Althouse: "Professors like myself hate ChatGPT and similar platforms because our students turn in artificially generated, robotic papers. But..."

August 1, 2025

"Professors like myself hate ChatGPT and similar platforms because our students turn in artificially generated, robotic papers. But..."

"... if we ordinarily gave vapid, shallow papers the D’s or F’s they deserved, this problem wouldn’t exist. The fact that such papers routinely get A’s or B’s shows that we have come to expect and to train humans to write robotic papers. Similarly, when I worry I can’t distinguish a colleague’s genuine sentiments from the vaporous generalities Gmail’s AI suggests, what am I really worrying about? Is it that the machine is so good? Or that my interactions with my colleague are so empty? Once we step back from the paranoid reaction, the problem presented by AI facial recognition assumes different contours. In posing anew the question of facial control, the technology provides us with an opportunity to think about how such control works in both its artificial and natural forms...."

Writes Michael W. Clune, in "Your Face Tomorrow/The puzzle of AI facial recognition" (Harper's)(Harper's gives you 2 free articles a month, and I used one of mine to read that. I doubt that you'll find 2 better choices and recommend that you go ahead and redeem your freebie on the first of the month).

27 comments:

Dave Begley said...: At the Eighth Circuit Judicial Conference, a lawyer drafted a cert petition in ten minutes.; 8/1/25, 6:43 AM
BUMBLE BEE said...: Zappa was prescient?
"Who Are The Brain Police?"
It is shaping up!; 8/1/25, 6:57 AM
n.n said...: A correlation with finite degrees of freedom decoded by automated Intelligence for a progressive model of automatons hoping for change with benefits.; 8/1/25, 7:02 AM
FormerLawClerk said...: Here's the problem with judging people in this day and age (and by judging people, I mean giving honest grades): Someone will come along and count who you gave all the As to and Bs and Cs and Ds and Fs. Then they'll compare these groups for racial disparity.

They will find that you give lower marks to black people and then BAM, you can now not pay your mortgage.

It's simply too monumentally dangerous to give honest grades because our society is rewarding fucked up people who cannot accept real data.

Imagine one day: you lose not only your job, but your entire career that you had to pay monumental sums of money to achieve. You lose your house. Your car is repossessed. You can no longer send your children to child care (which now costs 22% of your income.) Your family is now homeless. Your wife is going to ditch you and take half your shit, and get maintenance payments that amount to modern-day slavery.

Or you could just give everyone Bs and none of that happens.

Different races have different IQs. That doesn't mean some races are inferior. Blacks do far better in the NFL than whites do. Doesn't mean whites are inferior. Alternatively, blacks do less-well in physics than whites.

We cannot accept that. So everyone gets Bs.; 8/1/25, 7:07 AM
JRoj said...: 44% of papers turned in over my summer class were clearly AI generated. They were well written, but never referenced the core course materials necessary to address the prompts in the assignments. Rather than accuse the students of using AI, they were told to rewrite the papers to incorporate the requisite material. It worked. Slightly poorer writing with much improved content. AI can’t reference what it does not have in its data.; 8/1/25, 7:20 AM
Mike (MJB Wolf) said...: I might click through to read later as I am titillated, as in mildly tickled, to see how the writer connects the headline of the article with the post title/excerpt that Althouse selected. Before I could get to the formulaic set-up ("provides us with an opportunity to think about") I tripped over the disentangling of "vapid, shallow papers...that routinely get A’s or B’s."

Whoa. Wait. Why? Why do poorly written papers deserve such high marks? This is more of the Participation Trophy mindset where every tiny ballhandler gets a tiny prize just for showing up, not for showing excellence.

Demonstrating competence in discussing a subject is the purpose for long-form answers. Writing reveals thinking. Lack of writing or verbal skills (one can dictate term papers now easily as well) indicates a lack of critical thinking skills and if the students can't write then they also can't distinguish crap AI writing from good stuff.

Education is all the way down the toilet apparently and well on its way to the wastewater treatment plant if this professor's thoughts are anywhere near the mainstream.; 8/1/25, 7:25 AM
Mr. T. said...: Wait.

Students still have to write papers in college???
Whatever happened to just sitting on the grass, petting your stuffed rainbow unicorn toy, while listing all 3 or 4 ways you have white privilege for an A whilst taking Lauren Duca JournoSocialJusticeisming 101 class?; 8/1/25, 7:27 AM
WK said...: “Professors like myself” are grading the papers. Huh.; 8/1/25, 7:55 AM
Mary E. Glynn said...: Nope, you're the early adapter, ann.
You stick your face in the machine, ask it your questions until you manipulate the prompts that gives you the answer you want...; 8/1/25, 7:55 AM
RCOCEAN II said...: "I went to my department’s annual holiday party, at the home of the chair."

Good to know his Chair, isn't like all those lazy desks and tables and saved up to own a home. And what "Holidays" does its celeberate? Ugh. Guess Harper's cant say the word....shhhh "Christmas".; 8/1/25, 7:57 AM
Mary E. Glynn said...: I doubt that you'll find 2 better choices and recommend that you go ahead and redeem your freebie on the first of the month)
-----------
Ah, the arrogance of a tenured DEI professor who doesn't read much... Keep your "freebies". We've got libraries, ann.; 8/1/25, 7:59 AM
RCOCEAN II said...: AI wants your face brightly lit and unsmiling. It measures the distance between your eyes and other parts of your face, and then goes to a "Database" to get a match. And then there's this:

"Given adequate lighting and a relatively clear picture, software can match your passport photo with your appearance in a party photo on a friend’s Facebook page, for example, 99 percent of the time."

