Gagan Biyani CEO of Maven on AI, Learning, and Building Big

Online learning, AI adoption, and what it really takes to build a company worth building.

Today's guest is Gagan Biyani, CEO and Co-founder of Maven and co-founder of Udemy, one of the largest online learning marketplaces in the world. Gagan has spent over 15 years building at the intersection of education and technology. His perspective on how professionals learn, adopt new tools, and think about ambition is unlike anyone else in the space.

You'll hear:

– Why Udemy pivoted from live to recorded learning in 2009 and what that decision unlocked for an entire industry

– How Gagan built Maven from scratch a decade later, starting over with first principles

– His honest take on AI adoption, including why even tech-forward CEOs still call their lawyers

– The business case for investing in employee learning and why L&D has a reputation problem worth fixing

– How to think about ambition, risk, and picking the right wave to ride in your career

– Why startups are less risky than most people think

Whether you are a GC thinking about staying relevant, a senior leader trying to build a learning culture, or someone curious about the future of work, this one is worth your time.

Follow Gagan: 

@Gagan Biyani on LinkedIn

SHOW NOTES

Books, Authors & Thinkers Mentioned:

– Ben Horowitz (What You Do Is Who You Are, cultural lore, Amazon door desks)

– Bob Iger (career ambition and picking industries that can go big)

Other References:

– Maven (maven.com) and GC AI cohort classes (gc.ai/classes)

– Udemy, Coursera, and the recorded video learning era

– ChatGPT and Claude as first-research tools for legal and business questions

– Jevons Paradox applied to AI and legal work

– Exponential adoption curves and why AI is moving faster than the early internet did

– WSGR (Wilson Sonsini) referenced as Gagan's corporate counsel

Follow us on all social platforms to get each new episode when it drops.

⁠@Cecilia Ziniti⁠ on LinkedIn

⁠@CeciliaZin⁠ on Twitter/X

⁠@GC AI⁠ on LinkedIn

@gcai_co on X

⁠gc.ai⁠ website

Transcript

Gagan Biyani (00:00)
Learning is something that doesn't happen by scrolling TikTok, right?

learning is in depth and it requires a lot of counter motivation.

you have to kind of sit down for like, you know, an hour, two hours, four hours over many weeks, you know, to really learn something. And so the internet has both

improved access and dramatically increase the amount of distraction

Cecilia Ziniti (00:32)
Welcome back to CZ and Friends, where we talk with founders, operators, and lawyers shaping how modern companies work. I'm your host, Cecilia Ziniti Today, I'm joined by Gagan Biyani, the CEO and Co-founder of Maven. Gagan is building one of the most interesting platforms in education today. He's rethinking online learning from the perspective of live and how people learn today with all this media coming at us. We've got TikTok, we've got lots of things. What does it mean to actually learn as an adult and how does it replace traditional education systems?

Before Maven, Gagan co-founded Udemy, one of the largest online learning marketplaces in the world. He later co-founded Sprig, stepped away for several years to travel. Super fun. And then having the founder blood, he chose to build again. I met Gagan through actually teaching. I teach GC AI classes on the platform and have admired his leadership ever since. This will be a conversation about building in the same industry, what success teaches you, and we'll get into AI and learning and adoption and other fun things as well. Let's get into it.

Gagan, welcome to the show.

Gagan Biyani (01:32)
Thanks for having me, Cecilia

Cecilia Ziniti (01:34)
You helped build Udemy into one the largest EdTech marketplaces in the world. When you look at that experience, what do you take from it?

Gagan Biyani (01:41)
Well, I think we got the most important thing right, which is we had product market fit and it took almost a year and a half to get there. Our original idea was live learning online via a Zoom like interface. we quickly, my job was to go and sell this live learning platform to potential instructors. This is 2009. You have to imagine that like,

There was barely educational content on YouTube at the time. YouTube was, think, a year or two old. It was very early in the internet, particularly when it comes to education. so my job ended up being extremely difficult. Most people did not want to teach online at all, much less teaching live. And about a year in, I talked to my co-founder.

and told them, look, I don't think this is going to work. Like we cannot sell live online learning. And we ended up switching to recorded online learning. And that was probably one of the biggest decisions we made that resulted in not just ⁓ Udemy's success, but also an entire EdTech, know, entire generation of EdTech companies that were relying upon recorded video rather than live education.

