Dear Auburn Students:
Over the past two years I have read a lot of essays, papers and discussion posts and thought “I am 97% sure a machine wrote this.” I’ve read research papers citing books that do not exist. I’ve read in a short biography that I serve as the President of the Alabama League of Women Voters. (This is false). These all represent the downsides of generative AI.
I don't think you are a bad person or even a bad student if you use ChatGPT. If you gave 19-year-old me ChatGPT in 2003, I cannot tell you with a straight face that I would never have used it to complete an assignment. It would have given me more time for Napster, Friendster and AOL Instant Messenger (ok, boomer). I don’t think you are a bad student but I’m worried for you.
To be clear, this is a powerful technology with many useful applications, including in the university setting. There is evidence that it can aid in some aspects of learning. However, there is also strong evidence that it can significantly harm learning and even lead to a decline in human ability. And it is becoming increasingly possible—or even likely—that this technology is not as transformative or revolutionary as some of its boosters have argued.
You need to be literate in this technology, but I’d urge you to ask yourself some questions about it: When I use ChatGPT what am I gaining? What am I losing? When I graduate, how common will AI literacy be? On the other hand, how common will the ability to write, read, communicate, think critically, and make human connections be? What will set me apart in my career? What skills should I invest in beyond the ability to prompt a Large Language Model? One of my colleagues put it better than I ever could when they said, “When all you can do is tell AI to write something for you, all you are doing is demonstrating that you are replaceable by AI. You need to be better than AI.”
The good news is that it’s probably not that hard to be better than AI. Ask ChatGPT how many “b’s” are in the word “blueberry.” Ask Microsoft Co-pilot for a list of the states containing the letter “R.” You will be surprised at the answer. (Or maybe not — they might have fixed it by now). Better still, give ChatGPT a text that you know very well, and spend some time asking it detailed questions about the text. How often is it correct? When it’s wrong, does it admit it? Does it ever say, "I don’t know?"
You need to be better than AI. You can be better than AI. But to be better than AI you need to lock in and do the work. You need to write on your own and, through that, think on your own. Again, I’ll borrow another human’s words — those of author Ted Chiang: “[W]riting is…strength training for the brain. Using ChatGPT to write your essays is like bringing a forklift into the weight room; you are never going to improve your cognitive fitness that way.” Improve your cognitive fitness. Write on your own—or better still, ask another human being to help you improve. Try leaving the forklifts out of the weight room and the robots out of your writing.
Do you like this story? The Plainsman doesn't accept money from tuition or student fees, and we don't charge a subscription fee. But you can donate to support The Plainsman.
Peter B. White is an associate professor at Auburn University in the department of political science.