AI Made Practical for Teachers and Designers With Sairam Sundaresan

Think AI can do everything? We put that assumption under the microscope with AI engineering leader and author Sairam Sundaresan, and walk away with a playbook that’s practical, ethical, and built for real classrooms and design teams. We break down why narrow, well-scoped tasks are where AI shines, how to turn prompting into a repeatable workflow, and what it looks like to treat a model like a new hire you’re onboarding—clear roles, examples, and tight feedback loops.

We dig into the big wins for educators and instructional designers: personalized learning at scale, faster feedback cycles, and smarter revisions between sessions. Imagine a 24/7 teaching assistant that adapts to your students’ levels, flags weak spots, and helps you adjust the curriculum without waiting for the next term. Pair that with your human superpowers—reading the room, motivating learners, and connecting dots—and you get a learning ecosystem that’s both efficient and deeply human. Along the way, we share hands-on tips with tools like Notebook LM, Canva, Gamma, and Genially to prototype content, translate assets, and build interactive experiences without months of overhead.

Enjoyed the episode? Follow the show, leave a quick review, and share it with a colleague who’s experimenting with AI in education. Your support helps more educators discover practical, ethical ways to use these tools.

🔗 Website and Social Links:

Please visit Sairam Sundaresan’s website to subscribe to his newsletter.

Sairam Sundaresan’s Website

📢 Call-to-Action: Want to explore AI in a way that feels clear and approachable? Connect with Sairam Sundaresan and check out his book AI For the Rest of Us. You’ll find practical insights, real-world examples, and guidance on how to use AI responsibly in work, learning, and life. Visit Sairam’s website to learn more and access resources designed to help you confidently navigate the AI era. 

Join PodMatch!
Use the link to join PodMatch, a place for hosts and guests to connect.

Disclaimer: This post contains affiliate links. If you make a purchase, I may receive a commission at no extra cost to you.

Transcript

Copy

Welcome & Guest Introduction

Jackie Pelegrin 0:01

Hello, and welcome to the Designing with Love podcast. I am your host, Jackie Pelegrin, where my goal is to bring you information, tips, and tricks as an instructional designer. Hello, instructional designers and educators. Welcome to episode 94 of the Designing with Love Podcast. I'm thrilled to have Sairam Sundaresan with me today. Cyram is an AI engineering leader and the author of AI for the rest of us. With his passion for making AI accessible and practical, he brings a valuable perspective on how this technology can be understood and applied by everyone. Welcome to the show, Cyram.

Sairam Sundaresan 0:41

Thanks for having me, Jackie. It's a pleasure to be here.

Jackie Pelegrin 0:44

Yes, thank you. And I usually like to give a shout out to Podmatch when I uh when I have guests that come on that have been matched through Podmatch. So I'm glad we got connected and had a chance to have this wonderful and engaging conversation today.

Sairam Sundaresan 0:59

Yeah, thank you, Podmatch. Yeah, excited to talk about some cool topics.

Jackie Pelegrin 1:03

Great, yes. So to start, can you tell us a little bit about yourself and share what inspired you to focus on AI engineering?

Sairam Sundaresan 1:11

Sure.

Cyram’s Path from Photography to AI

Sairam Sundaresan 1:12

Um so I am an AI engineering leader. I currently lead a team of engineers who are trying to help cars um drive safely and uh park safely by themselves. Um and prior to this, I was working um in both the data center and uh mobile phone um industries, and I was working on AI-related topics there as well. So I've sort of spanned uh 15 years in AI, which makes me sound like a dinosaur considering how fast AI is moving. But uh that being said, it's given me the opportunity to see several inflection points in this amazing technology and also work with some of the brightest minds in the world on this. And uh what's got me into the field was actually my passion for photography. Um when I was a kid, I was uh taking a ton of pictures with my uh parents' camera. It was a film camera at the time, and I used to love manipulating the images um in the lab. But what I realized along the way of that journey is that I could use those images to teach computers to understand the world. So I found that there was this entire field that focused on this particular topic, and I was hooked. And since then, I've spent every moment uh trying to learn as much as I can. And that's how I've gotten into AI, and that's how I continue to learn.

Jackie Pelegrin 2:37

That's great. So you didn't expect to actually go into this field, and it it was a love of another another subject in another area of photography that led you to do this. So that's really great. Wow. I I love that when you be kind of you go into it accidentally or it falls into your lap and you don't realize, wow, this is something I really love and I want to continue doing it and uh and advocate for the for the field, right? So it sounds like that's what you're doing too, because you work with a a team of engineers and you're able to help them hone their craft and their skills and help move that this field forward. So that's that's really great. Sounds like you're advocating for the profession and for the industry as well. So that I love that.