Cecilia Ziniti (02:55)
So the internet was still, or least connecting online was still pretty new then. What was the insight that you were going into? Was it just like, okay, now we have this potential to have teachers, the best teacher in the world that you can reach? I'm thinking like telehealth was probably similar at that time. So what was the insight?

Gagan Biyani (03:16)
Yeah, absolutely.

yeah, my co-founder, Eren grew up in a small village in Turkey and he didn't have access to, had a one room school house. So every kid, ⁓ attended, you know, went to the same classroom, matter what age or level you're at. And, he basically self-taught, himself, computer science and mathematics on the internet. you know, that was a lot harder back then. And so I think he always had the itch that.

we could make this a lot easier. And so the insight of Udemy was that the democratization that the internet provides would dramatically increase access to learning. And I think we succeeded in that.

Cecilia Ziniti (03:58)
Let's fast forward to today. So what is learning online like today? And I mentioned the TikTok overload and the like, there's these studies that people's attention spans are lower. I think back to law school and I don't honestly know how I sat through as much just instructor-driven education as I did. It's like a shock to me. So what's different and was that original promise, has that promise been met?

that anyone can learn anything on the internet.

Gagan Biyani (04:25)
I think that we live in a world where the availability of knowledge is ubiquitous. There is no limit to what you could learn if you are highly motivated and knew where to look.

The problem we have now is a problem of curation and motivation.

So today on the internet, it's hard to know what's good quality content and where to find it because there's so much of it. And it's extremely hard to self-motivate.

Learning is something that doesn't happen by scrolling TikTok, right?

Like you really don't, like there are many people in the audience who are going to be, know, X users or LinkedIn users or whatever. Like

you cannot learn anything by like reading X posts or on LinkedIn.

And when I say learn something, I mean like a subject matter, right?

You can...

improve your existing subject matter and stay up to date via those platforms, particularly when it comes to news and like what features are being launched or what's the new thing. But then when you hear about OpenClaw and you're like, I want to actually become a user of OpenClaw and spin up my own OpenClaw on a Mac mini, you can't really learn on another, on one of those platforms. So what I believe is that

The reason for that is that

learning is in depth and it requires a lot of counter motivation.

You have to avoid the natural addiction cycles of social media and the ADHD. And

you have to kind of sit down for like, you know, an hour, two hours, four hours over many weeks, you know, to really learn something. And so the internet has both

improved access and dramatically increase the amount of distraction

that prevents you from learning as well.

Cecilia Ziniti (06:17)
I mean, I'm thinking like in law school, I used to turn off the Wi-Fi. This was like when we still had Word and books and stuff. And I would turn off the Wi-Fi and be like, all right, you get one hour. And then I just furiously work on whatever it is and then turn it back on. is that, I mean, do you see that? Like obviously, you've more thought about adults learning. But are our brains different now? And what do you even do to fix that?

Gagan Biyani (06:45)
I think our brains are different, but I'll also say I'm in no way an expert in that. And, you know, I'm, a victim of ADHD, but not an expert in it. So, I, for what it's worth, think that, there are, there are two separate problems here. One is the illusion of learning via Twitter, TikTok deal, Duolingo, whatever all these, you know, sort of.

⁓ addictive products. And I think many people now recognize that that is not actual learning. That's fine. and then there's the availability of learning like, okay, well, do I know of an alternative? Can I, can I, if I want to go learn about OpenClaw or if I want to learn how to completely change my legal workflows so that I'm using AI, you know, through, your course, for example, ⁓ do I know where to go? And.

I mean, this is where you're getting at, but that's why we started Maven is to create that opportunity for more in depth expert led training that takes you out of, know, puts it on your calendar. You paid money for it. Like it takes you out of the scrolling and doom scrolling and puts you in a learning mindset. And that's why, you know, Maven's been effective.

Cecilia Ziniti (07:56)
So Maven was your third company. ⁓ So third startup that you founded. So congrats. But how did you, you know, obviously you had spent years, you know, in the space of trying to, you know, both help people learn and then monetize that. You mentioned that getting people to teach in the early Udemy days was a challenge.

Gagan Biyani (07:59)
Third startup.

