Episode 4 - Answers to Better Questions:
Designing Learning for Mastery

Join host Aaron Burnett as he sits down with Edan Shahar, serial founder and CEO of Wild Zebra, to talk about lifelong learning, building a one-to-one AI tutor at scale, designing for mastery (not memorization), empowering teachers—not replacing them—and how EO’s experience-share model sharpened his leadership through big life shifts and company pivots.

This podcast features real entrepreneurs sharing real challenges and solutions. No pitches, no sales - just honest conversations about the moments that shape successful businesses.

Aaron Burnett: What’s your entrepreneurial story?

Edan Shahar: Okay. So I have been building companies since my freshman year of college. I've been through quite a few of them at different iterations. Most of them have been consumer facing technology internet companies. I did spend a couple of years in quantum computing Today, what I'm working on is a company called Wild Zebra. It consumes all of my interests and passions and I spent a lot of time working in it. And what we are looking to do is follow the trajectory of my career building tools that help in education. I've been working on trying to better understand students to help them in more personalized ways for a long time.

And today the manifestation of that is Wild Zebra, where we're bringing together. AI and we just launched recently. We're now working with 6,000 students who've been on the platform for initial pilot schools, and we're expanding all the time. And the core mission, and kind of what I'm most interested in in life in general is the process of learning.

I think that. That also delves back to, you know, why EO is actually great for me. I'm very interested in learning and in metacognition and sort of how a human learns and how we can engage people in their learning. How we can give ownership of the learning process to people. So specifically, you know, most of education, when we think, when someone says education, you're kind of thinking K 12.

Of course it happens throughout your life, and EO is a great example of that. But for us right now. At Wild Zebra, what I go to sleep thinking about and wake up thinking about is how can we create more engaging platforms that understand students very, very deeply? Both what are the things that they need to work on and how they work on them, and also what are their students' interests and can be sort of that encouraging aspects? 

It's multidimensional when you think about learning, keeping people engaged, connecting a student, like where they want to go and why they want to go there to the activity. That they're doing now and then having a, a place where you can explore yourself. I think that the best kind of learning happens when there's a, in our platform, it's an AI tutor that has a Socratic approach, but it's, it's, it's a place where you can sort of ex explore the map yourself, right?

That's where the best education happens. So that's my career journey to date. Kind of glossed over some things, but that's, that's where I am. So this is not your first learning company? This is not my first learning company. Yeah.

Aaron Burnett: How did you come to this focus on education and learning? Why is this so important to you?

Edan Shahar: I think at the, maybe the most superficial level part of the journey for me was that I am, or originally from Israel, like your previous guest. Coming from a different system, I didn't know as much about, you know, I needed to learn about it.

My parents were not, you know, involved in the same way that maybe the average American parent might be involved. So part of the journey was my own journey through school. I came to learn about this, , when I was applying to colleges and kind of that whole process. I think at a deeper level, I am just somebody that loves learning.

I concluded undergrad with 249 credits, and the context there is that at UDub you needed 180 to graduate, and at 250, they just tell you to, to just get out. And, and so what I did is I, and I had two major, like two majors and three minors, and I, and I kind of staggered it. So like, it was only in the very last quarter that I was, I concluded all of the, the classes and I did graduate in four years.

I just really liked stuff. So I took summer school and took, you know. 20 credits every single quarter and yeah. Came in with a bunch. So, I think there's, there's that level just about myself, I'm interested in education 'cause I think that is the journey of life is a journey of education, of understanding the world, understanding oneself.

And I think that is what makes us human, you know? What differentiates us from, from other beings is our ability to learn new things. I think another aspect of my fascination with learning has to do with my heritage. You know, I mentioned I'm from Israel, I'm Jewish. I think that it is deeply steeped into the culture, this concept of learning.

It's been, you know, around for thousands of years. With like, that was, the objective was to always learn and, and dive deeper. You know, it's really understanding the nuances of the Bible and arguing with other people that came before that and argued about it. It's sort of this ever evolving journey there as well.

So I think there's also that element that comes into play. And then I think now I also have two young kids. So now there's also a new element that is really interesting where it's not about my own learning, but about seeing learning happen in my kids, which is some of the most exciting aspects, like the excitement they have when they are learning things.

