839: How to Prepare Faculty for the AI Revolution - with Chirag Tailor, CEO & Co-Founder, Instructify

Joe Sallustio: Welcome back everybody. It's your time to EdUp on the EdUp Experience podcast where we make education your business. Dr. Joe Sallustio back with you again, again and again. No, we will never stop. In fact, we're recording this - it is now the month of February 2024 as we record this. Elvin has got me booked out daily, and I mean daily, other than Saturdays and Sundays, through the end of April. I podcast every day at 12 noon central standard time, same bat channel. So we are busy and we're busy around having the important higher ed conversations that need to happen as administrators, as EdTech CEOs and founders, as business and industry. And we do it all really for students. We have to celebrate the amazing parts of higher education because our industry, I think, is on a comeback. And we've got engaged students, faculty, staff all around us. We just want to get them on the microphone and have them tell us their story.

And one person that's really, really good at telling stories is my guest co-host today. In fact, he's been a president for a really long time, at least as long as I've known him. He's been in the same place. He's been there for years. When he came on screen today, he had a full-on picture of a beard that you would be envious of. Hemingway, I don't know who, maybe Santa Claus, although without the gray. Ladies and gentlemen, we're going to bring him back to the microphone. It's been probably three years since we interviewed him the first time. Here he is. He is Dr. Nathan Long, and he's the president at Saybrook University. Nathan, welcome back.

Nathan Long: Great to see you. Good to see the robustness of the podcast coming in, what, four years old almost, I think. We need to talk to Elvin and get you on Saturdays and Sundays too.

Joe Sallustio: No, no, we're not going to do that. We're not going to do that, Nathan Long, no. But a quick story, Nathan, you will always have a special place in my heart, my friend, because as I was getting down to my dissertation, I was in Southern California. I worked at an institution down, say down the road in Southern Cal, which could be like two hours away, but it was down the road. I was like, I need people to get me faculty who could be interviewed and contribute to my survey that I had going on. And I wasn't having much luck. And I called you and you opened it up to your entire institution. And I got great feedback and it allowed me to finish my dissertation on time. Thank you so much, my friend. Without you, I'd be ABD still to this day. I promise you it would have never happened.

How does it feel to be back? You were interviewed so long ago, probably before I had sound effects. Now you're coming back to interview somebody else, kind of full circle.

Nathan Long: It's exciting. And I've been really looking forward to interviewing Chirag Taylor and seeing you again. I think we've all grown a little bit. You're at a new institution. I've been at Saybrook now. This is my 10th year. So lots of great things that have been happening. I think you hit the nail on the head. Higher Ed's at a crossroads and today's interview is no exception, right, with the advent of AI. But I think, you know, kind of more broadly, we've got a lot of stressors, but there's a lot of opportunity and reason to be optimistic. I love your boundless optimism. I think that's what we need in the field. A lot more of that to show what we can do and what we're capable of doing. So really excited to be here and looking forward to hanging out with Mr. Taylor as well.

Joe Sallustio: Well, we're gonna find out a little bit about what the future holds for higher ed today with our guest. He is the one, the only Chirag Taylor. He is CEO and co-founder at Instructify. What's going on? How are you?

Chirag Taylor: Hey guys. Thank you so much for having me on the call. I'm really excited for today's conversation and really excited to talk to Dr. Long and Joe. It seems like you're very busy on this podcast too. I heard about it almost late last fall, early last fall, after attending some conferences. We did a whole circuit, conference circuit, like six conferences. And one of our colleagues, Derek Newton, told me about this and he was like, you need to go and sign up for this podcast. And I think I was looking at like September for a slot. And then I went through your calendar and it was like September's full, October's full, November's - I went all the way to February and I was like, wow, all right, this is exciting.

Joe Sallustio: We do Saturdays and Sundays. I know. Yeah, we gotta get Joe on the weekends.

Chirag Taylor: It's not gonna happen. Not gonna happen. I think I would be divorced and my wife would, I may be dead. My wife would kill me.