Cecilia Ziniti (08:21)
So what did you do differently? did you, you know, people talk about first principles, the big thing here at Silicon Valley, I've read on Twitter or X a lot about it, but I see it now and I would say it's something that I try to do as a founder. But what specifically you're like, all right, I'm going to go at it again. I'm going to do another education company. What did you kind of white sheet where you're like, all right, or clean sheet, I guess is the term where you're like, here's I'm going to rethink all these things or maybe you didn't. But what did you do?

Gagan Biyani (08:52)
I, we redid almost everything. mean, when we started Udemy, that was 2000 and, you know, they had started working on it in 2007. I joined in 2009. And then, you know, fast forward when we started Maven, was 2019, 2020. And the world was a completely different place. This is even pre-pandemic, but then of course, when the pandemic happened, it was even more changed. So,

You have to rethink the go-to-market strategy and the product strategy and the company strategy in the lens of the new world order that you're in. So what are things that have changed? Well, first of all, whereas Udemy was simply providing access to something for which there was almost no ability to find, Maven was providing access to something for which there is ability to find.

Pretty much anything you can learn anywhere now is available in multiple places. So if you want to learn OpenClaw, you can spend hours on YouTube. And there are tutorials on there. And you can also learn from Udemy courses and Coursera and 100 other places. Maven had to take into account that context. The second thing is people were willing to spend a lot more money online than they were before.

Particular and then third, there's a much better quality online native potential instructor. So we started Udemy, we were lucky if we could just convince anybody who knew how to code to teach coding. Their credibility was not, we couldn't, we could not curate for credibility. We just picked whoever was available and over time,

credibility became more based on, you know, reviews and ratings and sales than anything else on Udemy. And that's, I think, very effective. But 2020 is a totally different world. Like in 2020, I can look someone up. Every single person who is a successful professional basically has a LinkedIn presence, has a digital presence. And there's like thousands and thousands of people like you, Cecilia who have

built up online followings via podcasting, via a substack via LinkedIn, via YouTube, whatever, who are now deemed to be experts. so instead of relying on just anyone to teach, one of the biggest differentiators of Maven is that, your average instructor is a bona fide expert in the space that they're in. They've been in it for 10 years, right? and that's completely new.

and that leveraged the go-to-market strategy of using social media as a go-to-market strategy, which Udemy could not and did not use as a strategy. We did use Facebook advertising, but social media as a direct-to-consumer strategy really didn't exist in that time period. You could do it via product integrations, but you couldn't do it via human promotion, essentially. So that was one thing. Second thing is...

Cecilia Ziniti (11:39)
Hmm.

Gagan Biyani (11:57)
I was older and your own personal context matters a lot. I am much more aware of the diversity of needs of learning that exists in the workplace. And one of the things with Udemy was it was all about basic knowledge. was all about, if you're doing video-based learning, it's very hard to learn how to lead a team via watching 10 videos. Leading a team requires

requires practice. It requires communication and discussion. It requires customization and personalization. Whereas learning Python is, you know, it's like learning math. Like you can just like it like you can just read a textbook and do it.

Cecilia Ziniti (12:36)
Just do it. Yeah. Yeah.

Yeah. I mean, it's fascinating because now I teach. So I teach a class on Maven for context for Listeners. So I started off with GC AI literally teaching a class on prompting for lawyers and taught it on Chat GPT, which I realized very quickly that it was inadequate for the needs of lawyers. So the teaching actually was what helped me see the needs of this persona even deeper. So that's kind of a fun thing. But what what Gagan is describing around like

the experience, I definitely think that's true for AI. So I say that it's like learning to ride a bike. Like you can read a book about bicycling. You can watch a YouTube video about it. You can read biking regulations. You can debate the safety of bikes. You can lobby helmet manufacturers. But the reality is if you don't get your butt on the bike, you're not going to learn how to ride a bike. And so like that's literally that's the example that I use. ⁓

I think with AI, lot of folks like the bike, if you're using models that hallucinate or you're not using an appropriate tool, it's like, oh, and by the way, the bike's going to randomly fall over half the time. And so you can see in my case that the learning is actually a big part of change management for folks because the risk of like, and you throw in the undertone of people are afraid to lose their jobs because AI does at least some things that

human lawyers did better. So that was a lot, but I guess bringing that to AI, like, what's your advice, you know, seeing and knowing how, how professionals and experts in their fields learn and teach, What's your advice on AI adoption?