Sairam Sundaresan 3:21

Um I I really appreciate that. And and what I like to uh tell people is that by day I'm teaching machines to understand the world, and by night I'm teaching people to understand machines. So Oh, wow.

Jackie Pelegrin 3:33

So it's a cylindric process, right? Yeah.

Sairam Sundaresan 3:36

It is a cyclical process, yeah.

Jackie Pelegrin 3:38

Wow, that's great. So um in talking about your book, as I mentioned earlier in the introduction, you've written AI for the rest of us with the goal of making this technology approachable because it can be a little bit uh scary at times, right?

Narrow vs General AI: Clearing Myths

Jackie Pelegrin 3:52

So for those who might feel overwhelmed by AI, what are some of the biggest misconceptions and how do you help people move past them?

Sairam Sundaresan 3:59

So there are a ton of misconceptions, and I'll probably try to stick to a few. Um the first thing is that AI is going to solve everything. Um, it's not. And at the moment, um the the if you think about artificial intelligence, there's like two broad classifications. There's artificial narrow intelligence uh and artificial general intelligence. The latter is what people are concerned about, and that's something where uh an AI can do um all kinds of tasks at a level that is uh you know at least human level or significantly higher. And m my honest take is we are not there yet. And I uh and in full follow-up of people might be asking, oh, when is AGI coming? Should we be worried? Um, I don't know. And the honest answer is there's a lot of smart people looking into um how AGI might come about, and I don't have a timeline for it. Now coming back to artificial narrow intelligence, most of if not all of the things that we see in modern AI is you know is going to fall into this category. It this means that these are AI um models that are good at one specific task and are insanely good at it. And uh, for example, your um chatbot is really good at giving you responses, it understands your question. Quote aircotes understands your question and responds and responds to you in a way that you know feels like uh it's getting what you're saying. Um your um smart speaker at home listens to your voice, is able to decipher what you say, and then reply to you with either an action, like you know, turn on the lights or you know, play my favorite song, something like that. So these are things that are good at one specific task. And with AI, the idea right now is if the task is very well defined, is narrow and well scoped out, chances are AI will be very good at it. If the task is very broad, generic, and not well scoped out, AI won't be as good. And the misconception I see a lot of people having is AI is going to, you know, I I'll throw AI at any problem and it will solve it. It might, but chances are unless you are very particular with the constraints, it's not going to do a good job. So that's probably the biggest misconception. And happy to pause here and maybe talk about uh other things if you have a follow-up.

Jackie Pelegrin 6:30

Yeah, wow, that's amazing. Because I think some people just have this misconception I agree. People have a misconception that they can just give it anything, give it any prompt, and it'll just give them what they want, right? And it'll give them the answer. But that's not always the case. So yeah, that's interesting. Because we have a uh where I work, we have a closed system model. It's it's our own LMM that they created because we have proprietary information that's has curriculum and policies and procedures. And I've been actually working with this model and it has chatbots within it, but I've been working for it, Cyram, for about three or four days trying to get something specific. And I I was very specific in my prompt. Like I said, you're an instructional designer. This is what I'm trying to do. And what I'm trying to do is get a curriculum map. And I'm trying to get a downloadable. I tried an Excel document before, and I tried uploading all the syllabi documents. And there's just so many in this program. I think it it I don't think the system is able to handle it right now. So I tried copying and pasting it from the syllabus in there and just putting the text in there. And it seemed to work better, but uh still after three days, I'm still not getting an output. And I keep asking it, are you how's it coming along? And it's like, oh, I'm doing this. And I'm and I'm like, okay, is it am I really gonna get an output? And so even today, before I got off work, before I logged off, I've checked with uh with it and I said, How's it coming along? Oh, I'll get it to

Prompting Limits & Workflow Friction

Jackie Pelegrin 8:00

you by the end of the day. And I'm like, still don't have it. So I'm so I've tried different tried tried different formats and I I've tried different strategies and I'm still I'm get stuck at a certain point. So it's very interesting. I'm like, well, maybe the technology is just not caught up to that yet. And maybe they, you know, they need to our tech team needs to finesse this LLM model a little bit more.