Joe Sallustio: Shout out to Derek Newton. He is a good dude and we're glad to have, you know, he's writing for Forbes and he's got his blog that he writes, the cheat sheet. So he's got a lot going on out there. And one of the areas that he's focused in and around is AI in colleges and universities. He has, he's very opinionated dude. Let's put it that way. But tell us about Instructify. What is it? What do you do? How do you do it?

Chirag Taylor: Absolutely. So Instructify is an AI platform. We are the developers of a teaching and learning intelligent assistant. The acronym we've kind of given her persona, we call her Talia. So Talia is a co-pilot for higher education that serves both students and faculty in the higher education system. For students, she's an additional resource that helps you navigate the course, get personalized, individualized, in-class assistance for the specific classes that you're in. And then for faculty, she's there to help you do kind of the tedious tasks that are required to conduct a large course, to serve a large number of students in a virtual or hybrid environment, and kind of free up some time for them to do more human-to-human interaction. A big slogan that we're trying to run with right now is we're helping faculty win back their Saturdays.

Joe Sallustio: Amazing. Yeah, we're very excited to work with faculty and administrators to learn a little bit more about how we can help them save time throughout the year. What is a - so like I know Nathan's going to have tons of questions here too, but you think about any AI integration of any kind as a supplement or complement or as a hope or necessity is freaking people out. You know, faculty in general and staff. We don't know how to use it, whether we should use it and then students the same, right? I had a conversation with a student the other day, literally on Friday, who is having trouble in a subject and says, I know I could use AI, but I'm deathly afraid to even ask it questions because I think they're monitoring me and someone will find out and I'll get an immediate F. And I'm like, I started looking around going, well, I'm using it. I don't know what's going to happen to me. But what's the response like so far of schools that you've talked to, faculty that you talked to on just the integration of AI?

Chirag Taylor: Yeah. I mean, do you want to go first Dr. Long?

Nathan Long: No, I'm excited to hear from you.

Chirag Taylor: Okay. Yeah. Exciting. No, we started this journey actually a few years ago. I have almost a six-year career in AI. I got my master's in artificial intelligence from Georgia Tech in 2017. And since then I've been in industry. So I've kind of followed along the trend of industry, but it's really exciting to see how higher education has been adopting AI. When the large language models came out and ChatGPT blew onto the scene in November, the first immediate reaction was just fear, right? So there was a lot of people that are like cancel ChatGPT, like block the domain from university campuses. You can't access it while you're on our WiFi. And then like last week or almost two weeks ago, ASU announced this enterprise license partnership with OpenAI. And they're kind of, it's a full 180 from where we were exactly a year ago. We're blocking it to like, now let's look for creative ways to integrate it into our platform and our teaching environment.

Joe Sallustio: I think that's unique though, right? I mean, Arizona State and Nathan, I'll pass it over to you in a sec, that was kind of the first, I would say groundbreaking. The first university goes, okay, we're going to take this, we're going to do something with it. We're going to integrate it. We don't even know what that means yet, but still, I think is there resistance to this idea of AI and how, should, when, I mean, there's so many nuances, right? What do you think? Is there still resistance?

Chirag Taylor: I think there are still some people that are wary of it and they're concerned for different reasons. The biggest problems are around like hallucinations and, you know, trust and whether the AI understands their specific problem and their specific concepts. Like a lot of priority that we're taking internally is one building a, you know, a platform with high trust and high visibility, but also like helping faculty customize the experience to their specific course so that they understand that this is not just some general model that's trained on everything on the internet, but it's like a personalized model for their course, their assignments, their lectures, their course materials. And Talia is really there to serve as kind of like a right hand to help them pass on knowledge to students.

Nathan Long: Yeah, that's really cool. I'm curious, you know, I've seen some not so great implementations of AI so far, like with AI chatbots and whatnot. Instructify, I have no idea yet. I'm sure we'll learn more and may even come to adopt it over time. One of the things that I've been really, I think, kind of mystified about is how tough it is for the machines to actually learn, right? Like the algorithms themselves. We've seen chatbots that they still are like, can't find that information, don't know what you're talking about. Financially, they'll push you over to like Bank of America or something like that. And that's like the wrong answer. How do you, how are you managing through some of those issues, especially where machine learning is concerned with Talia? And making sure and improving the work that the AI puts out in support of faculty and students. Does that sound like a question?

