KAYLIE SIROVY: Hello, everyone! I’m Kaylie Sirovy, your host from the Minnesota Daily, and you’re tuning into In The Know, the podcast that brings you all the latest on the University of Minnesota.

This episode we will be talking about artificial intelligence or AI for short. AI is a vast field that brings together computer science and data to create systems that can perform tasks and solve problems that typically require human intelligence. One of those AI’s readily available to the public is ChatGPT, which is a language model. This specific AI is designed to understand and generate human-like text responses. Here to discuss ChatGPT and AI with me is Maria Gini, a professor in the department of computer science and engineering. 

MARIA GINI: So ChatGPT has been designed as a system you can talk with. It would be like a person in a sense you talk to and you know, you can ask questions, any kind of questions, and get some answers.

It’s also designed really for conversation, not just, not just like a query, you know, like, like when you search something on the web, you say, you know, you write the sentence, whatever you’re searching for, and then you get the answers. Then you have to go through and click all the links and figure out which is the right things that you want. With ChatGPT, you can talk. It’s more like a person.

SIROVY: Of course, AI is a rapidly evolving field, and ChatGPT is just one example. However, with the potential to revolutionize teaching and learning, its influence on education becomes even more significant. 

GINI: Now, for education, first the main things that, you know, immediately as soon as ChatGPT came out, immediately started seeing messages saying, oh, we have to detect if the student has used ChatGPT to write a paper. So we punish them, right because they shouldn’t. We say you cannot use, punish them. 

And to me, when I started seeing this approach I say, many, many years ago when the calculator started getting into, you know, academic, people used to do, you know, all the computations by hand. Then you have this pocket calculator and there was, for a period of time in academia, they were forbidden. Now, you couldn’t have a calculator in your exams because it’s cheating, right? It didn’t last long because, of course, it’s kind of stupid things and I think, again, in some sense the same idea here.

SIROVY: The conversation is far from straightforward, as each educator holds their own unique viewpoint on leveraging AI to enhance student learning. There’s no one-size-fits-all answer to how teachers can utilize AI effectively in the classroom.

Moreover, the question of what aspects of AI to embrace and what to restrict poses a significant challenge. AI’s influence can be both a boon and a bane, depending on how it is wielded. Stevie Chancellor, an assistant professor also in the computer science and engineering department, aptly points out that the matter is complex and calls for thoughtful consideration.

STEVIE CHANCELLOR: I think a lot of professors need to really sit down and think about how their courses and assessments are structured so that we can truly assess that students are learning what they should be learning. But also acknowledge the role that these tools can play in helping us, because I think blanket bans on them disregard their ability to help people who may think or speak or organize their ideas differently for lots of different reasons.

SIROVY: In her own classroom, Chancellor finds that AI, particularly ChatGPT, can serve as an aid for students. It can assist them in generating original ideas, refining essays, and clarifying complex ideas. While Chancellor acknowledges the potential of AI tools, she also highlights the need for thoughtful and balanced implementation since using ChatGPT to write entire essays is not helpful to a student’s learning experience. She suggests that AI should complement, rather than replace, traditional learning methods. 

CHANCELLOR: Do I mind when my graduate students use ChatGPT to help edit their essays and clarify, you know, shorten their content by 10%? Not really, to be honest. Editing and refining your own ideas is, it’s a skill you need to develop and you kind of learn that by seeing other people do it. Do I mind that in a freshman writing seminar where the goal is for you to develop your own capacity to write and edit yourself? I mind that more. 

By replacing ChatGPT, or replacing your learning with ChatGPT and these other systems, I think it kind of shoots yourself in the foot in some ways by preventing your own intellectual growth. And so having a blanket policy in this case may not actually accomplish the educational goals of different disciplines or, you know, levels of the student.

SIROVY: How should institutions, such as schools and universities, approach the regulation of AI?

GINI: How to control is a complicated part. The U. S. has a lot of discussion, lots of documents from the government, and so trying to figure out how to regulate it. Regulations are complex because if you over regulate, you prevent the development of the technology. But if you under regulate, then you have bad, bad uses. So again, there’s an issue of figuring out what’s the proper balance.

SIROVY: The European Union has taken initial steps towards proposing regulations for AI technology, though these measures are currently under review and not yet approved. However in 2021, UNESCO had successfully approved a document that discusses the ethical concerns associated with AI and emphasizes the need for their proper handling. While it may not offer a definitive solution, it raises awareness about the potential disruptive nature of AI compared to past technologies and its potential to cause harm to individuals.

CHANCELLOR: So one of the things I love to tell people is don’t buy into the AI hype. Hype prevents us from critically reflecting on our own opinions about new technology, whether that be AI, 3d printing, CRISPR, or other innovations or policy developments that make it really hard for us to assess how we actually feel and what we want this technology to do. 

And with that comes, unfortunately, some social consequences we may have to face such as displacing jobs, taking away people’s agency and making decisions. If we overbuy into AI hype, we can risk not thinking about the world we want to live in and what AI provides us in that world. And so my opinion is very driven by critically thinking about what we want our world to look like with AI.

SIROVY: The growing awareness surrounding the complexities of AI is undoubtedly a positive development because those sci-fi movies about robots taking over the world may be fiction, but those fears about AI are still there.

Acknowledging its capacity to be transformative yet potentially damaging underscores the importance of cautious and well-informed decision-making in shaping AI policies. 

GINI: I think, you know, AI is kind of complex, the more we talk about, right, the more people understand, the better we are, because there are a lot of people very afraid of AI, right? I mean, they think about the, you know, they call the artificial general intelligence, we’re going to build machines that’s smarter than us. I mean, all those fears are coming out. I don’t think they’re justified, but at the same time, it’s a serious question. If we invent something that is smarter than humans, can it take over the world? It’s possible, right? I think it’s very far away, but that’s something good to think about. And the more we think about and discuss, you know, knowledge and information is always a solution for a lot of problems.

SIROVY: Numerous AI systems exist and operate in the background of our lives, impacting our everyday experiences. A prime illustration of this is credit card fraud detection, a familiar encounter for many of us. 

CHANCELLOR: When your bank, you know, calls when you’re accidentally, like, trying to make a charge in a foreign country, or in a place that you’ve never been before, and your bank says, ooh, we don’t know if that’s you. That’s actually a form of artificial intelligence that’s been around for a really long time. Now that one seems less mundane, but there also are very mundane things that you interact with every day. Like how TikTok organizes and sorts and ranks the videos it shows you and infers what you might like. Content moderation systems that automatically detect when you’ve used, you know, a certain kind of language.

SIROVY: Although the focus of this episode mostly revolved around education, it is vital to recognize that AI’s influence reaches far beyond. Its integration into various industries, including healthcare, automotive, security, agriculture, and even sports, is happening at a rapid pace. Gini explains that half of the workforce may need to be retrained to account for the uptick of programs like ChatGPT.

GINI: Because ChatGPT now is going to affect the white collar jobs, not only the blue collar jobs, right? And so there are lots of jobs that are going to be affected, lots of jobs that will get lost and people will have to get retrained. So that’s inevitable. I’m not saying it’s a good thing, but there’s not much that can be done.

So AI is big and it’s in everything. So I think the difference, you know, some years ago, yeah, it was really for the, for the experts, you know, very specialized topics. Now, I think everybody needs to know something about AI because it is going to affect everything.

SIROVY: This episode was written and produced by Kaylie Sirovy. As always, we really appreciate you listening in. Feel free to email us at [email protected] with comments or questions. I’m Kaylie, and this is In The Know.