Sairam Sundaresan 8:26

So that's that's where the the interesting bit is because um I feel like with AI, especially the use cases, we are kind of writing the user manual as we go. Um so there isn't a right way to do things or a wrong way. That I mean, there are just you know, there are best practices around, but um there isn't like a a recipe that if you copy paste it from one model to another or um one interface to another, it just works out of the box. And that's kind of where spending time with these tools and seeing how they respond to smaller tasks. Uh when when you get past the initial excitement where you you give it a toy task and it does spectacularly well, you're you're you feel like, oh wow, this is going to completely change the way I work. Uh then you hit this uh wall of frustration because you try to give it a little more uh uh you know tasks that are uh you know complicated and then it doesn't give you the response you want. And then you try uh finessing your prompt, you try to go back and forth, and there's a lot of trial and error that comes into the picture. So this is where spending time with these tools and then figuring out what type of response, like how do you break up the response? And generally, in especially with chatbots, uh the I have a rule of thumb which is never take the first response. Um and the way that I think about it is if you had a new employee join your team or an intern join your team, they're joining your team without any context about the role, about the policies, about the where they can find things and how they need to do their job. And if you are asking them to do a task for you, uh their first attempt will not be anywhere close to what you expect uh or what is usually the the standard of the results, usually, right? Unless this is an exceptional employee who's uh yeah you know knocking everything out of the park. Uh so then how do you get this employee uh to ramp up and then support you? You actually give them examples, you show them how it's done, like this is how I do it, here's where you can find these documents, this is what's inside these documents, here's how you connect the dots, here's the task I want, here's the input, and this is what I expect as a result. And when you do this, you notice that the the results they produce are closer to what you expect. And the same is true for any of these AI models. Think of them like an intern or a new employee that you are onboarding onto your team. And every chat, if you treat it in that way, it becomes easier because yes, these models have this um notion of memory where you know it it it kind of knows your preferences or like based on your past interactions, it can say, Oh, Jackie preferred this type of output or the did this type of task. So I'm guessing she would um you know expect it in this format. But uh my take is always you know, think of it like you're onboarding a new colleague and be easy on them, give them as many instructions as you can, as specific instructions as you can, and then go back and forth. So that's that's when you'll get more uh useful outputs.

Jackie Pelegrin 11:41

I like that. That's great. And that's something I've been practicing too using that. And I actually just upgraded with ChatGPT, I was using the free version of it, and I was using it so much that I would, of course, uh hit my limit every day. So I was like, oh no, I'm hitting my limit, and then it would go to the lower model, right? After it hits that limit. So I decided to take the plunge and and get the $20 a month paid one. And I would notice such a huge difference between the free and that. And uh it's amazing. So there was, I think it was the other two days ago, I spent about two hours, two and a half hours on with the tool with Chat GPT, and I was refining some things and working on some materials, and I just kept going back and forth with the model and just saying, okay, yeah, you kind of got that there.

Treat AI Like an Intern

Jackie Pelegrin 12:30

But, you know, maybe we could try this or, you know, repurpose this and and just work through it. And so it really was nice because, oh yeah, you got a good point there. And so, you know, I just kind of like you said, just worked with it as if it was my assistant or, you know, or a new employee, as if it doesn't know some of that content. And so yeah, it was nice. And it kind of encourages you on along the way as long as you're kind of giving it that encouragement too. It will it will respond in that way, right? So it's it's neat to kind of have that experience and go, okay, I I know this is a machine, but this is pretty cool. This is neat. It's it's it's it's kind of a fun experience too, to kind of experiment with it and see how where you can go and what you can get from it. So that's a lot of fun. Yeah, I love that. That's great. So many of my listeners are educators and instructional designers who are curious about how AI can support their work. From your perspective, what are some practical ways AI can enhance teaching learning or the design of learning experiences?

