Conversation with Dr. David Zaretsky - AI in Education Part 1

How is AI transforming education, critical thinking, and the future of work? In this episode of RHEMINISCING with Dr. Tony Rhem, Dr. Tony speaks with Dr. David Zaretsky of Northwestern University about generative AI in universities, the risks of overreliance on tools like ChatGPT, the importance of critical thinking, AI as a tutor, workforce disruption, reskilling, entrepreneurship, and building a personal AI strategy for the future. This is a must-watch conversation for educators, students, business leaders, and anyone navigating the impact of AI on learning and careers.

#aieducation #generativeai #ai #chatgpt #criticalthinking  #futureofwork   #ailiteracy #aistrategy  #workforcetransformation  #highereducation   #entrepreneurship #aiethics #northwesternuniversity

Key themes: AI and the future of work, AI Strategy, Generative AI, AI governance

Transcript

Copy Transcript Copied

SPEAKER_00 0:04

Welcome to another episode of Reminiscing, where we explore AI, past, present, and the possible today with a very good friend of mine and uh amazing individual, uh Dr. David Zaresky. He's been gracious enough to join um the podcast today. Um, let me tell you a little bit about him. He's a he serves as a fractional executive, fractional C AIO, chief AI officer, as well as a chief technology officer. Um he's an AI entrepreneur, speaker, professor, and futurist. His innovative and visionary entrepreneurial executive mind has uh led him to do a lot of advisory. Um he's an accomplished uh engineer, data scientist, researcher, lecturer, as you can imagine, and published author. He has a passion for growing businesses with strong leadership experience with both professional and academic technology expertise in software and computer engineering with over 20 years of experience in leadership strategy, marketing, sales, operations, product development, and consulting, pretty much the end-to-end of running uh a business. Um welcome, David. Appreciate you being here.

SPEAKER_01 1:28

Always great to connect with you and spend time talking about things. You know, it's uh one of their my favorite things to do is engaging with brilliant minds.

SPEAKER_00 1:38

So Yeah, yeah. I I I I appreciate that because one of the things uh I want to bring up is that I was gracious enough to be invited by David to speak at Northwestern in a few of his courses. Um the one recently was with the um Northwestern Masters of Science and Law at the Prisker School of Law, and we talked about AI ethics, the state of AI ethics. We did some case studies, very, very engaging um students always, and and and David leads the discussion, and uh uh I really appreciate you bringing me in for those those opportunities.

SPEAKER_01 2:18

Uh that was probably one of my favorite classes uh that I've taught. And uh specifically having you come in there and really grounding the conversation. You can tell by the conversations and the questions. Uh it was there's no time for lectures in that class.

SPEAKER_00 2:36

It's just I know you have uh a time uh slot for lecture.

SPEAKER_01 2:43

Yeah, it gets smaller and smaller, depending on how good the uh guest speaker is.

SPEAKER_00 2:49

So yeah, so that that's always cool because uh uh really amazes me as the students and and what they catch on, latch on to and want to take a deep dive on. So it's it's really good. Let's talk about how AI is influencing education, um, I guess, especially from uh uh university and college level and how you guys are um tackling that, and especially at Northwest, especially what what are your views on that?

SPEAKER_01 3:23

Yeah, it's a great, great question. I uh I have I have uh probably the outliers view on these things because uh I I'm uh as as you said, I'm a futurist. So I uh I'm very optimistic about the future, but I'm also looking back and seeing the trajectory. I'm very pessimistic about um you know how how students are handling AI and how they're engaging with it is very uh disruptive, I think, to their well-being. Um I think the universities in the most part are not being very realistic about the expectation of what uh the real world uh is asking for. And so they're sort of like set in the past, and uh students are already well beyond, you know, catching up to what uh industry is asking, and there's I think there's a disconnect.

SPEAKER_00 4:22

Right, right. And uh yeah. Because they, you know, being a student and want to experiment and curious about the technology, they're probably playing with a bunch of different tools, learning what they can do, learning what's best, passing those experiences to other students, showing them what they can do and and what they shouldn't do. Yeah. I I think universities really have to have a policy around AI, uh specifically genitive AI, uh and how the uh what the university stance is and how they expect students to to leverage it, and also to be able to identify if you've done a paper or something in AI to have you know indicate that as a source, or uh some kind of policy in order to put some guardrails around that.