So the AI can easily create pictures and drop you in and no one would know the difference. Scary.; 8/1/25, 8:01 AM
RCOCEAN II said...: "When we consider the way AI makes use of faces, we inevitably contrast it with the way humans make use of faces. The flip side of the paranoid revulsion at AI is an idealized, even romantic sense of the comforting familiarity of human face-to-face interactions. As with so many situations in which we confront new technology, our tendency is to project our fears onto the new thing and to cling to the old, natural way."

Wonder if this was written by an AI? Verbose and says little.; 8/1/25, 8:02 AM
Jersey Fled said...: I’m glad I’m not teaching anymore.; 8/1/25, 8:06 AM
Aggie said...: How come they're not using A.I. to grade papers yet? This sounds like the age-old arms race between cops with radar guns and drivers with radar detectors. Both were made by the same company, in Connecticut - where radar detectors were illegal. One technology would advance, the other would catch up......; 8/1/25, 8:31 AM
Tina Trent said...: This article reminded me a lot of the article by that woman in the NYT magazine who wrote a granular mininovel about what is wrong with men who won't commit to her and her peers. In fact, they are both written in a pattern I've come to associate with AI: each paragraph reiterates the one before it and anticipates the one after it, just barely moving the story along, looping slightly back and slightly forth. A real slog and featureless read.

In college I had a brilliant professor who could actually teach people how to write. He combined a nebulous, over-riding idea -- voice -- with imitating and memorizing and analyzing the voices of the real canon, non-stop writing, and structured workshopping. It was like writing boot camp. Wake me up when AI can achieve that.; 8/1/25, 8:59 AM
Mike (MJB Wolf) said...: RCOCEAN II said, "AI wants your face brightly lit and unsmiling."

Ah yes I recall an appearance of the term "ring light" earlier this week in which I pointed out the shadowless lighting was the result. Should have known AI as behind it.

And whoever said it above, I also am glad I am no longer teaching. Now I can split infinitives at will.; 8/1/25, 9:17 AM
Peachy said...: Kamala and Biden lead the way... to vapidity on display.; 8/1/25, 9:18 AM
Mike (MJB Wolf) said...: WK said, "“Professors like myself” are grading the papers. Huh."

So no one has fed Strunk and White's Elements of Style into AI yet, teaching it to be concise and clear: "Professors like me are..."

OTOH can you imagine coaching AI towards "active writing" when it already has fabulist and aggressive tendencies?; 8/1/25, 9:24 AM
Tina Trent said...: As a graduate student instructor, I gave one student an F. She nearly never turned in work, all of it illiterate; skipped ever other class, and skipped make-up essays and tutorials, which involved me driving and finding a space to park for 45 minutes. She entirely skipped the final.

At 10:00 p.m. one night, her father, a big lawyer in NYC, called me at my home. I had an unlisted number: the school had to have given it to her. He said he was suing me for being anti-semitic because I refused to give his daughter time off for Jewish holidays. I didn't even know she was Jewish; she had never asked me for a day off for religious observance, or anything else, and her absences didn't align with these holidays. I was terrified. He listed my paltry assets and said he was going to own my house and car, so he had researched me.

My department chair assured me that I would not face consequences and that this creep had also threatened to sue them. I was required to never give a student an F again and to change her grade to a D. From that point on, the lowest grade I ever gave anyone was a B minus. It was a horrifying experience, but I've heard worse.; 8/1/25, 9:33 AM
Narr said...: The grades I gave in my history classes ran the gamut and generally fit a bellcurve pattern. I had only one student ever go over my head (although many groaned and whined at me), and she lost. Said my grade would keep her out of law school in the fall, to which I replied that it wasn't my grade, it was hers. (I took many D's and F's myself, and knew who to blame.)

The department evolved an unspoken rough, two-tier system
for the grad students, which amounted to befriending and encouraging the better students with good letters of recommendation for further schooling or jobs, and tolerating the housewives and hobbyists.; 8/1/25, 9:40 AM
Narr said...: Of course all my experiences were long before the AI.

OTOH, the campus and library PTB decided about 2004 or so that library faculty would devote a few hours a week to "bibliographic instruction" for the grade school student/guinea pigs of the Ed School. We got to show 3rd and 4th graders the wonders of the intertoobs, and how to evaluate information sources online.

That ceased as suddenly as it started--I suspect the honchoes were just copying what the guys down the road were doing.; 8/1/25, 9:53 AM
Lazarus said...: Similar to the problem people have deciding if others are being ironic or not. In universities and other bureaucracies, there already is the problem distinguishing expressions of real feeling from merely formal letters and remarks.; 8/1/25, 11:39 AM
PM said...: AI should always appear with a watermark.; 8/1/25, 11:56 AM
Jaq said...: "Was that over the top? I can never tell." - The Best Joker.; 8/1/25, 2:55 PM
Tina Trent said...: Here's a weird thing. I've been playing with AI, admittedly to my skill set.

I expected it to be better at rhyming and form poetry, ie. sonnet forms. I'm not finding that to be true. It produces both good and bad, free-verse and form poems, randomly. As someone who does not understand the technology behind it, why would that be?

And sometimes it's really snarky. I'm not, at least about this. It's like the computer is laughing at me. I find it concerning over a trivial thing. A non-trivial thing...; 8/1/25, 3:02 PM
Jaq said...: We are just asked to accept so much absurdity, that the ability to cogently analyze evidence and come to some rational understanding of it and then to express your understanding concisely and precisely does not have the value for US citizens that it once may have.

For instance, there was a research paper written that had to do with using medical scanning of the structure of the human skull and the claim was that they could predict IQ very closely based on skull shape, yes, phrenology, and that they could determine your ancestry from the morphology of your skull too, but not your race. I am sorry, but rigorous academic training is not going to teach you how to accept that conclusion, and accept it you must. Better preparation would be for you to write all of your college papers using AI.; 8/1/25, 3:05 PM