You

Gagan Biyani (14:17)
I

mean, look, in my late 30s at this point. In internet years, well, I started my first company when I was 21, right? So it's been over 15 years. And even for me adopting new technology, there's a lot of groaning and annoyance and I'm set in my ways, et cetera. And I think that...

I can resonate and empathize a lot with people who have been working one way their entire lives and now all of a sudden something, you know, a tidal wave comes in and just, you know, shifts the entire playing board in front of them. It sucks. But

you have two choices. either One you can claw onto whatever you have and wait for retirement. And like I've seen that movie before, you know,

My parents certainly both had some of these challenges as the internet was coming up and having to shift into an internet defined work world was challenging. And I've seen what happens when their generation sort of didn't adopt the new tools, right? Sales went from calls to emails. That was like a huge change. And so it went from the verbal word to the written word.

the, you know, and it happened very slowly over, over, over a decade. but if you were in a field where that change occurred and you did not make that transition, you would be left behind and the same thing is true with AI. So first you can either claw, you know, try to, try to hang on and hope that things don't bother you. And by the way, what will end up happening is There will still be many, many customers out there who are not going to use AI for the next decade.

And so you can slow, you can still over the next decade, like you can go, I'll call it down the chain of adoption to those who are less adoptive of this new technology and you will still continue to serve those clients. But at the same time, what you won't notice is that there's a new set of clients that are spending more money. with their lawyers. They have more needs, they're growing accounts And those new businesses, and new people, and new customers,

are not calling you because you're not AI adoptive. And so then your second option is like you adopt it and you learn and you recognize that like your expertise is as valuable as ever. You know, the reality is, Cecilia so we at Maven, I started Maven and immediately I'd already done three companies. I had done a lot of legal work for my companies. I would say that from a certain perspective, I'm,

You know, I'm not a lawyer at all, but I can be dangerous in a legal situation. I've been in enough. I've been deposed. I've been in lawsuits. I've, you know, had trademark issues. I've, you know, incorporated probably six or seven different businesses and nonprofits. Like, you know, like you learn things. I've negotiated tons of contracts, right? Hundreds probably. and so my default is already not to call a lawyer when I have a question.

and, ⁓ that has only increased now Every time we have a legal or HR issue or a trademark issue or a tax issue, like we've actually had all of those in the last couple of years, Of course, it's kind of normal. I will first ChatGPT it. the first step. And, I'll ask anyone on my team to do the same, But you know what? I still have a lawyer in every single one of those fields. ⁓ like I still have a lawyer, right?

Cecilia Ziniti (17:41)
Yeah.

Gagan Biyani (17:44)
And there are multiple reasons. First of all, you know, it's worth like, I don't remember how much every lawyer costs, but it's like 500 to $2,000 an hour, right? Kind of typical range. It's like, it's definitely worth $500 or a $1000 an hour to discuss the potential options and solutions that an AI may or may not be thinking of

to make a decision on what to do. So I will come into these meetings far more informed about what the law says, but I still need the advice of a lawyer and really the partnership of a lawyer to discuss and determine what the next steps should be. so the profession, I think, will continue to have its own, you know,

its own opportunities, but those opportunities are going to change. So that's just my perspective as a business owner. mean, I don't think about the legal world very much, if I'm being honest. But my perspective as a business owner is I still have lawyers in pretty much every major field, but I'm using them less. And you know what? I mean, that's not that big of a problem, I think, because the truth is there's a lot less, there's a lot more demand for legal work.

than there was lawyers anyways, in my mind, good lawyers at least. Like my corporate attorneys, WSGR, like, you know, they struggle to keep enough attorneys on staff to manage me. I'm always like, I don't have, I have more work than they can do at the rate that I'm willing to do them that for. And that's been true for a long time. So now finally, like I don't, I don't need them as much and they're still super busy. You know what I mean? So I think good lawyers who adopt AI.

will absolutely and change what they do as a result will absolutely continue to be wildly successful as lawyers have been for a long time.