SPEAKER_01 5:12

Yeah. The the class you were that you uh last uh came to, the uh it's called the uh evolution of AI. Um a lot of that class focuses around critical thinking. And uh that is probably the most important thing you can teach students today because you don't without critical thinking, you really have nothing to offer anybody, right? If they in a world where AI can do all the coding and run businesses, and we could talk about that too. But um somebody has to orchestrate, somebody has to direct, somebody has to be in charge, and that requires critical thinking skills. And uh so a lot of that class was really around you know, let's look at what's going on in the world, where things are heading, you know, how do we tackle these problems? Do we just let AI run rampant and do whatever it wants to do, or do you set guardrails, boundaries, and so on? Now what I have found just in the last, you know, um the last two years, but really the last year, uh predominantly, I've seen a massive decline in critical thinking skills. And I think a lot of that comes down to the fact that students, more and more students, not everybody, but more and more students rely on uh the chat GPTs to be the first source of information. Instead of starting with their own ideas, they go straight to Chat GPT and say, what is the answer? And then they decide whether or not what to do with it. I really think it should be opposite. You know, chat GPT and uh LLM should be a research tool. It should be a tool to help you optimize to, you know, uh use it in your day-to-day, whether I don't care if it's coding or um even if it's to help you uh research a paper or come up with an outline, you know. Um I I try to be realistic about those things in my courses, but um what it does uh create an opportunity for is uh the ability to elevate the um the workload uh and uh the direction and um the critical thinking skills. And uh that that really takes uh a lot of work to overhaul you know courses and things to reshape uh and and catch up. I don't know that enough professors are doing that, and that that worries me.

SPEAKER_00 7:46

I think that is the most prevalent problem with AI in education. If we are limiting or kind of circumventing the critical thinking of students and they're not really exercising that critical thinking muscle and then becoming over-reliant on those tools, then that does the student a disservice. So when they do go into industry, you know, with working with a company or even developing their own business, if they lack critical thinking, none of this is going to work out well. What what but I think the to your point is the professors have to be intentional about having critical thinking exercises uh built into the courses.

SPEAKER_01 8:42

Right. You know, and unfortunately that requires a lot of overhaul of you know courses that you may you maybe have taught for 10, 20 years, um, you know, with certain updates and stuff. Some of them have exams, some of them have projects, but I think, you know, taking theory and applying it uh and even giving them more challenging problems, you know, so things that you could have easily coded up uh as a s a solution now, now coding is just one part of the output. Really now it has to be a broader, um broader uh type of project where you have to do all the research, you have to understand the broader impact and the problem and what you're trying to solve and why you're trying to solve it. And um you have to, I think, elevate the you know, now you have all these tools that you can that can help you to solve the output. So coding could be just one part of the output, but I think it's that whole journey that's really important to understand. You know, and it reminds me a lot about uh sort of the entrepreneur's journey. You know, a lot of people say uh, you know, when you uh create a startup and then it fails, it's like, oh, it's not really a failure. It's that lesson that you learned creating a startup and failing that you that it helps educate you and become a better entrepreneur on the second time, third time, and so on. Um but had you not gone through that journey, you would never develop those critical thinking skills, right? And so if you used an AI to just build your company and put an autopilot, you're not really uh solving, you know, that that problem, that developing those crit those muscles, you know, that you were uh uh talking about. So I think it's it's uh it's a careful balance, you know, it's a juggling act. And um I don't know, I think there's going to be some uh you know, people will go through some hard times first uh before they find that that proper balance and how we how we juggle that, how we do that at uh university level or even at you know, grammar school, um middle school, you know, uh these tools are not disappearing.

SPEAKER_00 10:56

Right, they're just getting better. And I can't, you know, I I just can't see um the lower grades using these tools. I mean, I I'm pretty sure they do, or at maybe at home or whatever, but I I I I can't see that uh as a part of the uh the classroom experience. Not in the the elementary and middle school level. They got to develop that. They have to understand what this is there they're learning, what's the steps of figuring this stuff out before they use a tool. And then on the other side, the vendors of these tools have to also educate on how to best use them. Um the worst thing you can do in my mind is use chat GPT for search. You know, that's that's not that's not what these tools really, uh, the strengths of these tools is not is not that. Um understand what how LLMs work at all, you know, because it's not you know learning anything. It's the only thing that's learning is how what's that next word or next phrase that you're typing in to when you generate text? Or um you have to be you have to build up skills in our case, the English language, and and have a chain of thought prompt to understand, okay, what you want to create and making sure that you create what what is intended, that the tool creates what is intended. So that takes a level of understanding the language, the constructs, just like programming. Yeah. And and so you can guide the output, um, uh, even though these tools get better and better. My belief is that we would we are going to need very skilled people at critical thinking, subject matter experts to be able to discern if the tools are actually delivering what's intended. Right. You know, if you don't have that, how can you know what the tool is doing is correct?