Chirag Taylor: That's a great question. I mean, the first thing that we've done is we've implemented this technique that's being used very widely in the large language model space called retrieval augmented generation. In a nutshell, what it does is it constrains the worldview of the language model to a specific set of documents that we call the knowledge base. And the first operation when a student asks a question or a faculty asks a question is the model retrieves subsets, chunks within those documents that are most related to the question. So it retrieves relevant information from a very constrained document set that the instructor or the administrator predetermines. So it's not going out to the internet. It's not going to Bank of America. It's looking at the documents that you've selected, like, "Hey, these are trusted sources that I want you to refer to and then give me answers from that." And then once it pulls relevant context, you know, it could be like 10 chunks of documents, 20 chunks of documents, that context, along with the question is forwarded to the brain, which is a large language model. And then, you know, from that, the answer is determined. And because we have the chunks and we have a vector store that keeps track of the documents and all the chunks of the documents, we can also do attribution. So this is kind of like our first step. I know it's not a perfect solution, but it's our first step towards fighting problems around hallucination, fighting problems around like misattribution of answers, or, you know, making sure that Talia doesn't, you know, just give any outlandish response. So she refers to a fixed knowledge base. And on top of that, she provides references to those knowledge bases. So if she pulls, you know, page two from a syllabus to answer a question, she can provide a little tag in the user interface that says, "Hey, here's the answer, but I also got this answer from this specific page of this document."

Joe Sallustio: All right, nice, nice. That is amazing. It really is. That's very cool. So you're then in the testing phase, I assume, you're kind of working this with faculty and students to see how this is all threading through and kind of mapping out appropriately? Is there some rigorous pilot testing that goes forward with that?

Chirag Taylor: Yeah, absolutely. So we were doing an extended pilot with Georgia State University right now. We're working with the math department over there to put Talia into three separate math departments, over 300 students are using Talia right now. And one thing that we're all really excited for is not only to see the platform in action and see how students are interacting with Talia, but more so also have an opportunity to collect feedback. So one of the mechanisms that we've implemented as part of this pilot is allowing students to flag responses, right? Hey, that's a good response, that's a bad response. Allowing instructors to endorse responses and be like, "Hey, that was a great response by Talia, I'm gonna endorse it so that my students know that I, as an instructor, agree with what she's saying." So we're also implementing these various feedback mechanisms and endorsement mechanisms to help build trust within these interactions. And then from this, we can probably go back and do more advanced machine learning techniques to improve the recommendations and answers.

Joe Sallustio: Nathan, I want to follow on and pull the thread a little bit that you were kind of leaning into earlier. I'll try and put you on the spot a little bit here, right? So, because I think what Joe brings up is an important aspect. You've got, yeah, you got the ASU's early adopters first, ASU is always the first adopter, right? Michael Crow's the man, if you will, in higher ed. But for the rest of us out there, right? We've got faculty and staff, students who are freaking out, right? Or, you've got a whole continuum. I'll give you an example. We were in a faculty meeting about three months ago and one of our faculty members, big adopter of technology, had the Zoom AI chat meeting assistant coming in to take meeting minutes. So he was going to be late. Everyone was like, "What's going on here? Oh my God, you know, the world's falling apart." Eventually it took them like two or three months as they started looking at this. They're like, "My God, we don't have to do minutes anymore for meetings. Our lives are a lot easier."

And it really was like a small ramp up with our faculty because they're adopting technology. So I have a lot of respect and kudos for them. But in general, I think we're seeing a lot of anxiety. Maybe it's unfair, but I would say some draconian measures around plagiarism detection with AI and kind of coming at students with a heavy hand on this to not really thinking through the opportunities that come with AI to enhance teaching and learning rather than seeing it as more of an impediment to. So it's a long-winded way for me to get around to the question of asking, how will you bring the value add of this to faculty and staff and students beyond the "it'll make life easier" kind of thing, which it will, and it sounds like an amazing tool, but what are you all working on in terms of I guess the marketing pitch to really pull people in on this?