Cecilia Ziniti (19:44)
Wow, so much to unpack there. basically, the perspective of a CEO in 2026, of a tech forward CEO, is I do go to Chat GPT first. So I mean, that's great. I think this is literally Jevon's paradox in action. You got into a little bit what lawyers can do for you that ChatGPT GPT can't. Can you unpack that just a little bit more? So obviously, you're being deposed.

put us in your head when that's happening? Presumably you were an exec at one of your companies and what's going through your head and what did lawyers do in those situations that AI couldn't?

Gagan Biyani (20:19)
Yeah, I I was deposed a long time ago, but I'll use it as an example. First of all, there are lots of creative solutions to how to deal with the deposition, right? mean, everything from the personality one needs to take on to training someone on how short the answer should be, that's classic.

know, advice like be factual and be direct to like helping get the facts straight and collect all the facts so that you're not, you know, lying under oath, which would be a problem. ⁓ You know, human to human training is still super valuable. Like, I mean, I still respond far more to a human than I would to an AI training me on any of these things. And I think we're pretty far away.

Cecilia Ziniti (20:58)
You

Gagan Biyani (21:12)
Like I do think eventually it could happen that there would be a deposition trainer that's like, you know, digital and, uh, but still I think I'd probably for a long time want a human to view that and watch it, watch the recordings and give advice, et cetera. So that's like one version of it. Um, another version of it is, well, if you're the, you know, the, the GC who is, um, extremely

up to date on what AI can do, then you can be the one bringing me the tools. mean, there are like thousands of tools out there now to solve various problems. If there's a deposition trainer, show it to me, you know, and that like continues to make, provide value, right? Show it to me, review the transcripts, hold me accountable, like get me through the training for it.

And so I don't know everything about how this is going to evolve. You one question in the back of my mind is like, what will happen when the AI gets better and better and better? And it, it will. And my sense is that we will continue to see more and more opportunities for, for adaption adaptation. So, yeah, my, my view is that there is still at least a decade or two of runway of

essentially doing similar sets of things as what lawyers were doing before. And I can't predict further than that what's going to happen sort of, you know, five, 10, 15 years. If all this stuff gets fully built out and you know, we don't need lawyers as much, what will happen to the profession? I think there's a lot of open questions for me. And I can say that there, but there's only one obvious solution, which is, well, if you're, if this is possible.

which I don't know if it's gonna happen, but if it's possible that there's a dramatic shift in the number of lawyers we're gonna need, then the best thing a lawyer can do is learn the new tools, be at the cutting edge, because those are gonna be the people who gonna stay relevant. Second, become more business forward and more business understanding. The more you understand about a business, the more you have multiple skills that you can bring to bear together that you can provide value in multiple different ways.

And then third, would say, the underlying skill of staying fresh will actually help you with whatever changes occur anyways.

Cecilia Ziniti (23:29)
So I might consider that a spicy take, right? So 10, 20 years before things are so crazy, before AI pulls back, or you really have to change completely. But also this skill of learning new skills. So this is obviously something you're an expert in. Maven has, obviously, live classes. You also have the ability for instructors to go into companies and teach kind of like directly.

to Acme Acme is doing an AI change transformation. You've got an AI-ifier company, instructors. You've got people that teach various engineering things, leadership coaching, sales coaching, et cetera. As a leader or a senior leader at a company, which a lot of our listeners are, how do you develop that curiosity and that learning and that desire to do things in a new way?

in your company.

Gagan Biyani (24:21)
There's this great quote from a, I'm gonna give you the general answer for people in general and then I'll talk about it. Well, what I'll say about the company is that I don't have to do this at Maven because Maven is an education company and everyone at Maven just wants to learn mostly. sometimes you have to nudge them a little bit like AI adoption amongst the engineering team did require a little bit of nudging and effort, you know, and, but it's not too hard. You just nudge.

Everyone has a base assumption that learning is a good idea, but sometimes they forget to. So my job is simply to say, Hey, did you ask, did you ask Claude that? Did you ask Chuck? By the way, Claude and ChatGPT interchangeably, cause we often will prompt both. that's a whole nother conversation. But so have you asked Claude that have you asked ChatGPT that I will, I will have to ask them that. And that's enough in Maven, but more broadly, you know, the challenge of learning

is, as I said earlier, it's one of motivation as well as access. And Maven solves the access. The way that I solve the motivation is I have this great quote about reading. It's like, you don't like reading, then just read something that you like until you love reading.