So yeah. That's my view of the world and the meaning of life. I'm just very interested in how we can bring the best technology to  achieve this goal and to understand the nuances of education and learning, because it's not just about consuming knowledge.

It's not just about the ability to regurgitate knowledge, it's about the ability to internalize a different worldview and have a richer, fuller view of oneself and the world that we live in when you actually are able to learn things.

Aaron Burnett: I once heard someone say, and I can't remember who (and I'll badly paraphrase) that the difference in traditions is that the Jewish tradition is that the Bible is the questions you should ask and the Christian tradition is, it's the answers.

And I think that's an interesting juxtaposition and I like the notion that it's the questions 'cause the questions make you open to exploring and finding your own answers and thinking deeply. Whereas if it's just the answers, you don't have to think as deeply.

Edan Shahar: I'm kind of jumping off of that. One of the things that I think about is.

Like in my team meetings and the companies that I've run, I like having debates. You know, like I preface many conversations as like, let's debate this. I really like having my mind changed on things. I feel like that's, for me, the most enjoyable experience. There's something very viscerally amazing when I can learn something new. I have a new perspective and I see the world. I see the universe in a different way. Like, I don't know. That's just how I'm wired up. It's a gift when somebody does that for you. Yeah. Yeah. It's really cool. 

Aaron Burnett: So what problem are you solving? What gap are you filling in the market?

Edan Shahar: The gap that we are filling and the problem we're solving is we are trying to find a new way. To teach students and initially we're, we're targeting math and English. And we are trying to find this new way to teach them that meets them where they are and focuses on mastery. So we think of mastery as again, that idea where you have actually learned this thing.

So you, you know, you have learned knowledge and you're able to apply it and you understand why you're applying it. And you do this in a highly differentiated way. So one of the things that's happened over, you know, the last many years actually, you go all the way back to Alexander the Great and Aristotle and you know, you, you go on through, through the years where there's been private tutors, right that have existed. So one of the things that we're really interested in is can we create a private tutor for each and every student. And when we think deeply about what that means and why. There's this like famous study called, Bloom's Two Sigma Problem from the eighties.

It was a researcher that did research on the most effective way to teach students things. And, and of course you found that when you have a private one-on-one tutor, you can get a two standard deviation improvement. And of course it makes sense because when you're, when you're just talking with somebody and somebody's teaching you something.

If you get stuck, you don't move on until you understand the nuances and they might take different approaches and you just stay on there. So that's on. When you don't know something and when you do know something and you're passionate and interested in it, then they dive deeper, you know? And the idea is, well, right now just this, you know, the ratios just aren't there.

Like if one teacher is doing all of the education delivery, that's not going to be as effective as if you can sort of superpower this teacher with. An army of, you know, learning specialists that's sitting with each and every student. So that's, that's the problem that we're, we're trying to solve is can we get a two sigma improvement for each and every student in the world?

The ultimate vision that we have is that tutor and that tool and that agent that works on your behalf, that's teaching you things throughout your education. Maybe they start in middle school and then they go to high school, and then they follow you into college and maybe even your career. So for example, you know this, this teacher understands how you like to learn best.

This teacher understands, or this tutor, this AI agent understands how to motivate you, how to keep things interesting, and also what information it is that you're trying to learn. You could imagine that when somebody in five years joins your company, this agent will ingest all of the content that's relevant for your company, all the content that's relevant sort of for your industry, and will go through a teaching process so that this new employee will actually know all the things.

And that's kind of the vision that we have for the future is that the ideal AI tool will be with you in your whole lifelong learning process.

Aaron Burnett: You also described something to me when we were chatting the other day that some of the initial exploration that's done in this context for a user is to find out what their interests are. If they're super niche interests and to contextualize learning across those interests. Can you describe that?

Edan Shahar: Yeah, so there's a, there's a student that recently gave us feedback where one of the things you can do in the tool is you tell us the things that you're interested in.

So, you know, it could be Taylor Swift, or it could be, you know, the operations of a 737 jet engine., and the idea is if you look at the world through deep fascination, which was one of the things that I do is like, you can, you can, everything that you'd ever wanted to learn about mathematics, for example, you could probably learn in just a jet engine.

Now there's a student (actually, I didn't go into, into the student's account to see exactly what their interests were) but their feedback to us was, this is incredible. I have this very specific interest that no one else understands, but through it I'm learning all of mathematics. And it was just like, I mean, that's one of the dreams here is like, no longer should trains be leaving Chicago and going to Boston and, and that's figuring out they're why no one unless.