Chirag Taylor: Yeah, there's a lot of feedback that we've gotten from faculty. At Georgia State, one thing that they're really excited about is personalizing responses to students. So there's this part of the science of learning that you guys are probably more familiar with than I am. But I'm learning this as I'm talking to more teachers and educators, instructional design, and the science of learning. Students and anyone who's trying to learn something is more capable of grasping a new concept if they can relate it to something that they've already experienced. So I was talking to the assistant dean of academic innovation at UCLA, Dr. Worley, and he's actually doing research on this. And he showed me a very concrete example in ChatGPT. He asked me, he's like, "Are you familiar with cell division mitosis in biology?" I was like, "Yeah, I learned it at some point in seventh grade, eighth grade. I understand how the chromosomes find out they divide." And then he's like, "Well, if you ask the textbook definition, there's all these phases, prophase, telophase, metaphase, and they're all labeled and students have to understand all these things when they take the class." It's just biology jargon, right? When you're trying to pick this up, it's a block of text and you're like, I can't make any sense of this. And Dr. Worley asked me a question. He was like, "What's your favorite TV show?" And I was like, "I like West Wing by Aaron Sorkin. I watch it. It's like this presidential drama." And then he asked ChatGPT to explain mitosis as a screenplay from West Wing.

Yeah, it was so interesting. So it wrote out this entire script of like, you know, there's the president and his like, you know, staff and cabinet members inside the Oval Office. He's trying to, you know, attend some kind of lecture about like a biology class and he needs to understand this concept so he doesn't make himself look like a fool. So then all of his like cabinet members are helping him understand different aspects of this process. And then just, I can see the whole scene purely in my head. I can hear the characters talk. I can relate it to like watching the TV show. And then it sticks a lot more to me. And he was like, "In that context?" I'm like, "Yeah." And I was like, what if we can do that for every single student? Individualized to their preferences and their learning. 

So one thing that we're trying to experiment with at GSU is, you know, they have, we're rolling out Talia through a Discord server. So it's like this kind of communication platform and there's channels and stuff. The professors always create an off-topic channel to just talk about random things, build, you know, like a little bit of, what do you call it? Build a little bit of banter with their students, understand a little bit more about what their students are into. And then we're trying to see if Talia can host certain conversations in off-topic channels. Like, "Hey, tell me a little bit more about if you have any pets or if you have any music interests or TV show interests." And then once we've learned these things about students, can we use that information to bias the output? So if Timmy asked the question, it would be a very different answer than if Sarah asked the question based upon what their personal preferences are.

That's like one of the things we're experimenting with. Of course, there's also like lots of instructional design concepts, right? I think next year, GSU has an initiative where they're trying to customize assessments for students based upon where their particular roadblocks and obstacles are in a course. That is an intractable task. If you're trying to teach 200 students how to do algebra and you have to design a customized assessment for each of them, but with AI, it could probably be done.

Joe Sallustio: Are you kidding me? No, I'm not. For a third straight year, the EdUp Experience will be recording live at Ellucian Live 2024 this year in San Antonio, Texas, April 7 through the 10th. Illuminate, innovate and inspire. That's the framework for the conference. Leaders from institutions around the world will converge at Ellucian Live 2024 to discover game-changing technology, share industry insights and build powerful connections. It's time to explore higher education's greatest opportunities with future-ready ideas, solutions and best practices designed to drive transformation. You can register now at elive.ellucian.com. Epic.

Nathan Long: Yeah, for those who are like in the graduate education space, we're having this conversation now too around the changing face of the thesis for master's degrees and dissertations, right? You know, this notion that you have to write or type a final product, do we really need that anymore? Or is there some other way to get to that sort of synthesis of knowledge leveraging AI? And I think you're onto something really fascinating here. I mean, would love to see Talia kind of weigh into that space as well and kind of leverage some of that. I know graduate education is in need of an overhaul.

Joe Sallustio: Poor Joe is probably freshest off the train on that one. And I go, by the way, want to say a note that that whole long answer that you gave, Chirag, the one takeaway for me is that you had to look at Nathan and I and describe what Discord was. Just because we have gray hair doesn't mean we don't know what Discord is. I know it. I use it. OK, most people in higher education are a little bit kind of... Nathan, Discord is a conversational platform. I just want you to know.