Sairam Sundaresan 14:12

Um, this is a you know a gold mine of a topic, and there's uh a lot to dive into this. Uh so I I I'm I'm I I don't come from an educational background, although I'm a passionate educator myself. Uh so here's let's let's let's see. Um, one of the things that you immediately recognize is that with learners and students in general, no two uh students or learners are the same. They have different needs, they maybe learn better with different media, uh, they may be at different points in their learning experience and journey. And uh therefore, if you design a single curriculum that is going to cater to each of these learners, it may not land as effectively. And uh the other thing is there are as teachers, you probably have uh you know finite time interacting with these students during the course of a term or um if it's an online course for the duration of the course. And the other challenge is for the teacher, it's a lot of cognitive load because one learner might get things really quickly and they may ask you for advanced material. They they probably might ask you if uh you have what like what's next, and how do I take this to the next level? And then you have another learner who's probably um not there yet and trying to catch up. And for them, you may have to invest a ton of time and energy trying to explain how something works or how they can um understand something better. And uh if you now add AI, which is um a 24-7 private tutor that's available that can tailor the learning experiences based on the current level of a learner and based on the media that they learn best from, you suddenly have this tool that uh scales education for these teachers, right? Because all of a sudden they can ensure that each student gets the care that they need without themselves, with the without these teachers sort of stretching themselves too thin. You um get courses that you could have custom quizzes designed by these AI models based on the material that are tailored to the learner's weak spots. Because you can it's not something where you have a test and then you know have a test a few weeks later. If if you want, you could have quizzes every day. Um not saying learners should have quizzes every day, but my point is that if they have uh certain areas where they're having a bit of trouble with, you can spend more time on that. And this can be done outside of the classroom environment. So all of a sudden, learners now have access to a infinitely patient 24-7, 365-day private

Practical Uses in Teaching & Design

Sairam Sundaresan 17:04

tutor who is able to probe and ask and teach at a level that is um sort of helping them learn where they are. And for the teacher, it allows them to then design experiences around that. So they have now a TA that's available to help. And and for designers, uh all of a sudden, you can get instant feedback on like if you have slides for a particular course, you can get feedback on where the students in general didn't get the material. So maybe some slide needs to be redesigned, for example. Or you can think about um like is there a particular topic that was not covered in depth? And was there a demand? And you can have a turnaround of a week as opposed to a semester. And I feel like that's opening up entire new opportunities in education. But I'll stop teaching.

Jackie Pelegrin 17:57

Yeah. Well, that's wonderful, Saram. I I love that because I'm a I I work in curriculum design and development, uh, my daytime job. And then in the evening, I teach online classes in instructional design. So hearing that perspective is great and it reinforces what I'm already doing as an educator and as an instructional designer. So it's fabulous. And I I really hope that um those that I that work with me listen to this and also my students, because a lot of them are educators, a lot of them are teachers. So being able to tap into this technology, I think, like you said, it personalizes the learning and it allows us to make changes on the fly. So, like you said, we don't have to wait a whole semester or anything like that. Um, because sometimes when we wait too long for this, the curriculum, it's that you have you're trying to play catch up and then you feel like your students are already behind the curveball. So it's great when you can have that opportunity to make sure that that curriculum is up to date and it's uh it's applicable to what they're going to learn on the job. So that's great. Especially in fields like I would think healthcare, like nursing and um, you know, like what you do, engineering. We have Grand Canyon University has engineering programs and the science science and engineering are really big areas at the university, but also education is is really big right now and and so uh counseling in those areas. And so trying to help them to understand that AI can come alongside them and be their, like you said, their tutor or their assistant and not look at it as uh as something that uh they should be scared of. So yeah. Right, exactly. Right.

Sairam Sundaresan 19:40

So if you if you think about what teachers can do best, um they they can look at students and see how they're responding to uh material or how they're responding to certain topics in conversation, and you can see it in their body language, their eyes, their face. Um, and you can see when somebody is motivated or is getting what is being said versus somebody who is a little uh, you know, maybe they they don't get the material at first go. And that is a qualitative signal that only teachers can get. And on the other hand, the quantitative side comes from these AI tutors because they know how many times a student had to retake a quiz, exactly um which questions they tripped up in and where they need extra support. And you all of a sudden you have both sides of the coin, which allows you to design and develop a holistic learning experience. So for me, I'm just so excited about what is available for education in this day and age versus what was available um back in the day.

Jackie Pelegrin 20:42

Right, exactly. You know, one thing you mentioned, Saran, was uh cognitive load. And that's something in our in my field in education, but especially in instructional design, that we're always trying to keep at the forefront. And I always remind my students about that because it's it's based on cognitive psychology that our brain, our working memory can only handle so much information. And so if we put too much on the students, if we info dump on them, uh, you know, and the technology is too much for them, you know, if if that's a barrier to them, then they're gonna they're gonna struggle, right? And they're not gonna they're not gonna get that. So we don't want to put any barriers in front of them. We want to make sure that it's accessible for them and that they feel like they're uh they're not having to work around things or anything like that to try to get to what they need to. So yeah. So it's nice when these tools um can help them with that. And like you said, then the the teachers and the educators can focus on uh what they need to do to to help that learning move forward and to help them have