Your passion is the railway network of the United States? Now you can have the context of whatever it is that you're interested in and learn all the different things, you know, all the different skills. So that's on the skill side. You need to learn history and maybe you can learn about modern history through the perspective of modern air travel, for example. Why not?

Aaron Burnett: You describe kind of an agentic experience where there's a one-to-one relationship between an AI agent and a user. Is that what you're doing now or is that a vision of a future state?

Edan Shahar: Oh, that's what we're doing now. That's what thousands of students are doing today.

If I'm  a user, the tutor is my tutor. The tutor is your tutor. The information bubbles up and of course the parents can review the conversations or the teachers, but yes, the interaction happens there. One of the other things is that the concern that this teacher brought up is like, you know, she'll, she'll talk to other teachers and some of them are concerned that they're going to get replaced by this tool.

I have no concern about that. I think that actually the jobs of teachers become significantly more important and they focus a lot more on. Facilitating. Group projects, it's about, you know, collaboration. It's about these other things that are way more important. Mm-hmm. As you think about, you know, sort of where the world is going.

And we can have and teachers can really focus in on that. And then they also have the ability to, to dive in and help specific students. When the AI has said, Hey, you know, the student seems to be stuck in this place, maybe you can take a different approach and, and look at things. So that's very, very core to the system, is we want to empower teachers to be more successful.

Right. But we also do have a product that is directly B2C.

Aaron Burnett: Common concern with AI is inaccuracy or hallucination. How do you protect against inaccuracy or hallucination?

Edan Shahar: There's been some techniques that we found around when we pre generate questions as students are going through it and we give enough context to the, to the ai, we fill it in.

And so it's been very, very good and we it's getting better all the time. It is still a concern that of course we review. But again, one of the things that we found is also as you're going through conversations, if you can do sort of these spot checks, that's great. One of the things that actually happens with a lot of teachers is they have hallucinations or any educator, you know, it's like you'll be up on the board, right?

You'll make a mistake, and then you go back and fix it. Now, actually, that process, when the, when the. Person does. That is an incredibly invaluable and important process for the learning of the student, right? And so that's one of the things that we're also working on is if there are mistakes, how can we catch them and have the AI demonstrate this corrective behavior?

Like, Hey, actually a couple steps for behind, I made a mistake. Maybe we can actually have you, can you point out to where I made this mistake here? So this is a long term view, is that the, we can make the hallucinations in mistakes. A feature of the process rather than a bug of it, because it's a conversation.

It's like the experience is not, give me the answer to this question, and then it gives you the answer. And maybe there's a mistake in there, right? Like. How many students are using AI now to just cheat, you know, basically get the answer. This is, you actually have a full conversation with the AI and in that process the AI is asking, you know, what answer did you get to and how did you get there?

And can you explain and explain your reasoning? And so sometimes the AI will make that mistake, but sometimes it's able to catch itself. And correct itself and demonstrate that you know it, just like you, the student sometimes makes mistakes and you can correct yourself. I assume you're using a proprietary LLM?

Well, we're using, a whole bunch of LLMs actually.

Aaron Burnett: Okay.

Edan Shahar: Yeah. So what we found is that there's different sets of the process, right. That are optimized for, that are just better for different things. Yeah. So like right now we use Gemini. Mm-hmm. Open AI models, Maverick sonnet. Which is anthropic. So we were kind of, what we found is it's a whole orchestration system where there's certain parts of like question creation and verification of.

Verification steps that are better for some models, other models that are more conversational., there's latency is a really Yeah. Important thing. So, you know, that was a lot of initial feedback. It was like nailing that aspect., then there's these processes that run post conversation where we sort of do an analysis and you can imagine their latency might be less important.

Yeah., and so we're, we're. We're constantly testing out new ones. I mean, that's the other thing is like we are incredibly blessed by the AI gods, the big tech companies that are constantly releasing these new models., and that's actually part of what we do as a company, is we're constantly trying to figure out, okay, can we use this new model to solve this specific part of the process?

Does it make it better against these objectives that we have? It's a, it's a process that I think will never stop. Right. So long as there's new better models. Yeah. Are you using a publicly available workflow platform? A lang chain, I think is the, is the one that we're using. Yeah. But a lot, a lot of the internals we've needed to build up.