Nathan Long: Thanks, Joe. Thank you.

Joe Sallustio: No, that's great because some people don't know what - and you know, the funny thing about AI is that social media moves kind of similarly, there's always something new, always some new adoption. What's the value prop for the individual faculty member? You walk in, I think if you get an administrator or a dean or somebody that looks at this and goes, wow, I'm seeing all the possibilities. You get down to the original, the individual faculty member and you say to them, "Hey look, you're going to take, I'm imagining how this might go, you're going to take all your course content, all the stuff that you use, all the stuff that you, what a student would interface with and you're going to basically shove it all into this technology. And then what the student is going to be able to ask that technology, Talia, questions and get answers back about the course. Instead of you taking all of those questions by email and phone call and having that turnaround time where you're panicked to get back to somebody and then you can't catch them and instead that student's going to be learning on their own in a way and you're going to be able to focus on the lesson." I mean, you know, really explain to me how - I made that up, but explain to me if I'm even on the right track or if it sounds different.

Chirag Taylor: You nailed it. You nailed it. You hit the nail on the head there. That's exactly the first value prop that we approach faculty with. A continuation of that, the only thing I want to add there is if you're a faculty member and I'm sure both of you have experienced this, you just finished your dissertation and Dr. Long's been in academia for a very long time. And so when you teach a class, you are an expert in that class. You have a PhD on the topic. You know this concept like the back of your hand. But then when you're trying to teach it and you're especially trying to do it in a hybrid or online environment at scale, you can't reach every single student with your knowledge. So what do you do? You employ the help of teaching assistants and you say, "Hey, here's either a PhD student or master's student, somebody who is familiar with the topic. They're under my tutelage. They understand the concepts and you can go to them first. If they stop, then you can come to me." Right. What happens when these teaching assistants graduate and they go into, you know, the real world? Industry or to go into, there's a lot of turnover there. I was a teaching assistant myself. So a reason why we got into this, my co-founder and I were both graduate teaching assistants at Georgia Tech. We saw what it was like to go through this process behind the desk for the OMSCS program at Georgia Tech while they were scaling up their online learnings. Like we started with like a hundred students in a class and now that same class has 900 students online. One of the things that Talia provides is continuity in that teaching, right? So you upload all your contents into Talia. You are essentially building a second brain for yourself. And then students can go to Talia for references, but each interaction with Talia allows her to improve. And then she can come to the faculty and say, "Hey, the students are asking me X, Y, and Z. I was able to answer 80% of these questions from the knowledge base you gave me. The other 20%, I don't know how to answer. How would you answer them?" Teach Talia how to answer them once. She updates her knowledge base. She re-learns the concept. She, you know, augments the content that you gave her with this additional information. And then when the student comes back again, she has an answer for that question.

Joe Sallustio: So if I, and I can say, okay, Talia, I know I've loaded you with a bunch of information and I'm the faculty member. Now some of it's old. So a year later, two years later, I don't want you to use it anymore. Yeah. Do I? Can I have Talia learn not to provide certain information and is that going to disrupt outputs potentially? You know what I mean? Like how do you control inflows and outflows and all that?

Chirag Taylor: We have a robust web platform that we've designed for administrators to be able to manage the knowledge base. So essentially what you were doing is as a user of Talia, you are providing her a knowledge base, a set of content that she can reference. And if things change in your course, you want to either update it or deprecate it or remove it or add to it. But that is the interaction there.

Joe Sallustio: Chirag, but now I'm going to be the cynical faculty member and say, but why would I use this? Because I'm just basically giving the student everything they need to cheat. They're just going to copy and paste the stuff that they ask it into the paper. What am I supposed to, how am I supposed to even know? I've fed it 100 documents. How am I going to know? And students in general, that's what they do. They're just going to look something up, and they're going to copy and paste it, and that's the paper. And you know what? Everything is AI these days. So says me, the faculty member.