Personalization, Feedback Loops, and Speed

Jackie Pelegrin 21:44

those moments where they go, oh, I understand that now. And now I can take this into the workplace and I can apply it because the program I teach for, they they do project, it's project-based learning, which is really great. I love that because they're able to they don't there's only maybe one or two papers in the whole entire program of the courses I teach. All of the rest of the assignments are all project-based where they have to literally do this. And I have a course I'm teaching right now, you'd love this, where they they have to use an AI tool twice. So they have one one assignment that's about situational leadership and they have to and they have these prompts for these, uh, for the four readiness styles. And they have to take the prompt, put it into an AI tool and converse with that tool and get that individual. So the AI tool like Chat GPT plays the employee and then they play the instructional designer. And so they have this conversation with them, working with them as the leader and trying to get them to move from one leadership style to the to the next or one situational model to the next. So it's very interesting to do that. And then another one they did that's that I'm grading this weekend is where they had to build an e-learning module on um uh it's it was for like a fake tech company uh about their policies and procedures. And so they have to use the tool to help kind of build that content and re refine it. So um, but it's funny because some of my students um they they feel like, and I think even educators do too, but some of my students actually told me in messages and in their reflection afterwards that they felt like they were cheating. And I was like, really? That's surprising. Yeah, exactly. And it's like, hmm, you know, but yet educators, they're like, they think in the back of their mind. So many of them, I've been to webinars where they think in the back of their mind and they'll type it in, they'll be like, how do I know if my students' DQ responses are genuine or not? How do I know if they're not just utilizing AI? And I'm like, well, yeah, yeah, exactly. And I was like, but before, and I thought to myself, to and I put some responses in there, and I said, Well, what happened before AI? They found ways to cheat, you know. So if they're if students are gonna cheat, they're gonna cheat no matter what, whether AI is there or not. You know, if you remove the tool, they'll they'll find ways to do it. I mean, we I could search quiz questions for the classes that I help design, and you can find them online, or you go to Course Euro and you can find them. So absent of AI, I think they can they can still do it. Exactly. I mean, they they've had Turnitin, Turnitin's been around for a long time, and institutions have used those to check for plagiarism, right? So yeah, so before AI was around, Turnitin was around. So it's funny. Um, but yeah, you know, my institution, they don't want to police the students, they want to make sure that they're using it ethically. So that's the the most important aspect. So yeah, that's great. So that kind of goes into my next question about you know, this being such a powerful technology. There's always those ethical considerations. I think just a society as a whole and not just in education, but what are uh what do you think are some of the most important things that educators and designers, or maybe even anyone, you can so you can open this up to anything, that they should keep in mind when using AI responsibly in their classrooms or their training programs, or maybe just in general in life. Um feel free to open that up to other things too.

Sairam Sundaresan 25:13

So I I I feel like um there's a lot of um there's there's a lot of ground to cover here. And uh from an ethical perspective, the very first thing that comes to mind is like the material itself. Make sure that um if you know respect copyright and ownership of the material, the lineage of the material, so that way um don't um like use material from you know, without consent on training these models to create better educational experiences. Because at the end of the day, um this is somebody else's work that is, you know, that they've probably put months or years into to develop, and uh you need consent or appropriate compensation before you consider using it to train or you know, build an AI learning experience. So that that's one that comes off the top of my head. The second one is is more related to sort of um from like a learner's side, I feel the ethical side we you just touched on it, which is don't coast or you know, mail it in um by just using AI

Ethics: Copyright, Hallucinations, Energy

Sairam Sundaresan 26:24

to respond to everything, because it will respond in a very plausible way. But but here's the kicker. Um chances are it's hallucinating part of the response. Like um, there's been cases when the AI model has hallucinated textbooks that don't exist um as part of a citation or uh research papers that don't exist, for example. So if you are just you know throwing um the every question into an AI model and expecting an answer, in the long run that's actually going to hurt you. And and don't do that to uh you know teachers or educators who've spent uh hours trying to figure out how best to teach um and make sure that you know you you understand the material. That's the second one I can think of. The third one is um in terms of usage itself, because there is like AI consumes a ton of energy, and so using it responsibly is is sort of important for us to ensure that we aren't uh um you know from the from the perspective of electricity and carbon emissions and all of those things, it's it's it's crucial that we use it. Um you know, don't use it just for dad jokes. Um use it uh responsibly for something where you need that extra assistance. So uh the the the fear that I always have is uh like there might be an you know like if you have a 24-7 TA next to you to help you with everything, what happens when that TA is suddenly not available? So don't make it a crutch and you know, don't be overly dependent on it. Use your uh thinking first and then see how it lines up with what um the model has to say. And the other ethical thing um I would just generally say is um this this sort of applies to um things beyond education like interviewing and all of those things. I see that there's a sudden rise in people um either impersonating um others like by using AI to change their appearance or by um using AI to answer the questions that are asked in an interview. And those are just areas that you know it it leaves a very bad taste in the mouth for for the companies doing the interviews. So that's just uh a few that that came to mind.

