-
-
-
-
Photo by Growtika on Unsplash
I’m skeptical about artificial intelligence, in part because we need to find a place in our world where it works for people but it does not replace us nor our capacity for thought.
The truth is that we’ve evolved to be lazy, as a whole. We created agriculture because it was ultimately easier than wandering around the countryside looking for game. We created machines to do various tasks because we didn’t want to have to work as hard as we did. We are, on the whole, a lazy species that has to force itself to take on work.
Tilting At Windmills is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
Which is fine, because the desire to minimize labor gave us time for great works of art and literature. It gave us empires and heroes. It took us from just another animal trying to survive and into the undisputed dominant species on this planet.
So AI is likely a tool that can help us if we learn how to use it right. What “right” means is still a matter for debate in many cases, but we know what isn’t.
This story has been going around on social media for a bit, and it’s worth thinking about.
For the first time since he started teaching Welfare Economics and Social Choice Theory nearly two decades ago, Brown University economics professor Roberto Serrano gave his students a take-home midterm this spring. Quite a few students had expressed anxiety about being in a classroom after a gunman killed two students and injured nine in a December mass shooting at Brown , and so “it was appropriate,” he said, to allow students to take their exams at home.
But by the end of the semester, Serrano regretted the decision. Dozens of students in the class likely used artificial intelligence to cheat and earn perfect or near-perfect scores on their midterm, he said. Serrano in turn made the final exam in-person, which led more than a dozen students to drop the course and even more to fail it. Administrators’ response to the widespread cheating event has been “meek,” he said, and the incident has raised questions about how universities can—and should—respond to AI-enabled cheating at scale.
His welfare economics class typically attracted up to 30 students, but this spring he taught 86—an increase he attributes to the promised take-home exams. When the midterm came along, the average score was 96 percent.
“Historically the average grade in the midterm of this course has ranged between 65 and 80 [percent], and this exam was harder than the exams I wrote in the past, because … take-home is an opportunity to challenge the class a little bit more, given that you’re giving the students unlimited time,” Serrano said.
Serrano then ran the test through ChatGPT, and it gave the exact same answers. More than that, it did so in some rather bizarre ways…which was exactly how many of his students answered the questions.
In class, Serrano challenged his students, told them about his suspicions, and said that he wasn’t going to invalidate the test, but would instead give them a chance to defend themselves by taking it again.
Now, I remember test-taking, and I’d have griped about having to do it all over again, like anyone else. Instead, 18 dropped the class immediately, and others never came to take the final.
As for the grades, the average went from 96 down to 48.6, with three of the students getting a zero. Nearly half of the previous scores.
I’ll put part of that on Serrano for letting them have a take-home midterm exam, but the onus should still be on students to do the work themselves. This is flat-out cheating, and it’s normal these days.
More and more students aren’t turning to AI as a tool to help them fix their grammar and spelling, to translate text into a language they speak, or to collate data from numerous sources into something cohesive. They’re using it to write their papers.
Then, of course, the teachers are using AI to grade those papers , which means AI is grading AI, and the entire educational system is becoming more and more devoid of humans except as a go-between.
What’s happening now is that AI is the new hotness, and people are seeing a wealth of possibilities within it. They’re not wrong to see that, but the issue is that they’re operating as if AI can’t be wrong. They’re treating programs like ChatGPT or Grok as infallible, when they’re not. Google AI, for example, has given me names of supposedly influential people in the Second Amendment writing community who don’t exist, whose names don’t even exist, and Grok has made claims about historic texts that I definitively knew were inaccurate. We all know that AI makes “mistakes,” which should probably be called “hallucinations,” and yet what do we have from the younger generation?
They’re worshipping the idol of AI, treating it like a flawless god they can turn to in order to answer their prayers. In the case of Serrano’s students, it was for the answers on the test, but it won’t end there. It can’t end there.
I don’t hate AI. I think it’s an amazing bit of technology, even if it’s not what science fiction told me AI would look like. Some of the more bizarre stories I’ve heard about AI are closer to what I tend to think of as “artificial intelligence,” of course, though that doesn’t make me feel any better. Still, I don’t hate it. I use it regularly to spitball ideas for various things in my ever-growing series of thought exercises.
But too many people use it not just to escape having to do their own midterms, but as the answer to all things. They see it as the next step in technological evolution, which will let us actually do less and less in our daily lives and still function.
They worship the promise it brings, but they do not seem interested in tempering their demands or expectations. It’s just a matter of time before we have “Jarvis” from the MCU in every middle-class household, and that’s terrifying to me. The less we need to do, the less we will do, and the less we will actually use our own brains for anything.
Share
AI, in time, will remove not just our need to think, but our desire to think, and humanity will celebrate it.
For the first time, we won’t be the masters of the new technology. It won’t need us to function. It can just do what needs to be done, all without any significant human input, and we will, in time, forget the dangers we’ve been warned about for generations.
I’m not saying not to use it. I’m not saying it can’t help in a plethora of fields. It most definitely can, and you most definitely should use it if you have a need.
I’m saying you shouldn’t trust it with anything that matters. Otherwise, you run the risk of worshipping the false idol of technological “thinking.”
---
Tilting at Windmills is 100% reader-supported. If you enjoyed this article, please consider upgrading to a paid subscription for 15% off the first year or making a one-time donation here . Either would be greatly appreciated at a time when I know good and well I’m about to have some serious expenses coming.