Has AI reached the point where a software program can do better work than you? : NPR

Rob Schmitz, Host:

Has artificial intelligence reached a point where software programs can do a better job than you? Today we’re talking about ChatGPT, a powerful tool that can help you with a wide range of tasks, from generating human-like text to providing helpful answers to your questions. To do. We’ll explore what ChatGPT does and how it’s being used in different industries. In fact, the last two sentences he just read were written by his ChatGPT bot, not by me. Ethan Morrick, a professor at his School of Business at the University of Pennsylvania Wharton, is also experimenting with his ChatGPT. He’s joining me now to teach me how ChatGPT works. welcome.

Ethan Morrick: Thank you. I am happy to be here.

Schmitz: So Ethan, I came across ChatGPT when I was hanging out with some friends this weekend and they were talking about this very interview. Then suddenly the heads of all the teenagers in the room popped up and they said, oh, we have it in our phones. This is clearly a trend among teenagers.

Morik: I think it’s universal. Soon something of a fun phenomenon happened on TikTok. But also, whether the teacher knows it or not, he finds one school where not too many people use this in their lessons.

Schmitz: And how do they usually use it?

Morrick: I mean, there are a few things, right? So the most obvious thing people get from just playing ChatGPT for a few minutes is, wow, you can cheat an essay with this. And if you spend a few more minutes, they say, you can cheat writing software code or translating languages. But the real-world uses go far beyond that, and I was surprised to see how some students report how they use the feature.

Schmitz: Well, please explain a little. What other uses could it offer other than cheating?

Morrick: Well, you mean there are a lot of them. But just on the educational side, you can actually use it and paste it into your essay to ask for revisions. You can paste the entire academic paper and have it summarized. You can find errors in your code, fix them, and tell them why they went wrong. It really is a generic companion that can be used for all sorts of purposes.

Schmitz: So I have ChatGPT open in front of me on my laptop right now. Can you explain this? How do I use this?

Morrick: So the best way to think about this is if you’re having a conversation with an omniscient, all-knowing, pleasing, but sometimes lying intern.

Schmitz: (laughs).

Morrick: And if you take that approach, you’ll be in great shape.

Schmitz: Okay.

Morik: Let’s use it to explain something or explain something. What are your interests that would get you writing an essay? Let’s play a little.

Schmitz: Well, I’m on public radio, so is it a story or a radio story?

Morrick: Right. Select a topic.

Schmitz: Let’s talk about frogs.

Morrick: Great. Why not create his NPR story about frogs?

Schmitz: Can you make an NPR story about frogs?

Morik: Perfect.

Schmitz: Title – “Fascinating World of Frogs”. (Reading) Frogs are a common and fascinating part of our natural world, as anyone who has visited a pond or marsh on a warm summer night can attest. These amphibians, belonging to the order Anura, are known for their distinctive vocalizations and their ability to hop and swim with ease.

It goes on and on. A giant goliath frog. A small frog about the size of a fingernail.

Morrick: Yes, this is a little long. Why don’t you try it as a poem that actually rhymes? So can this be expressed as a rhymed poem?

Schmitz: Right. Could you write this as a rhyming poem? please. (Reading) On warm summer nights in ponds and swamps, you can hear the familiar croaking of frogs. They belong to the order Anura and have a unique life cycle for you and me.

To be honest, I expected a little more from ChatGPT. I’m — OK, I’m starting to understand both the curiosity and the limitations of this. How did you use ChatGPT?

Morrick: So I really tried to see how far I could automate my work in the field of education. And I’m a professor of entrepreneurship at the Wharton School of the University of Pennsylvania. I teach an MBA. Therefore, I requested the creation of an MBA syllabus. and…

Schmitz: Wait. Have you asked ChatGPT to create your MBA syllabus?

Morrick: Of course. In short, we have created a pretty good syllabus for an MBA class. And I said, “Well, that’s interesting.” Could you write the final assignment for that class? That gave me the final assignment. This was again a really great business plan challenge. Can you give me the scoring rubric? You broke down the scoring and gave me a table of points. I asked him if he could give me the beginning of his second lecture that will be given in this MBA syllabus. And it wrote me a lecture. What we don’t fully understand is this multiplier of truly amazing abilities.

Schmitz: When we started this conversation, it turned out that some of our first users were teenagers and they were using it to cheat on their essays. What are the concerns and ethical implications of this?

Morrick: I think we are just beginning to address this issue. Is it ethical to have the AI ​​write a draft and then fix it? Is it ethical to paste the text and ask for improvement? lie confidently. How do you treat the work? So I think we still don’t understand what this means, but I think it goes much deeper than people think. It will replace us all. So get ready for that.

Schmitz: Really? Will this really replace us all?

Morrick: I don’t think anyone really knows what the future holds. He has two options, right? Option 1 is that he can draft 10 stories and keep what he likes, doubling his work capacity. I mean, in the best version of the world, you’re out there and using this to do 10x or 12x the work. And I already hear people doing this all the time. They use it to write letters, bios, agendas, whatever.

Schmitz: Right. But, in other words, I’m telling you to create more jobs, but at the same time, I can foresee that if this kind of thing were a little more precise, it would actually take people’s jobs away. I may report the news, but I don’t know if it’s true.

Morrick: Of course. Me- Actually, I just had a live interview on a TV station and it turned out that all the questions were done by his ChatGPT, which I didn’t know. It sounded great. Here’s what I mean. In fact, I think there’s a kind of awareness that comes from playing with it. What should I do? So I gave you a turnaround case. The good news is that your abilities and intelligence are multiplied by 100. The downside is how many people are needed. And I have no answers to those questions.

Schmitz: Ethan Morrick is a professor at the Wharton School of Business at the University of Pennsylvania. Ethan, thank you.

Morrick: Great. Thank you for inviting me.

Copyright © 2022 NPR. All rights reserved. For more information, please see our website Terms of Use and Permissions page at www.npr.org.

NPR transcripts are produced by NPR contractors on a rush deadline. This text may not be in final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative recordings of NPR’s shows are audio recordings.