Dan Loney: In a world where artificial knowledge is the only thing that is being used, we are currently considering whether AI might affect economic literacy. It’s a delight to get joined in-studio by Michael Roberts, finance doctor here at the Wharton School. Before we get into the Artificial aspect of it, Michael, I’d like to know what your feelings are on the state of financial literacy in general.
Michael Roberts: I can start by saying that I’m not completely sure what financial education also means at this time, but if I consider the study on the subject and the way I interact with experts, I would define it in different ways.
Second, in general, it’s very bad. You consider the rate of economic education. They are about 57% in the U.S., but that hides a lot of variation in the community across women, across money, across training, across time, where we see rates varying a great deal. I believe the development of financial education has evolved over time, but at a slower rate. It’s no improving quickly much. Current government policy changes, in my opinion, are good, but perhaps more sobering than good. Who will instruct it? How is it going to be taught? What’s going to be taught? There are a lot of problems.
Loney: Where does AI fit in around financial education?
Roberts: It’s difficult not to recognize that it is fundamental to financial education and economic process in general in some way or another. And given how quickly it’s changing, apparently what I’m about to say may become outdated when this is finally published. However, I’ll take a kill.
In terms of a stand-alone, let’s say, investment adviser or economic guru, where it is right now is just nothing near where it needs to be. It requires a great deal of financial knowledge and the ability to question or indulge with the AI to produce insightful responses. When it does create answers, they can be spectacular, but they can even lead you astray. How long until there is a place where anyone can communicate with Artificial to remedy their financial issues? I have no idea, but it’s not going to nullify the importance of what I call economic skills.
You showed some thought to how conceptual AI might play a role, Loney. What have you discovered?
Roberts: I have been using it for about two centuries, ever since I had a group of students in one of my classes use it to help me with an interview. It was in a various environment, more software and data technology. However, it was a fast wake-up call.
I’ve been feeding it concerns, both common financial literacy issues, as well as questions from my test, and it’s exciting. In practice, AI is financially knowledgeable by any estimate. From that perspective, it’s outstanding. But when you ask it concerns such as only large, open-ended issues like, “How may I save for retirement?” “What if I invest in? It really isn’t sure where to begin, and I’m sure there are rules or restrictions on AI from becoming an unlicensed RIA (registered investment adviser). It provides a very extensive review. When will it be able to become our economic specialists, in my opinion? I don’t know how quickly AI technology will advance toward that goal.
Loney: What has to occur for it to get to that level? One thing we do know is that as AI grows, we must keep in mind that there are people working in the background.
Roberts: What’s important is that I don’t view AI as only a cure. It serves as a supplement and a starting point for our journey toward financial literacy, whichever it may be or economic skills. In other words, I believe it will be crucial for people to realize that financial principles and awareness are still relevant. They are going to be crucial in order to work with AI effectively, wherever it may be in the prospect.
Some people might know whether or not AI can be a useful tool to teach along the way given the fact that we know that financial literacy knowledge is essential and that it has been difficult to implement in schools.
Roberts: I think it is, and it probably will be a great help faster than we think. Sal Khan has created Khanmigo. There is no reason why we can’t have our classes using personal AI teaching assistants right away or soon. However, being a TA for a school differs from being able to respond to people’s particular personal or professional financial queries. It’s important to recognize that difference. Everyone is different. You and I both have unique retirement savings strategies.
Communicating With and Interpreting AI
Right, AI can’t fully comprehend how each person feels about that, and that’s true.
Roberts: Not without the aid of the end consumer. When I say that it is crucial for people to understand how to speak with AI, that is what I mean. But to do so, you have to know funding. You have to know what to ask it and, importantly, how to interpret what it’s spitting back at you.
Loney, I noticed the research you’ve done on AI and related topics. You might also change the way you phrase some of the questions to better understand the conceptual AI tool.
Roberts: Yes, positively. To give the AI the chance to obtain the proper response, I had to learn how to communicate with it. In the same way, I would assume that the AI is learning to communicate with me to provide the solution I need.
Programs and Limitations of Generative AI in Finance
Loney: What you’re saying is somewhat of a concept that we’ve heard from other experts — that AI, in the context of where financial education may be going, will be a way to help it along the means, but it won’t become the be-all, end-all.
Roberts: I think it has the opportunity to accelerate our progress, but it won’t be a replacement for financial literacy, financial proficiency.
You also mention the ability to interpret the responses provided by ChatGPT in your work, Lindsay.
Roberts: Yes. It’s not just spitting out a number. It’s giving a somewhat long-winded explanation of the answer. If you don’t understand what it’s returning, the information is not helpful. It’s not usable.
It’s progressing so quickly, it’s difficult to keep up with. That’s why I say it’s hard to see a limitation, at least from my perspective, on what it’s going to be able to do, other than I don’t see it currently as a replacement for the progress we need to make on the literacy side.
What has the study’s impact on your outlook on teaching finance and the importance of using ChatGPT as a tool in that process, Loney?
Roberts: Oh, I’m incredibly excited. I think ChatGPT or other AI models from the previous generation are just a fantastic addition to the toolkit we need to address financial challenges. I embrace it. I counsel my students to embrace it. I want them to use it on tests or homework, if they choose. But they quickly discover that they cannot avoid understanding the finance if they are going to use it.
Loney: How much has it benefited your students?
Roberts: It depends. For data science-related financial applications, when I teach my Data Science for Finance course, it’s hugely helpful because it almost completely eliminates the technical burden, in terms of programming and working with data. It has made things so much more efficient. On the banking and personal finance side, it’s a little less helpful because it struggles with broad, almost open-ended questions. At this point, the application’s requirements change.
Do you have any surprises about your interactions with ChatGPT, Loney?
Roberts: Yes, sometimes it gives impressively insightful responses. In fact, it found an error on one of the financial literacy exams, which I thought was really impressive. On the other hand, it makes some of the same silly errors that it makes, which also impress me. As it grows in complexity and with more interaction and data, that level of volatility needs to be determined.