SPEAKER_01 13:00

Right. Now, go taking a step back, you know, going back to you know, in the middle school and stuff, there, you know, we we here in Chicago, we have a huge problem with uh education. You know, a lot of uh most kids in the public school system are reading and doing math like below grade level. How do you catch those students up? You know, and especially if the parents are not engaged or they don't have the the skill sets. So I think there is a place for AI in that. Interestingly, I I was just a friend had afforded me this article. Um, a couple of professors at at Kellogg, a Northwestern, um trained a model uh to uh for students, but it wouldn't give you the answer. What it would do is help you think through the questions and and guide you towards an answer. I haven't had a chance yet to play around with this, but I think that that's the right direction is providing uh a like a tutor uh tutor, right? It wouldn't just give you the answer, but they would help you work through the problem and and such. So I think that that something like that could be very helpful for kids who, you know, maybe the parents are not strong in math and they need a tutor and can work with that to towards uh getting to it. But I think uh that you know things like that have to be more prevalent in universities. Like, hey, these are models that you could use because they're designed not to provide you the answer, and that becomes sort of like the um you know certified uh type of of tools that you know great schools and stuff and help get kids caught up. And because if you're not set up for education, you know, at that level, I mean you're you know uh unfortunately right now you're written off, you know.

SPEAKER_00 14:43

True.

SPEAKER_01 14:44

And um, so we want to try to you know reduce that gap as much as possible. And uh so I I think that there could be opportunities there. At the same time, obviously, like the at the youngest, you know, at those young ages, that's that's when the critical thinking skills are most uh important to help develop because you if you lose that you're also right. There are ways of developing those critical thinking skills, sports, um, you know, team building exercises, you know, different clubs and stuff where students you know can work on those things. But uh, you know, some of the schools now they've been cutting budgets and now they're there's no after-school sports. And there's and um so I I you know in many cases we're doing a disservice even to the young young kids these days. So it's uh it's a lot, you know, uh, I think the onus is on the parents to um make decisions on behalf of their children and uh be very vigilant about what they're getting into and uh uh but also try to set them up for success if they're not getting enough from the schools.

SPEAKER_00 15:51

Right. Yeah, pro ac proactive parenting and uh as a partner in the in your child's education is uh extremely important. And um, you know, finding that balance on where AI can be used. Like you uh I mean I agree with you a hundred percent that uh AI can be used as a tutor, as something that can uh aid the um uh instructor um in this case to to help this the students that are sort of behind or or below level to put that r rise that level, yeah, raise that level to to where they need to be. And I think um there'll be like you say they need to be more I'm not I'm not aware of what is out here as far as the educational piece for AI, but I know there's some AI tutors and things like that that uh can probably probably be adopted. Um so from from from there, uh I I just want to add that one of the things I uh I'm doing, and I'm um I'm part of the founding editorial board of AI and Ethics Journal by Springer. Uh recently uh became the associate uh editor for AI and education, that's part of the Springer AI and Ethics. And and and we um receive a lot of different research in uh uh AI and education. And and it ranges from K to 12, uh, which is uh a lot of focuses on, and also how uh universities can best leverage AI in different curriculums. So um if if you want to submit a research to that, to that, that would be fantastic. Uh encourage your colleagues to do so just to continue to um you know raise the level of attention in in that in that area.

SPEAKER_01 17:42

Yeah, one of the things I'm working on with uh some students, I have probably about uh a dozen or so different research projects going on at the same time. But one of them is on developing new skill sets. Because what you know, what inevitably is going to happen is a lot of people are gonna get, you know, either laid off or jobs are gonna change because they're gonna become almost unnecessarily relevant, but you know, media um uh duties that uh require uh a lot of sort of repetitive uh workloads and things like that, th those things will probably be optimized with AI. So some of those jobs will be be displaced. Well, what do those people do? So they need new skill sets. Well, what are you gonna go back to college at 40, 50 years old to get a new skill set? No. So I think uh learning new skills outside of a university uh system is going to be a huge need. Now there's c uh you know companies like Coursera and um Udemy and stuff like that, which are already doing that. People can you know create uh um but you know uh those are predefined topics. But let's say there's a new topic. I want to become uh I don't know, I want to become a plumber, I want to become uh a um mechan auto mechanic. You know, I don't know if you can find that in Udemy. Right.