Oh, sure. Yeah,

Aaron Burnett: Yeah, yeah, yeah. Yeah. Fascinating.

Edan Shahar: And I think it's also worth mentioning that part of this company is we've also taken a lot of  anonymized data and a lot of context to create this. So Yeah. From, from other, from partners of ours. You know, millions and millions of, of anonymized student data reports, right?

Tens of thousands of questions and sort of scoring reports on all the questions and, and you know, how are students answering them. And so all of that goes in and actually makes it very rich. The ais are. If you're trying to, you know, go down specific paths, you wanna give them a lot of context. So we actually, there's a lot of work that we do in the, the pre-work, which questions are students gonna see?

How are they going to see it? What context do we give the models so that they can create the right kind of experience. Yeah. And then also how do we understand our students to make it a really rich environment.

Aaron Burnett: The other thing that people worry about, where AI is concerned, aside from everyone's job speaking.

Edan Shahar: Mm-hmm.

Aaron Burnett: It's privacy. So how do you protect. Student privacy, parent privacy.

Edan Shahar: So right now, I mean, all of the models that we have, you know, we, we don't send any information about the student and as much as we can, all of the models that have information on the student, we keep that data. So that's, that's on, on servers that we have access to.

And then the other thing is we also have a view that at any point. All the information should be erased, should be able to be erased so folks can ask us and we will erase it, and we think that's very, very important results so far. You said 6,000 students using the platform so far? Yeah, yeah. So far. I mean, we're just getting started, but the results have been incredible.

I mean, it's mostly based on survey data, students logging in at 10:00 PM at night. To do work on a weekend, you're like, okay, that seems, that seems pretty crazy. And then teacher, teacher feedback as well. So, you know, the schools that we're in, they wanna renew. And I mean, one of the things that's also interesting is as we think about the future is like, how do we, the business model, how do we sort of find the right business model?

Because right now our view is all of the AI tools, they're not quite good enough. Meaning we are like, we're pushing the envelope on lots and lots of calls on the most expensive models. And so the inference cost is actually very high. So how do we sort of structure it by? Like, we want for it to be really, really good and we're willing to sacrifice cost in order to ensure that the experience is amazing.

But that creates some challenges potentially down the road with, as we think about scale. We'll see. Well, you know, I, I've got high hopes that the models will come down and we'll find optimization opportunities. Yeah, I think that's a good bet. We'll see. I mean, we just gotta stay, stay alive long enough for, for the curve to come down.

We will see what happens there. So let's shift gears and talk about eo. Sure. What brought you to Al, the previous guest that you had Shahar, he was, he was my accountant and he was like, you know, you should probably join EO. I joined in 2019. I joined in February. In March of that year, I, we discovered that my wife was pregnant.

In May of that year, I did my silent 10-day meditation retreat. I just, I go back, you know, and then nine months later was the pandemic. So the world has changed a lot, but it was like right at this sort of, this sort of turning point in my life. Worked out great, I guess, is what I'd say.

Aaron Burnett: What has EO done for you?

Edan Shahar: One of the things that we do in the, the forum meetings is we have a, you know, a business. Personal and family reflection. It's given me a lot of value in business. You know, we, my forum, we've got a WhatsApp group. We're constantly talking about like a how do you structure. Deals like this, how do you find employees in this category?

How do you approach problems like this? And so having a work group where other people are going through very similar things is very, very interesting. And how they've approached problems has been incredibly valuable. There's always been about half the forum are folks in my similar. Age, life stage 35 to 45 with one or two kids under the age of five, maybe three kids.

You know, like, so sort of how, finding that balance of keeping a marriage going during, while you're a CEO and you've got, you know, young kids. Like hearing about how people are doing, you know, how people are achieving all those things has been incredibly helpful for me. The whole process of the experience share rather than advice giving is like.

Fucking amazing. I mean, right. Like that's, it's, it is structurally just leads to an incredible opportunity.

Aaron Burnett: Why do you think that is? Why is it so valuable to do it that way? I ask 'cause it took me a while to sort of click in and see, oh, okay, I get why this is profound. So I think for

Edan Shahar: Most people, it's because they don't feel judged.