And so if that means you need to read Archie comics all day because you love them or Marvel comics, even though, I wouldn't consider that enriching your brain too much, it's still building the habit of, of reading. so the same thing is true with AI. Like if your first AI usage is going to be like, you know, for example, for me, actually my first day I uses was, was using AI to pair while I was playing video games. was playing Zelda on my switch.

Cecilia Ziniti (26:01)
Wow.

Gagan Biyani (26:02)
And probably like the most deep, obviously I'd already started using AI, but the most deep AI usage I was using was using AI to give me advice on what to do in the game I was playing. And that was actually like for me, the easiest way to get started. And of course it was unfortunately like with tutorials in particular, AI is really, really bad. So it was, it was wrong like so much that it was super annoying also, but.

It was also really fun. Like I just had a phone next to me while I was playing the game and I'd take screenshots, you know, of the game or I took pictures of the game on my TV and I'd send it to it and I'd chat with it and I'd be like, you're wrong. Like blah, blah, it didn't work, et cetera. But it was fun for me. And that led to my, that was the start of my AI adoption. So I'd say if you're having trouble with AI, just adopt it somewhere where you know it's going to be helpful.

Cecilia Ziniti (26:51)
some way. Yeah,

no, that's exactly right. And I also it's the same is true for evaluating different models, right? Like learning their personality. So my husband's same thing was playing a one of these video games and was asking ChatGPT for advice. And it kept hallucinating. And it made it very real to be like, you know, go to this level and you'll meet this boss. And then that wasn't the case. And then you guys Gemini and Gemini, you know, being Google's product is like super trained on this stuff. And so, anyways, Gemini came back and was like.

Gagan Biyani (27:11)
Yeah.

Cecilia Ziniti (27:18)
uber precise and that was actually one of the reasons he switched to Gemini as his primary was that like these taste things and we spent a lot of time at GC AI thinking about like okay what are the in-house use cases, which models perform better, what can we tweak and that's kind of part of our IP. But in terms of like for individuals I couldn't agree more when we teach the first class like that aha moment of like okay you know ask a simple question like give me some interview questions to interview a product counsel.

That's kind of table stakes. And then you can get into this deeper stuff. But I agree the experience of doing it. It really is. I keep coming back to that bicycling metaphor or cooking metaphor. And there's another there's another good one that's like anyone who knows how to eat well can soon learn to cook well. But it's like, OK, you have taste for what you consume and, you know, text that you consume clearly. And so developing that taste around AI, it's it's definitely a question of use. I love that. So.

OK, so I so I'm a senior leader. So let's bring it back to the very senior levels where, you know, learning you've thought a lot about the kind of like business benefit of a learning culture. Any good stories or like, how do I make this dollars and cents? And then, you know, I'm the CEO of Acme or, you know, the CLO or CMR or whatever. And I'm debating like, OK, I'm going to invest in learning or I'm going to nudge my people harder.

What's the business case for him?

Gagan Biyani (28:42)
It's a lot more expensive to hire new employees to replace your existing ones than it is to train them up. And it's really unfortunate, actually. It's great you're asking about training. I haven't written this publicly yet, but I this thought in my head, which is that L&D has taken a serious reputation hit over the last couple decades. People don't want to do learning and development within their companies.

And I kind of blame, like, you I feel like Udemy has some accountability here because it's one of the number one L&D software providers in the world. And I think that this is like the moment for L&D to step up and show that it's valuable. Right. And so if you're a leader at a company, you might've been burned in the past by investing in the training that didn't really work or moments where

You know, it was kind of unnecessary. Like people were just doing it for fun, et cetera. Like this is not that moment. Right. And like, as a leader, you have to be adaptive in what tools you prescribe and what you do at different moments in time. Like today and taking an existing employee and making them AI enabled will like two to three X their productivity. So let's just say you need to stop that employee's work for a month.

This is not the way I would recommend learning, just like to point, yeah, it would still be worth it. Like it would be worth it for six months, right? Like if you, you, if you had someone spend six months learning AI and you knew they were only going to stay for one year afterwards, you would still get enough productivity over that year that follows that six months. So one month is nothing. And actually the way that you can learn AI, which is you shouldn't learn it.

Cecilia Ziniti (30:05)
Assuming that that was the trade-off, it would still be worth it, yeah.

