How to stop worrying and love (or at least live with) ChatGPT
MARY LOUISE KELLY, HOST:
Ever since the chatbot ChatGPT launched back in November, educators have raised concerns it could facilitate cheating. Some school districts have banned access to the tool, and not without reason. The artificial intelligence tool from the company OpenAI can compose poetry. It can write computer code. It can maybe even get an MBA. One business school professor recently fed the chatbot the final exam questions for a core MBA course and found that, despite some surprising math errors, he would've given the chat bot a B or a B-minus in the class. And yet, despite, again, the clear potential for cheating, not all educators are shying away from ChatGPT.
Ethan Mollick of the Wharton School at the University of Pennsylvania has made it a requirement to use the chatbot in his classes, and he's here to explain why. Hi, there.
ETHAN MOLLICK: Hi.
KELLY: All right, so this is spelled out right in your new syllabus. Students are not only allowed to use AI, including ChatGPT - they are required to. I know we're probably early in the new term, but how's that going so far?
MOLLICK: Going great. I mean, the truth is, I probably couldn't have stopped them even if I didn't require it.
KELLY: Give me an example of when you say it's going great.
MOLLICK: For example, we had a class yesterday where people had to generate ideas for their class project. Almost everyone had Chat up and running, and they were bouncing ideas off of it, asking it to generate a list of 50 good project ideas and interrogating it on those things. And the ideas so far are great, partially as a result of that set of interactions.
KELLY: Hmm. It's interesting because, as I prepared to talk to you, I was thinking you would make the case, look, it's here. We can't get rid of it. It's not going away, so we have to just resign ourselves to ways to learning to live with this. You don't sound resigned. You sound downright enthusiastic about the possibilities.
MOLLICK: I alternate between enthusiasm and anxiety.
KELLY: (Laughter) Yeah.
MOLLICK: I mean, I think that this was a sudden change, right? We don't - there is a lot of good stuff that we are going to have to do differently, but I think we could solve the problems of - how do we teach people to write in a world with ChatGPT? We've taught people how to do math in a world with calculators. I think we can survive that. And I think, as educators, part of our goal is what changed about the world that we have to teach. So I think this is a tool that's useful. I think there's a lot of positives about it. That doesn't minimize the fact that cheating and negativity is there, but those have been there for a long time.
KELLY: One positive that I had not thought about, but that you note, is that it might help level the playing field in the classroom for students whose first language is not English - explain.
MOLLICK: Yeah. So writing is a skill, but not everyone is a great writer. It might be a talent issue. It might be that English is not their first language. Some people take a lot longer to write than others. And there was never really a tool to help you with that, other than just plain cheating. Now, everybody can write well with ChatGPT, and, in fact, I now have a requirement in my class that I expect every piece of writing to be good pieces of writing because there's no reason you can't do it that way. So ChatGPT really does level the ability of people to be good writers. And I've had my students tell me after class that they're now using this to write emails and letters and are being taken more seriously as a result.
KELLY: I'm not an educator, but I am a parent, and I have been a student. And the parent and student in me are screaming in protest, saying, does it count as good writing if you didn't write it? Where do you come down on that?
MOLLICK: So I think we have to separate out different kinds of good writing. I mean, at its current phase, AI will never be as good as the best experts in a field, right? And we still need to teach people to be experts. They still need to do a lot of writing to learn to be a good writer. But we also do a lot of writing in our day that is not that important, but could be very stressful and time-consuming for people who aren't good writers - whether that's an email to a professor or making sure that an essay is well formatted. I mean, these are things that are hard to do, very time consuming, have very limited value other than what you're trying to accomplish, and I think we could separate those kind of writing from learning how to write well.
KELLY: Have you had a problem yet with a student cheating using the software, and would you know if you did?
MOLLICK: So that's an interesting question. There is a lot of people claiming that they can detect cheating, and it just isn't clear that that's the case. And so I think everybody is cheating. There was a survey done of students at Stanford, and, you know, a large number of them had reported that they had cheated in the last month since getting this material in place. I mean, it's happening. So what I'm asking students to do is just be honest with me - tell me what they use ChatGPT for, tell me what they used as prompts to get it to do what they want, and that's all I'm asking for them. We're in a world where this is happening, but now it's just going to be at an even grander scale.
KELLY: Is that a depressing statement from somebody who has devoted their life to higher ed - who's a professor at Wharton?
MOLLICK: Yes, but there's an estimate from the U.K. that, right now, as of today, 20,000 people in Kenya are employed full-time just writing essays for people. I mean, there is a lot of evidence that cheating has been pretty ubiquitous for a while. I think it's bad. I think there's a reason we assign homework and assign essays, but I don't think that that is novel. And it is depressing, right? But it's also - you know, means that there's more people we have to reach. We have to teach them why it's important to do these things. We have to change assignments around to make this work better. We have to think of better ways to keep people honest. All of those things are true, but I don't think human nature changes as a result of ChatGPT. I think capability did.
KELLY: Ethan Mollick, associate professor of management at the Wharton School of the University of Pennsylvania. Professor Mollick, thank you.
MOLLICK: Thank you very much.
(SOUNDBITE OF WIZ KHALIFA, BIG K.R.I.T. AND GIRL TALK SONG, "HOW THE STORY GOES") Transcript provided by NPR, Copyright NPR.