It's a very take it or leave it experience. Like, this is what I did in this situation. You know, like I'm trying to figure out how to do comp structure for my I don't know, VP of sales or something and somebody else talks about it and like, okay, that doesn't quite work for me. No big deal, versus, “You should do it this way.” 

I think the other thing is that, when people give advice, they have an extrapolation of how they think the world will is, but when they give an experience share and they actually just talked about what happened, you know, it could be that like my add, like I give all the facts of what happened and my adding up is that it happened 'cause of this thing.

But you hear it, you're like this guy, like he would give me this horrible advice or whatever, right? But I'm just gonna do the fact pattern here and you know, you can sort of follow it. As it went, and extract whatever, value you want from it. Yeah. So I think there's something about that, that just makes it amazing.

Aaron Burnett: Yeah, I agree. There's something about it that neutralizes ego. And the experience share - you're right, it leaves, you know, it leaves it on me, the listener, to take what I will from it. And what I find curious and interesting about it is I often don't know what I've taken from experience shares while I'm in the room, don't know until a day or two or three or maybe a week later.

Edan Shahar: Yeah.

Aaron Burnett: When it sort of settles in.

Edan Shahar: Yeah, yeah, yeah. Definitely. Or, or even months or years later. I mean, that's one of the things is Jim Neal, was the like founder or seed moderator of our forum, and he would say things and like a year later I would still be thinking about like, I have this experience in this way. Like, it's just amazing. Right. When you, when you hear stories rather than get advice, they're just like, it's not personal, it's just can just learn from somebody else. Yeah. 

Aaron Burnett: How have you changed through EO?

Edan Shahar: Yeah. I've known you since before EO. It's hard for me to unravel the complexity of life.

Like what's happened since I joined EO? Well, you know, I've started, I started doing quantum computing. I started a new company. The previous company I had is, you know, Five X, that's from the business side and the, you know, on the family side, we went through COVID. I have had two kids. And then on the personal side, I don't know what actually, what personal means anymore at this stage, but that's a different conversation.

But, you know, I've gone through that journey and through COVID, so it's hard to say exactly what EO has changed. But I am certainly a changed man since 2019. One of the things that is most interesting about it is the layers and depths of it and how things can be true at different levels for different reasons.I just think that's really fascinating. And what I love doing is like, you know, digging in and finding the next level of truth. And I kind of give you my run from, you know, from the hip about education and learning. That whole thing is just like. I like thinking about that at, in lots of different ways, about different things.

Yeah. I am discovering, you know, one of the other things that's fun about Wild Zebra is like I get to discover new things about the world. Like, for example, brain rot, which sounds horrible, which is a very silly thing. It's like the memes that are coming out of TikTok, right? And again, it's one of these things.

So students can choose any interests they want. I mentioned Taylor Swift, I mentioned a jet engine. You know, any sports team, whatever else it is. But we also get these interesting memes or silly memes that students are are bringing in. And it's interesting, right? Like in this world that we live in today.

Your phone is just an incredible, you know, dopamine engine with like every possible thing you could imagine that might excite or interest your lizard brain all the way to your prefrontal cortex. We have to find ways to engage students. Like there is this, this battle that's going on, and so the question that I have is like, can we pull elements of that and bring it in and sort of mesh the things that you are wanting to learn.

And needing to learn with the things that are going to excite your dopamine receptors. And so, you know, can we have a world in which the amount of dopamine released from a. TikTok experience. Mm-hmm. Can be pointed at something productive that the student even chooses. You know, a lot of students and a lot of people self-report, they're like, yes.

Even though they're doom scrolling or doing whatever it is, they actually don't like that experience themselves. You know, at the end of it, they didn't get anything out of it. Right. It was kind of felt empty. Is there a world, and especially now with AI, where we have so much control over media creation, that you can begin to target that at an educational process for kids.

I don't think we're there yet, but I think we have some elements of that, and that's what we're seeing in our platform when we're seeing these really high levels of engagement. Like students are really into it. You, when you enter a classroom that's using Wild Zebra, you know, the process is really interesting.

You know, at first students are, they're logging in and then you just like hear these like little bursts of excitement. Like, oh, they got, like, I got this thing, and you, and you get these little bursts of like, oh, this is a thing that I'm interested in and now within this context I'm getting to learn the things that I need to learn.