Chirag Taylor: No, that's a good concern. The benefit of using Talia over another tool like ChatGPT is you get to put guardrails on it. So you can come to us and say, "Hey, Instructify, there are certain questions and there are certain types of questions that we want to just have Talia give a blunt answer, provide the attribution, do the references." For example, you know, what room on the campus am I supposed to take my final in? That's somewhere hidden in the syllabus or somewhere online. The student just needs an answer so that they can get to their final exam at the right time, in the right place. That's not something you want to be clever about how you answer. But if it's something related to an assessment and the student's like, "Hey, I need to understand how the central limit theorem works," like, you know, "What is the central limit theorem?" That's where Talia can be a little bit more Socratic and kind of use that as a teaching opportunity and say, "Hey, well, you know, what do you think it means? And like, you know, where did you learn this from? Or what did you learn in class?"

We're experimenting with different prompting techniques inside of Talia that will, the first decision is, is this an opportunity to use a Socratic method or is this somewhere where we just need to provide a blunt answer from the knowledge base? So if we can differentiate from that, be like, "Hey, this is a logistics question, it's somebody that just needs help immediately, boom, give them the answer, we're done." But if it's something that we can determine is like a test question, an assessment question, somewhere where there is an opportunity to provide understanding and learning, then we can trigger a Socratic mode within Talia.

The Socratic mode is currently under development. We've been hearing about that a lot from a lot of faculty members. A lot of people are very excited about it. In fact, like UC Santa Cruz, the UC Scout program - I'm not sure if you're familiar with that Dr. Long. We got an RFP from them. They work with 30,000 high schoolers, I think, to help them kind of get into college track, and they're looking for an AI tutor for them. And one of their hard stipulations is it needs to provide Socratic responses for assessment-based content-based questions.

Nathan Long: That is fabulous. It is. How long do you think it's gonna take for you to kind of hone the Socratic side of that work?

Chirag Taylor: Yeah, technically I don't think it's very difficult, right? So there's two challenges there. We need to understand what type of question it is, whether it's a question that we want to use the Socratic mode on. And that again, I think would be very simple because we could just look at like, where's Talia having to pull the references from in order to answer it. If the majority of the references are coming from syllabus during that RAG (retrieval augmented generation) mode, then, "Hey, this is a logistic question. Just provide the answer." But if a majority of the references are coming from, you know, your lecture notes or your textbook or something, it's like, "Okay, this is probably an opportunity to trigger Socratic mode." Then the Socratic mode is actually just a series of prompting, right? So you provide the large language model with additional prompts and say, "Hey, find the answer to this question. Don't provide the answer directly to the student. Now that you understand what the answer is, ask a question to try to extract the information from the students themselves."

So we're experimenting with what's the best method of doing, whether it's a prompting technique, whether it's a separate large language model that we fine-tune just to provide these kinds of responses. But behind the scenes, Talia is like this whole orchestration of decisions behind the decision tree of like, when do we want to use a certain machine learning technique? And when do we want to use a certain language model? Do we want to have multiple language models behind the scene? So that's the behind-the-scenes brains of Talia that we're working on.

Joe Sallustio: That's impressive. It really is. I have, before I know we're coming kind of to the close here of our time together, I have a more, I guess a broader question for you, Chirag, because you're an entrepreneur, right? You're like two years into this venture. You're a startup, you're a business guy, you're brilliant in so many ways. So what's it like running a startup? I've talked to a number of entrepreneurs in my time and it's always fascinating to hear from them and what kind of advice would you give to those who are out there, especially in the ed tech space? Like what to think about as a startup entrepreneur?

Chirag Taylor: The short answer is very challenging. I think that I could have probably picked an easier route towards my career, but I don't think I would have been as fulfilled. Something, and because we have to find reasons to kind of keep going. And I think my personal reason is this concept I read about, there's a Japanese concept called Ikigai, and the whole summary of that is you should find something that you're not only good at, but something that benefits the world, and that you can make money doing too. So it's a combination of not just it makes money, but it also benefits the world. You're very good at, and, then there's a fourth part of that Venn diagram too, that I can't remember off the top of my head, but it's a more holistic approach to how you should find purpose in life. And I think this is kind of my Ikigai right now. I'm very interested in education. I was a tutor before I was a graduate teaching assistant. You know, I got my advanced degree because I thought that like in order to do anything meaningful with my life, you know, be very learned. So I really enjoyed the process of watching other people, you know, kind of mature through the learning process, get their own degrees, go off and do exciting things with what they learn, and helping them kind of unlock these new opportunities and a very exciting experience for me.