Jackie Pelegrin 28:46

Wow. Oh my goodness, that is scary. I didn't I didn't think that was that that's something I haven't heard of before. So you brought wow, you brought that to my attention. So they're impersonating other people in the interview and answering them. Wow. Oh my goodness.

Sairam Sundaresan 29:00

So they're basically the creating fake personas. Um so the resume says they have all of these skills and everything, and then when you look at their uh when they come on a video interview, they use AI to change their appearance so it looks like some person and uh that way, you know, they're yeah, it's it's it's just it it's it's uh a very murky area that that has been um coming into the news of late, and I just find it, you know, extremely distasteful.

Jackie Pelegrin 29:31

Wow, that is, and it's uh yeah, it's so unethical because um they're not being genuine about who they are, and so that's really sad and what they know. Wow. That's really sad. So just like with any technology, we have to be careful and have those guardrails in place. Wow. And it's I'm glad that you brought up the part about um being mindful of how much we use it because of the the energy it consumes, right? The carbon emissions and everything. Because here in the United States, you know, we have data centers that they're building and they're they're huge data centers. I think there's one in Texas they're gonna build, and there's another one, I forget where the other one, there's uh some other one that they're gonna build, I think somewhere south, but I can't remember where. And uh so the administration's trying to do that, get ahead of it, because they as they say, we know China is trying to win the AI race. So it's like, oh, how do we get ahead of that? But yeah, it's like, well, we still have to be ethical about how we go about it. So we don't want to get ahead of it, but then and try to beat China, but then you know, hop, skip, and jump over things and do all of that. So there's always that ethical component. Anytime there's technology in place.

Sairam Sundaresan 30:45

There's always two sides to the coin, right? So um just being cognizant of that and being

Misuse, Impersonation, and Guardrails

Sairam Sundaresan 30:49

responsible, I think that's that's crucial.

Jackie Pelegrin 30:52

Absolutely. Yeah. Love that. Thank you so much for that. That's that's amazing. Those ethical considerations are always something that we have to keep at the forefront of our minds. Absolutely. So uh before we move into the final question, I'd love to slip one more in as a bonus question that I know will resonate with my listeners. So looking ahead, what are a couple of things that excite you the most about the future of AI and what should educators and instructional designers be preparing for now to stay ahead?

Sairam Sundaresan 31:20

So in terms of what excites me, it's just the number of things that we considered impossible just a couple of years ago are now suddenly within grasp. And um a lot of the um ideas that seem like fiction are some suddenly like possible if not plausible, right? It's uh it's it's important for us to the thing I'm most excited about is just being um empowered to do and try different things and actually see the outcome of those things without a long turnaround time. But because previously um for anything in tech if you needed to build something and see how it works, it would take like weeks or months and you'd need a large team and it would you know it's like you you couldn't try out a whole bunch of ideas all at once and see how they turned out because it would just be infeasible. Now all of a sudden with AI assistance you can quickly prototype ideas. And therefore you can um either you know veto a whole bunch that that aren't um you know go good to go forward yet or and you can find the the ones that have a lot of potential very very quickly and with a fraction of the time and cost that it it would take uh otherwise. The other thing I think I already spoke about is like education. I'm super excited about how AI is going to change the field. And um for me just being able to even even for me, right? I mean like there's this tool called Notebook LM that I uh use all the time and it's completely changed the way that I learn new topics or concepts because AI is evolving so fast and I try to keep up with a whole bunch of research papers that come out. And all of a sudden notebook LM allows me to do that more efficiently and I can create mind maps, I can create uh quizzes, I can even have like if I'm short on time, it creates a podcast for me to listen to based on the material which is absolutely mind blowing. And um just for me to have all of these tools at uh you know at my fingertips is is is a blessing and just uh I'm excited to see where this goes from here in terms of um education and for me to help uh teach others. So those are the two things I'm excited about. Now coming to what learners need to be um and instructional designers need to be preparing for I would just say the easiest way to start is spend time with these tools. Just pick a tool that you want to use it can be anything your chat GPT, Claude, something else if you're in design maybe a tool like Canva or gamma and just just play around with the AI component of those tools and just see how to use it, interact with