And so it's really, it's like a, it's a very fun experience. Yeah. Teaching my, my parents how to use AI, and then teaching these students how to use AI. It's like the speed at which they learn things. Oh yeah. It’s crazy, right? Like, so first, like the first seven minutes, sometimes they're raising their hands and usually the instruction is, okay, whatever you're about to ask me, whatever you just ask me.

Just write that into the AI. And again, like, I don't know what I'm supposed to do. Okay, great. Just say that. It'll help you out., and then, yeah. Within, you know, 10, 15 minutes, boom. They're all, they're all flying, right? And it's just. It's crazy. It's a wild experience. So, alright,

Aaron Burnett: So here's a philosophical question.

Okay. Should we, you're right that mobile phones have created, I think it's not hyperbolic to say an addiction to dopamine. Yeah. Should we be leaning into that addiction? Should we be hoping to moderate dopamine release? Because one of the byproducts of constant release of dopamine yeah. Is exhaustion.

Yeah. Psychological, mental exhaustion, but also emotional exhaustion. People who are really tuned into constant dopamine release also report much lower levels of satisfaction, happiness. They tend to have very muted experiences as it pertains to joy, so quality of life. Outside of the device is diminished by constant dopamine release.

Yes. So better to lean into it or better to find a way to, moderate to rein it in.

Edan Shahar: Well, I think, you know, dopamine is the, chemical in your brain that would be released when we were out on the, and you find yourself exactly in a, you know, in, in a. Under a fig tree or something like, oh man, these guys are ready.

Boom, you, you drop that first fig in, you get the sugar high, you get, you get that dopamine. I think we do need to find the way to moderate it, but I think the other aspect is that we need to connect it to things that are, that are progressing us to things that are good to, you know, the ideas. Like again, when we were in the Savannah.

You did need to go wild and get really excited about this fig tree. Mm-hmm. Right. And like, and then go for the dopamine chase to go find the next one. Right? Sure. On the ne on the, the next time. And so the question is like, are you being force fed dopamine or is it aligned with, you know, we, we use some confetti, for example.

Mm-hmm. But you gotta work at it, you know, you gotta like, you gotta get to the right answer. You gotta explain your reason. You gotta have a deeper understanding and then you get the confetti, right. So I think, I think the, the core of it needs to be. You know, we should be thinking about engineering tools that align with what it is we want as humans.

And then dopamine is one of those ways that we can leverage and point people at, at what they want. Now again, these tools, like we talked about with, TikTok or whatnot, right? They are just trying to go overload on dopamine and just keep you there. I'm not as interested in that. I'm not a neuroscientist, that's just my armchair perspective of the matter, right?

What do you think?

Aaron Burnett:  I think that we have to moderate dopamine release. I think that we're, we're headed down a slippery slope. If what we do is continue to find ways that. Also trigger dopamine will reach a point of dissatisfaction and exhaustion. You said people doom scroll, they're on their phones a lot and they don't feel good at the end of that experience.

Right. I've experienced that. Certainly, I watch my daughters do the same thing.

Edan Shahar: Yeah.

Aaron Burnett: I find that I feel healthier. I feel a, a deeper and more legitimate sense of, well, real sense of wellbeing and satisfaction. More serotonin. The less I'm on. My device. Yeah. The less I'm triggering dopamine. Yes. I feel the ability to think longer.

Yeah. More complex, deeper thoughts. The more I stay away from my device, the more that I just read or walk or do those things that. Created a deeper, more connected sense of, I,

Edan Shahar: I'm with you. You know, I mentioned earlier the, the 10 day silent meditation retreat. Yeah. You don't get to bring a phone on that.

That's a Vi Costa thing. Highly recommended. Where did you do it? In Alaska? Just down here. I think dharma.org or something is the website. It's like the DJ Goenka as I like to call him. Yeah, he's This Goenka talks to you. Sure. And I call him DJ 'cause he is.

Aaron Burnett: I dunno, by the way, most unlikely place to have a 10-day silent of meditation retreat on Alaska.

Yeah. I grew up around there. That's a logging town. That's not a meditation town.

Edan Shahar: Yeah. Well it's awesome. And I highly recommend it to everybody. You know, if what we're trying to optimize for is like happy lives. Certainly we of course ban many things that exist today, but I think in the process of banning things like.