Nathan Long: Well, and let me just say, there's a lot of courage that goes into what you do. And this is gold, man. I think, honestly, I can't wait to learn more about this from you and your team. I think there's a lot of applications, not only for Saybrook, but for higher ed in general. You may be onto something really super big. And as you know, I think you're kind of living just keep showing up every day and keep trying. I mean, not that you're looking for advice from me, but so inspired by what you've been talking about, what you all are developing. I hope you stick with it because you're in the hard part, right? The building. And once you get through that, could be, I think it's going to shoot off, shoot for the moon.

Joe Sallustio: Anyway, I got to say, I agree this, you know, the, and maybe that we could provide some perspective to for you, Chirag, because excuse me if I ramble here for a second. I think this is super cool. It would be for me. I look at this and go, why would I not add this immediately? Higher Ed though, by its nature, prevents new things from coming in. New entrants, new things, new ideas. As much as we are an institution that teaches, we don't like sometimes when new things come in and tell us how to do things. So you're going to have your really early adopters or people that look at this and go, wow. And then the majority of people are going to go, what is this? How is it going to affect me? Is it going to replace me? You won't even need me in the classroom anymore, and therefore I'm against it. I think if you could sell against that, boy, you're shooting for the moon on this one. But that is, break things down. A lot of times, that's where it all comes down to is what's going to, there's a lot of things. Do you think about that? That we see from people in industry.

Chirag Taylor: And, uh, you know, the example I'd like to give is, um, you know, 20 years ago, 30 years ago, before CAD architects were drawing all the diagrams by hand before we had learning management systems, faculty was dealing with hundreds of papers. They had to keep up with and grade manually and stuff, but you know, these systems are designed to not replace people, but augment their current capabilities. If we do it right and we really pay attention to human-centered design, which is a big priority on our end, is helping them do more with less time. Get back their Saturdays, focus on the real portion of what education is all about, which is building connections with students and allowing them to unlock new opportunities with what they learn. And then kind of have Talia take care of all the peripheries and the tertiary tasks that you need to do as part of like accreditation and grading and learning and stuff, but you don't necessarily have to spend all your time on.

Nathan Long: Well, and you have faculty out there who are totally willing to break the mold, right? So, I would just, for anyone who's listening to this, like break out of, you're in the mold and you're not willing to break out of it, like the assessments that we do today, those should be a thing of the past in tenure, the dissertation, the thesis, because what Chirag's building, what AI is bringing is something more powerful in terms of learning, human learning, that can really blow the lid off, I think, of the inventions and the innovation that we're looking at as societally. So pedagogy, pedagogy, pedagogy, let's up it, let's get better at it and leverage AI, what you're building, Chirag. I mean, that's going to be a huge catalyst for change, I think, going forward.

Chirag Taylor: So anyway, very excited. The one thing I'm really excited about is to work with like instructional designers and faculty to really understand the pedagogy. Right? So we're technologists. We care about education. We probably are not experts in the science of education. This is where we can get a lot of feedback from. So I love having these kinds of conversations with people and understanding like, what can we build into the technology that's, you know, science-backed and can really augment the learning experience, but like, especially on the dissertation phase, what has been, have you thought about like oral exams? What history is there?

Nathan Long: There's a whole lot, you know, went back in the day, right? I'm old and I actually have less gray hair than Joe, but...

Joe Sallustio: Yeah, it's true.

Nathan Long: But back in the day, where I was at in Cincinnati, we had started dabbling and then it just exploded into alternative theses and dissertations, multi-modal kinds of things. We had students not writing much of anything, but doing multimedia presentations in the old school way, right? You know, uploaded it before files, that kind of thing. That is the, in my view, the direction that dissertations and theses at the master's level need and should be taking today. We're still 20 years behind, right, where we what we could be doing. Now there are some schools blowing the lid off of that right now, but I think kind of collectively, when you can look at how people synthesize data and information in ways that are meaningful to the reader, to the consumer of research, to the public, it's going to have big implications and also demonstrate learning. So I think it's really changing the way in which we envision the dissertation, the thesis as a synthesis of knowledge that's beyond the written word. Because I can easily go to ChatGPT and have them chunk out a psychology dissertation or, you know, I'm an ed historian and sociologist. You know, complete this and it'll do it in, you know, record time.