Future of AI: Fast Prototyping & Tools

Sairam Sundaresan 34:04

it. Like I said there is no user manual we're kind of writing it as we go. So just as as you spend more and more time with it you figure out the patterns that um allow you to get to the desired results faster. And they also allow you uh you know a safe space to experiment, try new ideas out and then see in and maybe unlock new use cases that were previously uh something you hadn't considered or thought about. So from that perspective it's um just display of the tools and then uh the other aspect is maybe have a basic sense of AI literacy um in just understand how these models work behind the scenes. I'm not saying you need to uh know how to code these up or do any of those fancy things. If you can fantastic but um just understand how they work and how they're built because that gives you uh context into their strengths and limitations and then that in turn allows you to work more effectively with them. You know where they're not good, where you need to step in and conversely where they are good so that you can step back and then do something else that you are good at. So um that's what I suggested.

Jackie Pelegrin 35:12

Those are great tips. I love that that's great. And I I'd love that you mentioned for education Canva and Gamma because I use those tools quite a bit myself and I even did it for I used it for work recently with Gamma and I took some of my old PowerPoint slides because I've been teaching OneNote for for quite a while to help my coworkers to get more organized and kind of take maybe some of their ideas and just kind of organize them a little bit better because we have OneNote as part of Office 365. So I had them you know PowerPoint's great and PowerPoint serves great purposes but I thought can I kind of refresh these and give them a little bit of a boost and and everything like that. So I went to Gamma because one of my students had her presentation in Gamma and I was like oh what's this? Okay, I'm learning all kinds of tools here as an instructor it's amazing. And then my students are like hey yeah go look go use 11 labs and go and I'm like oh my goodness it's amazing how much I've learned in five years and then just how the tools just keep getting better. But yeah I was able to take that those presentations and really uh take them to the next level. So it was fun and it didn't take a lot of time because I already had the content. So I was just putting it in a different format and and kind of just reimagining it and what it could look like. So yeah it was it's fantastic. Yeah I love that. And Canva, oh my goodness, I love Canva. Wow they the Canva code, I've used that a few times and it's really neat to actually see that's it's coding and it's creating it for you. And then like you said, just like with AI, you have to kind of tweak it a little bit work with it. I've I've had to do that sometimes where I asked it to do a matching game and it didn't turn out that great the first time. It was okay but not all not everything worked and I was like okay well can you do that and then I asked it to actually do like a a little uh celebration thing at the end where it gave it and actually works now really well. But at the end it'll when you match everything it'll this little trophy will come up and it'll do the little confetti at the end. So it's really neat. Nice so I asked it can you please do this and then it I saw it as it was coding. It was really fun to actually see it in action and then just work with it a little bit and say okay you know not quite right can you can you do this? And so yeah it was a lot of fun. Love that yeah there's another tool out there have you heard of genially before it's spelled G-E-N-I-A-L-O-I. It's an interactive tool where you can do like interactive infographs and quizzes and escape rooms and stuff like that. I need to take this out yeah you'll have to check out Genially one of my students introduced it to me a few years ago and now they've got AI

Hands‑On Tips: Play, Iterate, Build Literacy

Jackie Pelegrin 37:54

technology in there where you can actually take your interactive piece and you can if you have you have to I have the EDU pro version which is $60 a year in uh here in the United States I don't know how much it would be where you are but it's really nice because you can get premium templates and you get some of the AI technology but if you go up I think I can't remember if it's elite I can't remember which one it is but you can actually have it translate your genially into different languages. So for example if I was teaching a Spanish class I could you know have it in Spanish too or if I have learners from all over I can I can translate it to other languages. So I was like wow that's pretty neat. And so it uses that AI technology to translate it.