Who gets to choose when to ban them and you know exactly, you get into other problems of, you know, of western democracy and anything you ban becomes forbidden fruit also. True. Yeah, exactly. One of the other things that's been, that has been really interesting in AI and education oftentimes, like the main script today is AI is used for cheating because it's, I mean, like that is, and it's true, right?

I was just interviewing somebody today. Who's in college, and I asked her, how many of your friends cheat? And she said, pretty much all of them. AI is a tool that really likes to give you answers and actually a lot of the work that we do. In our prompting and in our work to get Wild Zebra to work really well.

Isn't trying to like, extract that out. Yeah. And, you are not trying to give the answer, you're trying to facilitate a process. And so what we've, we've actually heard from some of the schools that we're working on that. Homework is kind of broken. I mean, you send a homework assignment that is, you know, this, solve the, these math equations or these math problems, or write an essay on this subject, and students are just cheating on it all the time.

And the idea is, so our view is that we need to flip that a little bit. Instead of the product being the end result, it's actually the process. That is the pro that is what you turn in. Yeah. So it's, it's the process of having a conversation of how do you think about a question and how do you solve this question.

That's sort of where things need to go because. Also as people become more and more literate with AI, you know, if you think about the aspect of education that is preparing them for the future workforce, it's not just about that result, but sort of how do you think about it and do you demonstrate your ability to master concepts and to, and to think sequentially and interact with this tool rather than just providing this result.

So that's been a really fascinating process that  we're seeing and hearing and in conversations with teachers. 'cause in the platform you can actually see, you know, every single step of the way. And we're not interested in just the end. It's actually measuring the journey itself

Aaron Burnett: If you're a teacher, class full of 30 students. How do you evaluate that journey for those students in the class?

Edan Shahar: Yeah, so right now the system is evaluating the journey. You know, you push it out, all the students are using it. The results come to you and there's tooling that shows you, first of all, is there any very concerning behavior?

Students are talking. So that will, you'll, you'll find out about that very rapidly. Right. And then sort of where are students demonstrating mastery and then where are they still developing, where the AI thinks they are. Now teachers can of course, jump in and, and read the actual conversation.

And what we found is that teachers love that. Right. They’re  able to really see the students' reasoning. You know, one of the teachers told us recently that. Homework is good, but bad homework is bad. And I was like, okay, that doesn't seem that interesting of a statement, but it makes a lot of sense.

Right? Sure. If  you're sending homework with a student and they don't know what to do, that's like a waste of time. Right? Right. Like you don't need to like have them like rack their brain, but here you can have a con, they can have a conversation and there's just a different stage. Of the learning process that occurs?

Like are you still learning the basics? Are you sort of gaining proficiency but you kind of know how to do it or have you totally mastered it and flat piece of paper. Can't differentiate that for the different students, right? Like some students need to work in each of these three paradigms. Some students like, hey, just you're working on this.

Like, go on to the next thing. You, you're clearly good, right? This tool can now do that. And like, this was really interesting, you know, for the, for the teacher who, who mentioned this problem that, you know, he likes to send homework by, like, if a student doesn't know, they just dunno. It's not actually helpful to, you know, feel like you're an idiot because you've been working on it for hours and you don't know what to do.

Right?

Aaron Burnett: So what do you want people to know about you? 

Edan Shahar:I don't have delusions of grandeur that make me want for everyone  to have a certain perspective of me, I guess. But people that I meet depends on the context.

I think, if you meet me in a business context, I think it's that I think deeply about interesting problems. I think if you meet me in a, I don't know, just person on a hike maybe that I can help out if you ever get injured. I don't know. I guess it really depends on the context for me there.

One of the areas that is also interesting that I've come to learn about is how much of EO is like the more you give. The more you get out of it, sometimes you might have an experience share about something that's very shallow because you don't want to be vulnerable and open in this way. If I'm a prospective member, what things should I know?

The process works best if you are, and probably only works, but certainly works best if you are somebody who is both deeply curious about your own growth and learning and deeply open. In that process. If you find yourself to be somebody that's not interested in learning, it's probably not the right thing for you.

And if you find yourself to be somebody who's too closed without an aspiration to become more open, because I actually, I think that's one of the things that EO has changed about me is that I'm much more of an open book. It's this muscle memory of doing experience shares has completely transformed my conversations with many, many people.