Joe Sallustio: Yeah, something that would take me two years to put together like a day, maybe less.

Nathan Long: So I think this is where it should be because how much better will students and professors be? And you can have different conversations, right, Chirag? I think that's what you're driving towards with this new technology and getting away from the, I won't call them tired, but they're just different and they're kind of old school. And so I think what you're proposing and what we're seeing in the industry to me is very exciting what it could unleash for students as well. It could make the entire experience better, improved and enhanced across the board. That's what I would want to...

Joe Sallustio: Chirag, last word to you, my friend. What else do you want to tell us about Instructify?

Chirag Taylor: We're just really excited to work with faculty on helping integrate human-centered AI into the higher education process. And along the dissertation front, something that you know I also want to highlight is my co-founder and I both graduated from Georgia Tech, but while we were there I was part of this EdTech lab that was kind of the crucible of where all of this started. I wrote a paper on like a recommendation engine we wanted to build back in 2017 for forums and discussions on online education. And that's where a lot of this started, but we still have a very robust relationship with this advisor there named David Joyner. He has a team of researchers that are working on new and updated EdTech approaches that we can integrate into Talia after doing like random control trials. And one of the things I'm most excited about is there's a student there named Ray Hung who's working on oral exams with AI. We're very excited to see where that goes and excited to see how we can work with other universities to kind of incorporate Talia into their day-to-day.

Joe Sallustio: Well, there you have it, ladies and gentlemen. First, before we close out this episode, you might have thought Nathan Long, Dr. Nathan Long, president of Saybrook University was just randomly great on the microphone, but no, he has his own podcast that he does. It's called Unbound, Saybrook Insights, and he talks all about the happenings at Saybrook University with alumni and faculty and all sorts of people. So he has perfected the podcast... see I can't even say pod craft... podcasting craft. Try to say that three times Nathan. I won't...

Nathan Long: Yeah, please don't. Please don't try that. But anything you want to say Nathan about this episode on the way out?

Nathan Long: I would just encourage if you are an administrator, faculty member, student, embrace this technology and give Instructify a call. I'm not a pitch man. First time I heard of this was literally last week as I was prepping for the interview, but I glommed on to have a chance to interview Chirag back in November because of course AI is the thing, right? So everything you're talking about Chirag is where we need to be and where we're going. And I think if it's not Instructify, it's AI, we need to get behind this and look at how we can use that and leverage it to improve teaching and learning in our institutions. Because I think Joe, you hit this at the top of the episode, right? Like we've got a lot of opportunity in higher ed and we've got a lot of room for optimism. Yes, there needs to be change, but yes, we have room for optimism. And this is one of those areas that can drive change, innovation, and I think excitement. So.

Joe Sallustio: Ladies and gentlemen, my guest co-host, Dr. Nathan Long. If you're not, he is actually not a pitch man, but Chirag, you might want to give him a call because he says it really nicely about the future. And ladies and gentlemen, our guest today, is Chirag Taylor. He is co-founder and CEO at Instructify. Chirag, did you have a good time on the podcast today?

Chirag Taylor: This is very exciting. You guys summarized it well earlier when you said that there are going to be early adopters, but the majority of the industry just has a lot of anxiety around this for a good reason. And the best way to kind of combat it is just talk about it, explain it, show people that it's not some voodoo magic that they should be scared of. And that's going to be very hard to do without people like Dr. Long and Dr. Joe to have platforms which will allow us to do this. So very grateful to being on the show. Very grateful to have this conversation with you guys. I'm excited to share it with all of your audience.

Joe Sallustio: Well, there you have it, ladies and gentlemen. I would say this was a fun episode and boy, boy, check out Instructify. You won't be sad that you did when you check out all that they have going on. You've just EdUpped!