Gagan Biyani (30:32)
you know, just in a dedicated slug like that. The way you should actually do it is tell your employees, look, you get two hours a week to learn AI or four hours a week or whatever for six weeks, you know. And like, you can do that three times a year or twice a year. Like that is absolutely doable.

Cecilia Ziniti (30:47)
It is, you know, this is classic.

I mean, it's also like one of these like, you know, even just stuff like how productive you are in email or like, you know, literally I had a thing where I created a bunch of hotkeys on my computer. So for like my phone number, for common things I do in sales, just like I literally do like at class and that'll be like, we teach classes and it'll like have a blurb or like templates in HubSpot and your CRM or.

For lawyers, we spend a lot of time on our templates. And it is the slowdown to speed up is a very hard case to make. It's also one of these, as a CEO, got to say, in hiring for us, we're scaling like crazy. And when I interview people and they have not even started the journey, it's basically an automatic no, I will say. It doesn't matter because we've got very experienced, very senior execs for

a lot of roles and I'm like, I don't want them to have to unlearn. Like it was literally we were talking yesterday that it's like, OK, if they're really, really good horse carriage operators and we're hiring for cars, can they even do it? You know, like and there are things that are transferable, obviously being on the roads and so on. But it's definitely like and it feels you know, I guess maybe that's a question for you, because I was early Internet as well.

I ⁓ my first job on the Internet was in 2001. I was a chat moderator and, you know, saw, you know, read it and Craigslist eventually get kind of Craigslist, all the different verticals of Craigslist became verticals and very much like an Internet enthusiast from the day. But I guess it did not feel as fast as this was was. Is it faster this time?

Gagan Biyani (32:31)
It wasn't.

It's way faster. mean, but that's the nature of an exponential. We're in an exponential, you know, and so just to really explain the exponential, you know, adoption curves are relying upon past adoption curves. let's say in 2008, sorry, in 2000, I think I remember hearing that there's about 80 million people on the internet today.

I think we're at like two billion, something like that. I don't know exactly, but it's billions of people, right? So any update to how you use the internet is going to be highly reliant upon the number of total people who are on the internet. And while AI is a fundamental shift, it is still an update to how you use the internet. It's not a new thing that you use outside of the

builds on top of the existing adoption curves. like, you know, for example, if you started a web company in 2000, you had to wait for the adoption curve to get to the exponential, because the adoption curve of the internet was exponential that whole time. And, you know, that's why Facebook starting in 2006 or something like that, 2005, like Facebook started at the right time for that adoption curve and it...

grew faster than anything else because the underlying growth rate of the adoption of the internet corresponded at the right timing for Facebook. And the same thing is true with AI. So yeah, of course it's moved faster. And whatever next thing, like when AI becomes ubiquitous and the internet adoption is still growing, basically it's now related to class. So as soon as you have enough money for internet, you're using the internet.

And so as that is, you know, and by the way, like the definition of internet users really important here because mobile phone users in third world countries are also using the internet, but just not in the way that we're defining it in this context. So everyone uses the internet essentially. But the point being that like, you know, that sort of like few billion people who are using the internet actively and are on a web browser or whatever, like

that becomes the base for any new technology. And we also have social media, which is sufficiently adopted that it created the information liquidity necessary for AI exists. now everyone knows about it because the speed from when someone creates an invention, whether that or finds a new piece of information,

to point at which it is ubiquitous has dramatically declined over the last, you know, whatever, over millennia, but certainly over the last hundred years.

Cecilia Ziniti (35:17)
I just I love this. I love this topic. I could literally talk about this all day. But yeah, I mean, it basically like is there something in this new move faster world? You know, you mentioned you can get things out more quickly. You're a multi time founder. You've had wins. You've had setbacks. You've had regulators trying to shut down one of your startups. You've done travel. Give us some some pithy advice or maybe a story about about ambition.

and what's different now if it is.

Gagan Biyani (35:44)
Yeah, I think, ⁓

You know, pithy advice is like, you can only achieve the ambition you have the mind for. you know, you have to give yourself the opportunity to reach the level of success that you will eventually reach. The opportunity has to be created somehow. So because of that,

My experience has been both positive and negative, Like positively, like, you you to me ended up being a billion dollar company at IPO, right? That was only possible because we had decided to go after a market in a way that had a chance to become a billion dollar company. ⁓ we didn't need to plan it. We did plan it, but you don't have to have it planned, but you have to have, give yourself that opportunity to play on that stage with that mindset.

