tire fire that’s been cooking so long some didn’t realize it really could burn the whole thing down.

I’d say the “us” there is messy. When I applied for my job, I made clear what, why, and how I teach. That’s my contract. But I don’t run admissions, or marketing, or the decisions higher up. Colleges can decide whether they want to be learning institutions or class reproduction machines.

Oh we agree about that! I didn’t take it amiss. Thanks for writing!

Heck, how about we ditch competition.

And that’s the very story on which Lapham taught me that lesson! I had written what’s basically the first chapter. But what to do about it?! I didn’t know. How could I publish without a solution?! Lapham said doing so was the point. So he was right about that.

I definitely agree, tho, that it’s an “accelerant not a cause.” I don’t think his description of the pre-existing crisis goes broad enough.

Fair, but I think my argument is w/ his account of the university. Terms such as “mastery” “debate” & thus the “certification” which he critiques describes to me one idea of what it is & shld be, but does so as if that’s a fact. Then there’s the conflation of AI w/ “idea,” & the absence of “labor.”

We may disagree — as a describer by trade (a journalist) — I think it’s never of more than limited use. In this instance, I know of many “middle grounders,” against whom
I’m arguing, that I know also to be brilliant in their spheres, progressive or even liberationist in politics, & in good faith.

Then I think of the many students who know classmates are raising hands with AI-produced comments. Nobody wants to be a rat. So they’re coerced into complicity. Some, meanwhile, came for the discussion; they’re robbed. This isn’t a problem of certification, it’s one of community.

I’m thinking of a student who used AI for a test, because other students were; this student knew they could do well on their own, but knew their classmates, with AI, were going to do better. The problem I see there is competition and its metaphor, grades. AI is just an acceleration of that problem.

When I talk to students who’ve used AI to cheat (on condition of confidentiality) most describe it as an experience into which they feel pushed, one they don’t like. Yes, some want shortcuts; but most think it would be better to learn, if there was time or circumstance.

If one implicitly understands one’s role as a teacher as a kind of agent for employers—vetting future applicants—then grades, & the way AI demolishes whatever shred of meaning they had, is the problem. But I think that misses a deeper problem, with more interesting possibilities…

Parallel to this, I think, is a common misunderstanding of the problem of cheating in education connected to AI. I don’t think it’s in essence a problem of certification, since cheating, & corruption in the form of grade inflation (or, perhaps, “grades” themselves) is already widespread.

Absolutely. He was all of those things, and antisemitic, too. *And* he was right that journalists don’t need to be problem solvers. That understanding was why Harper’s back then was so much more interesting than The Atlantic, the wonkery of which drew it into status quo complicity.

I see the parallel you draw there, but no, I don’t accept your category of “this lot.” The “middle ground” AI faction is extremely diverse and often surprising. Blunt categorical thinking is at odds with description. (Also, I’d argue, a mistake being made by AI advocates.)

AI is coming so fast—& in midst of a literary & media collapse+scholarly crisis it accelerates—that thisrough division of labor (description and solution), never stable, has largely been lost. Leading some solvers, I think, toward premature “solutions” in the form of acceptance of inevitability.

I’m deeply grateful to Lewis Lapham who long ago helped me see past my “conclusion” trouble with my first feature: my job, he said, was to describe the problem. That can be enough. To describe it without solution, which leaves the door open for problem-solvers.

A lot of the people making these arguments are very smart folks, educated in varieties of problem-solving. Maybe if they can’t see a solution, they default to “it can’t be a problem.” The “it” being AI itself; instead, they can bite off bits of it, solve those problems, & see the rest as inevitable.

That’s a monolithic view of a complex org.

Too many “middle ground” AI arguments—“I have concerns, too, but we have to adapt”—proceed from what is to me a peculiar embrace of “inevitability” which seems to be magical thinking, a way of depoliticizing the political, of self-soothing in the face of an overwhelming challenge.

With respect and recognition that he may be describing other situations, this is a clumsy caricature of the critique I share with colleagues. That may be specific to our discipline, but I don’t care at all about “certify” or “signal.” My critique is practical—I teach practice—and political.