SPEAKER_00 19:06

I I I know a lot of that is actually on YouTube. YouTube has is a is a great source for for learning. Yeah. And a lot of professors put their courses on YouTube. Yeah. And you have to search it, right?

SPEAKER_01 19:18

Yeah. But can I create a syllabus of uh, you know, uh, I want to become a quantum engineer. What does that require? Okay, so overnight uh quantum engineering becomes like the new thing, and okay, I need to figure out all of the you know, the the main things to to uh that I need to learn. And it could be maybe I'm a business person who needs to lead a quantum engineering company, or maybe uh I'm you know, uh somebody in grade school who's really interested in quantum engineering, and so you have to have content at different levels. YouTube doesn't do that, Udemy doesn't do that. Can I then take all that content, distill it down to something for an eighth grader, something for a high-level business person? And and um, you know, I think that could be a great way to develop new skill sets for people who want to shift jobs and professional in different areas.

SPEAKER_00 20:13

Right. Uh you know, talking about shifting of jobs and displacement, you know, AI, any technology is going to create new jobs. Right. But but like you said, you know, everybody's not gonna be transitioning to those new jobs. Yeah. But so what do they do? Like you pointed out, there are all the trades that are being overlooked, like plumbing, HVAC, um, electrician, all those things. Um they are still gonna be needed. Right. And so there needs to be uh sort of a pivot to to some of that. So and a fast track to say, okay, if I'm being displaced by AI and uh being an electrician is something I'm on that may want to pursue, or HVAC or whatever, what give people a path to that and understand what can be done to learn that um and make that available um you know online for free, whatever. Like the universities got you you have Harvard and and who else? Is it MIT in Stanford that just released all their courses on AI for no charge? I know Google has it, uh Microsoft. You know, one of the reasons they're doing that to your hit your point earlier is that they see this gap not only for people need to be reskilled, but it's a gap period for AI folks, people that versed in that understand AI, and it's not enough people to even take these jobs without the proper training. Right. So it's it's that too. And then students, I've talked to a lot of students at computer science that I you know wonder am I gonna have a job?

SPEAKER_01 22:00

Yeah. You know, um high schoolers are now learning this stuff. So you're getting out of high school and you already have the the knowledge base. You have companies like uh Palantir, right, who's just hiring kids out of high school and bringing in professors to teach them their learning on the job. They're and uh they're finding the best and the brightest right out of high school. There is uh, you know, I tell my students because I teach this um AI innovation lab course that that you've spoken at uh quite a number of times already. And uh that class has gotten you know even better. Every year it gets better and better because all these new tools and capabilities that come out. Um and I I tell my students there is absolutely no better time to start your own company. Right. Because you have all the tools, all the automation, all the capabilities. You don't need a massive team. You can hire two, three, five people and run a solid company. And uh you don't need anymore to hire a hundred people. You don't need these big companies. So when you're talking about like the new opportunities, I think what is going to happen is you're gonna have more and more small businesses popping up. Um and you can train people to become entrepreneurs, to run a business, you to run a Shopify, to run a, you know, uh if you got a great idea, now is there's no better time in the history of the world to start your company. And that's true. Or a few.

SPEAKER_00 23:28

Exactly, because you know, you can uh with some foresight, you can create those employees, those agents. Yeah. Um what I've I've talked about in the previous episode is is AI strategy, even from a personal level. Uh figuring out what's your strategy for this, because even with folks in the industry, we're finding stuff changes so often, we're like, oh my God, this we how do we keep up? Right. And I always tell people to pick a lane. Pick a lane and get very good at it. Yeah. And and then, you know, source those um courses or you know, those opportunities to learn more. But that's all a part of your personal AI strategy to be able to do this. And then if you run it, if you want to start a business uh and leverage AI, that's gonna be huge. Understand what's your AI strategy for your business. And so you can be that orchestrator of all these agents that's running parts of your business with your key, you know, key people, five those five or six key people, and understand what that what that is. So, you know, that I think that's that's gonna be popping up all over the place. Yeah. If you're enjoying this conversation, please stay tuned for part two. Uh, don't forget to share this with other inquisitive minds in AI, and we will be exploring more in AI in future episodes. Remember to like, share, comment. See you next time.