John Warner
@biblioracle
Writer, speaker, consultant. Chicago Tribune columnist, blogging at Inside Higher Ed. Coming soon, More Than Words: How to Think About Writing in the Age of AI. Previously: Why They Can't Write and The Writer's Practice. biblioracle.substack.com
I interviewed my old friend Dave Eggers about his new novel "Contrapposto," along with some other stuff, including what gets kids invested in themselves and their writing. biblioracle.substack.com/p/dave-egger...
If Harvard faculty need to artificially restrict A grades as an incentive to improve their class offerings they are in deep trouble. Maybe this whole op-ed is an admission that they're just not up to the task of educating.
Seriously? Do people really think Harvard's reputation for "excellence" has been harmed by the rise in A grades? Have job placements dropped? Applications drying up? This is elite pseudo-meritocratic thinking at its finest.
More logical nonsense. Every other example of the collective action problem has a finite pool of resources. This is not true of grades, unless you create an artificial limit, like they've just done.
I know these guys are economists, not rhetoricians, but classic begging the question here, assuming that the Harvard grade distribution is because of "easy A's." What an indictment of their institution!
Interesting framing from two members of Harvard faculty conflating their intramural angst about the challenge of ranking and sorting the spectacularly excellent with the entirety of education. Hubris much?
A guy who helped enroll a student population at the University of Alabama that was closer to the demographic make-up of the state is too woke for the Manhattan Institute. An explicit white's only stance.
What's maybe most galling is that at some level, this guy must have thought he'd get away with this shit. The spin that his fuck-up is an object example is the turd on top of the shit sundae.
This is the opposite of what I want in an LLM. What is the model that is against personalization? That's the one I'll use when I need its assistance.
Definitely email that guy! It's unfortunate that more people don't do this kind of thing because the literature on student engagement goes back 20+ years & we now a lot about what gets students invested. I use this slide in almost all my workshops. Note the date of the reference.
Some of the result of this failure to start at the roots is advice that is little better than pablum. What is "intentional" use of an AI tool? Why would using AI to practice verbal class components be a good use relative to alternatives? What does it mean to "recognize the limitations."
It accepts the framing of AI as a potential "co-intelligence" without grappling with the root differences of human vs. generative AI outputs in terms of origin and experiences. From the jump it makes assumptions about AI that are not necessarily supported when it comes to writing.
When I intentionally prompted with partially incorrect info (I knew that Jane Lynch had not been on ER) it confirmed my wrong answer with more wrong stuff. ER heads know that Mariska Hargitay played a character named "Cynthia Hooper" (Mark Greene's post-divorce fling).
The experiment is to illustrate the unique importance of experiences in learning as humans and it's clear that the experience of watching ER is superior to the way ChatGPT assembled its list from probabilities. Even when prompted to check it would still make errors.
An experiment with ChatGPT as an illustration for an upcoming presentation asking it to identify actors/actresses who appeared on ER and then want on to bigger things. Its error rate was significant. Because I'm a weirdo who has watched it all the way through 3 times, I caught them all from memory.
I got a chapter on this in More Than Words. The best way forward, IMO, is the collective model like Defector, Flaming Hydra, et al.
There's no way the actual video of Ben Shapiro at UATX can live up to the capsule description of the different parts of the presentation. If there's anyone I'm listening to about uniting the country it's Ben Shapiro and Joe Lonsdale.
Detaching the work of the class from assessment in the way Paul does it may seem radical, but I can testify from my own experience, it is necessary and from the perspective of most students, quite welcome. This is a college student expressing what should be ingrained by grade school as something new
Success does require a shift in how experiences/assignments and assessments are used, a shift perfectly expressed by @plthomasedd.bsky.social in his post here:
The irony of a tenured full professor at an Ivy League university chastising his colleagues for refusing to "look in the mirror" and dismissing critics while ignoring the overwhelming factors that have led to the problems he outlines (Which I agree with!) is off the charts.
This is a call for measuring teaching effectiveness that ignores the most meaningful criteria we could use: What percentage of your faculty makes a professional wage and has a teaching load consistent with disciplinary norms? I taught college for 20 years and never achieved either.
For a second I thought that Inside Higher Ed had written a feature article on my book, but alas, 'tisint true.
Oh, Louis, it's the rubric that's the problem. You're coaching your students to do school, which is not the same thing as learning. You already have identified the problem. Ditch the rubric, ditch the AI feedback.
The fjords are truly magical, and you can also go see where Edvard Grieg did his thing.