Sairam Sundaresan 38:38

But if I did that I would want someone that knows that language really well to double check my the work and not just assume that it translated it correctly but I love that you said that I was just going to say that's that's uh and and that applies to any educational experience right I mean all of a sudden you you you as a teacher are no longer limited by language you can teach uh students worldwide who come from different cultures but again trust but verify um so have somebody absolutely I love that great so as we wrap up based on your journey as an AI leader and author what's one piece of encouragement or advice you would give to listeners who are looking to enhance their skills in AI technology? I would say that pick a topic that you are really excited about or pick a problem that is bothering you a lot like scratch your own edge and if it's a a problem that is uh you know something that you do on day try to solve on a daily basis just write it out and I mean write out the steps basically and then try to explore how AI can replace each step one by one. It may be able to replace a step it may not but just see how you can bring in AI into that workflow. And chances are a lot of that might be automated away and as you do this you then start learning to use AI for different types of subproblems and then over time those become different types of problems. So you're building the muscles that are necessary to use AI and think in a way where you try to solve a problem first by yourself. You use your own critical thinking and and wherever you feel that there are inefficiencies or where you need that extra support you figure out okay here's where AI can step in and this allows me time to do something else or this allows me to completely automate away this this problem so I don't need to think about it. That's how I would approach it. And then the second one is like just um trying to keep up with what's going on I know that there's a lot of hype around the area there's like every day there are announcements on this tools come out and it's going to completely change the world and all of that. But uh try to keep a pulse on what's going on. Generally if a particular topic or tool is in the news for more than four to six weeks chances are there's something there. If it's there for only one or two days then I would exercise caution um so those are the two suggestions in addition to what I said about AI literacy.

Jackie Pelegrin 41:09

Right. Great I love that keeping keeping an eye on those things and not taking it for face value right and really keeping that critical thinking lens and I that's what some of the instructors I've talked to over time the last few months that they said they they're fearful that their students are going to lose that critical thinking. And I said well then our job as educators is to ensure that they don't and create authentic assignments assignments where they don't cheat right automatically. They don't want to do it because they want to actually do that assignment. And uh so we're trying to

Assessment Rethinks: Open‑Book with AI

Jackie Pelegrin 41:44

as an institution you know that we're trying to get away from so many papers and not have a lot of papers. I mean they still need to know how to write but we always try to think of what do they need to know in the field to be able to get into whatever particular industry they're trying to go into. So I think that's so important to have that um in place and so that they know that they can do that. So yeah that's that's great. I love that.

Sairam Sundaresan 42:09

Lots of great we've we've had open book exams um in the past and one of the things that affords the the the teacher is that you can set really um interesting questions that you don't have a direct answer to. And you'd have to do a bit of work you'd have to look at different sources and then come up with an answer that is sort of uh helping you assimilate the material. Now with AI that open book concept becomes you know it it goes to a completely different level because you instantly get the answer but again you need to fact check there's all those kinds of things. But I wonder if there is a way to create an experience that simulates open book but you know using AI in a way where you are practically testing things as opposed to like you said just having people done in reports and papers because those are the kinds of assignments where I feel students might learn the best because they are forced to use AI and at the same time they're forced to use their own thinking.

Jackie Pelegrin 43:12

Right. I love that yeah kind of like the assignments that I I'm braiding and it it forces them like you said it's got doesn't just have one part to it has multiple layers to it. So you you can see did the students really understand it and I can tell you know they're not just putting that input in there and just getting the output and doing a screenshot, but they're actually learning from it. And and then the reflection that they did afterwards when they did that first assignment was, you know, uh what went well what would you do differently next time? How would you converse with the uh with those different individuals you know in the AI tool next time to make it better and what did you learn through that experience? So that's something AI can't repu replicate is uh is those experiences. So it's really great to see that. Yeah I love that. Yeah just keep growing and keep learning right that's the that's the key.

Sairam Sundaresan 44:05

Great that's the key.

Jackie Pelegrin 44:06

Yes thank you so much Siram for sharing your insights and experiences with us today here on the podcast. The wisdom that you've offered for making AI more accessible to your guidance and applying it responsibly will no doubt inspire and encourage my listeners in their own paths as they continue to explore this tool these tools and learn to integrate them into their lives. I love it.

Sairam Sundaresan 44:28

Thank you so much it was a pleasure to learn uh from you as well I think I learned a ton about how um the thought that that goes behind building courses and learning experiences. So this was super interesting.

Jackie Pelegrin 44:42

Great yes I look forward to having you back on the show because once I have someone on my show once then we always think about there's something else we could talk about. Like you said earlier we could dig in and we could really uh you know hone in on some other areas too so if there's anything you think of along the way and you want to come back on the show feel free to do that. I'd love to have you back on and we can dig a little deeper and um and maybe even see where things take us. So I'd love that. Definitely great well thanks again Sarah I appreciate it and look forward to having you back.

Sairam Sundaresan 45:15

Thank you thank you Becky this is fun.

Jackie Pelegrin 45:18

Yes absolutely thank you for taking some time to listen to this podcast episode today. Your support means the world to me. If you'd like to help keep the podcast going you can share it with a friend or colleague leave a heartfelt review or offer a monetary contribution. Every act of support big or small makes a difference and I'm truly thankful for you.