Speaker 1: We've been talking a lot about AI and education this morning, and now we're bringing a little chat GPT into your homes. And we're checking out how the groundbreaking app works in real life situations. Say, for instance, you're working on a book report for school.

Speaker 2: Well, with the help of our social media producer, Cynthia, we're asking chat GPT in real time this morning, write an essay about character development in Huckleberry Finn. So putting in that question right there. And this is what happens on the screen in real time. An American classic, Huck Finn. And look, this AI generating technology is writing a paper that if a student was not honest, could be submitted to their professor, to their teacher. And this would be the assignment and how fast it would come together. So you can see why in the poll that we put out on social media, 52.9% of young people that we asked, do you use this? Are you aware of it? They said, yes. And now you're saying, why?

Speaker 3: Crazy. It looks like a human wrote that document.

Speaker 2: We just did it on real time.

Speaker 3: Until you put it through a scanner

Speaker 1: and you realize a human did not write that document and your child is now in trouble.

Speaker 3: Right, and that's probably why educators are very worried about chat GPT being a part of an AI in education.

Speaker 2: Yeah, the program writes that document that looks just like a human wrote it. And that has a lot of educators of all ages worried this morning.

Speaker 1: So the question, is this a time saver? At first draft that you work on when you turn it in, or is that cheating?

Speaker 2: So we are asking an expert here now to address some of those concerns about AI is Professor Jane Rosenzweig, Director of the Writing Center at Harvard College. I have to disclose to our viewers right up the top, you and I know each other because you helped me get my master's degree. And I never would have dreamed of handing in something to you or to Christina Thompson or any of my Harvard professors that had been written on chat GPT, but you're in the LA Times and talking about that there's a very real risk, Professor.

Speaker 4: Yes, hi, Reid. I'm glad to know you wouldn't have turned in a chat GPT generated document. So I think where the risk here, I'm less interested in cheating, although that is certainly something that we think about and more interested in what it means for what you're not learning if you're using chat GPT, right? All of these companies that are putting out these chat GPT tools are telling us this is a real time saver, right? You can use this and save all this time so that you can do the real work. But what if writing is the real work? And when you're in my class or any writing class, the real work is actually doing that writing yourself. And the reason why that's the real work is because we think about writing as a way of thinking, of figuring out what you think. So the idea that the chat bot is gonna generate your first draft for you means that you're not gonna be doing that thinking, that figuring out of what you think. We don't need more papers in the world, right? I don't assign papers to my students because I'm in need of papers. I assign papers to my students because I want them to have the experience of working through a problem, asking a question, figuring out the answer and working through that in their writing.

Speaker 1: That's a really good point. I mean, it's the thinking, it's also the retention. How do you expect this to affect how students compose a paper moving forward though? Do you think AI is gonna be accepted as a trusted resource?

Speaker 4: Well, I think right now we all know, those of us on the educator side, that you can't trust it in the sense that it's doing what they call hallucinating, right? Some of the output is just wrong. You'll get a list of sources and you'll go and look them up and they won't exist. There's also bias built into these systems and that's a problem that's gonna be difficult to solve. So right now the concern I think is that students will see it as a trusted resource and that educators really need to be focusing on how we're gonna help our students understand what this tool can do and what it can't do. And I also think that it's very related to the situation you're in, right? If you are trying to write an email and you write 30 of these emails a day and they're really not a writing situation in which you're gonna be doing a lot of thinking and working through ideas, then maybe this will become the tool of choice for people. But in the classroom, we really have to be thinking about what we're gonna lose if students turn to this tool and helping them understand what choices they're making in their learning process.

Speaker 3: And how are teachers going to figure this out, tell the difference? How are schools gonna be able to spot chat GPT in student work?

Speaker 4: So I know one of you mentioned that you can put it through a detector at the beginning of our time together. The detectors are really not trustworthy. That's not gonna be the way forward. We're hearing about all sorts of situations where students are being accused of using a detector to generate whether their work is their own and it gets it wrong. So I think as educators, what we're really gonna be having to do is rethink our assignments. And we've already started to do this. How do we help students move through a process of writing and understand why the process of writing matters rather than focusing on something like trying to catch people for cheating?

Speaker 2: You are not there to catch us for cheating, but what you are there is to do what is the genius, I think, of your program there at Harvard. And it's to come from a place of yes and get that critical thinking in the students' minds. And that's what artificial intelligence will never do. It will never teach us to think critically. Professor Jane Rosenzweig, thank you so much. I'm so glad I no longer have to hand in papers to all of you, but I am glad that I have lifelong advocates and friends. Best to you in Cambridge and say hi to everybody there for me.

Speaker 4: Thanks so much. Great to see you.

Speaker 1: I really love that she brought that up when it comes to critical thinking. And it also, creativity, it really limits your creativity as a writer and as a person if you're just cutting and pasting.

Speaker 2: Yeah, ideas are our currency as writers, right? And if we're getting those ideas curated from some computer program, you can't do that.

Speaker 3: No, and it's also, I think, a little bit of work ethic. You know, there's something to be said in the time you spend working on a project and writing it and researching it. You learn a lot, you put your effort in, and then you're proud of that. So I think that has a lot to do with it too. I hope kids think about that before they make that choice.