At the same time, YouTube is a, I mean, it's gotta be a half a billion, half a trillion dollar company at this point. And so we didn't play on the stage of like building the underlying video ecosystem of the internet. We built things on top of that, know, a certain type of video, specifically paywalled videos for educational purposes. And so I think like one big lesson for me,

on watching Udemy versus YouTube is well, YouTube obviously has way more impact and it's bigger business. Even if you just looked at the educational arm than Udemy did. And that's because in part, because those founders put themselves in the line of the ambition of like a bigger ambition. Like if they succeeded, their success would result in more than what Udemy's potential success was. And I think many people

are setting themselves up for ambitions that are even smaller than that, smaller than a Udemy or a Coursera or something. And so just taking that lesson and thinking about it over my career is, it's amazing to me just how big companies and ambitions can really be. And it puts into perspective any decision that I make about what I'm trying to build.

Cecilia Ziniti (38:02)
Yeah, it reminds me a little bit of, so Bob Iger wrote a book and the longtime CEO of Disney and talks about growing the business. And he makes a similar point where he says, just from a career standpoint, I forget the example he used. I think he said guitar picks. He's like, if you decide to go in and make guitar picks, you can be. ⁓

you know, really good, have the best guitar pick or whatever, but ultimately it's never gonna be that huge. So he would filter both internal ideas and his own career in that way. So a lot of our listeners are folks who, you know, are maybe the early career or kind of manager level, you know, shy of founding a startup, which of course is your and my bias.

What can they do to put themselves in that good path? Is it picking your companies? Is it getting on new tech waves early? What would be the consequence of the advice you just gave?

Gagan Biyani (38:58)
Yeah, mean, it depends on the scale of your ambition. It's important to qualify that ambition comes with risk. you know, and lawyers are great at evaluating risks. you you decide for yourself what level of risk you want to take relative to your ambition. But then once you're in a risk category, right? Risk category could be, hey, I'm willing to have a zero. If it means like a big opportunity that's found in companies, right?

⁓ Or, hey, I want to make sure that my family is taken care of no matter what, but I'd love to be at bare minimum revenue for my family with the potential of becoming a multimillionaire or whatever, becoming really big. I'm just giving examples here. And then the lowest level of risk is, I just want a stable career where I'm happy. And once you put yourself in a risk category, you know what category roughly you're in.

there's still a lot of variance in outcomes within those categories. I think that's what people really miss is like You can have a safe job, but then what you should do is evaluate the ambition of the company that your safe job is at. And evaluate the ambition of the people around you and put yourself in positions where even though you're riding the wave, you're not creating the wave.

Cecilia Ziniti (40:09)
Yeah.

Gagan Biyani (40:19)
If you ride the right wave, it becomes really big. like, you know, I know plenty of people who have been far more successful as managers and executives than they have been as founders, than I have been as a founder, I should say, or anyone else. And I'm just talking about financial success here. I think it's a good proxy. There are so many other forms of success. And personally, like I wouldn't trade my situation for these people or they probably wouldn't.

vice versa, right? it's not meant to be broader than just financial success, But in terms of career and financial success, you know, if you were one of the first GCs at OpenAI, I mean, you're worth $100 million. right? Like,

Cecilia Ziniti (40:59)
You're in good shape. You're in good shape. Yes, this is

true. And I mean, that was my law school. So my law school friend, Brian Israel, I'd to get him on the podcast, was the first lawyer at Anthropic. And that's exactly right. literally, he and I worked together on his paper about early internet. I was his mentee on that. yeah, I mean, think that what's hard is I think the lawyer personality is very, there's a lot of folks who are going to be in that boat. And then there's this idea of, well, it's luck.

OK, that like he picked that. But then it's like, OK, you know, look at your career as a portfolio would be my advice of like, OK, you might have my first startup ANKI was was, you know, had all the indicia of being a very successful, ambitious company. had index on the board, Andreessen Horowitz. I met Mark through that job, you know, but then it was hardware and it went to zero. But, you know, what I learned has now put me where I'm here now. So I think it's like

The advice I give is it's like a mandala of like, what is the experience? But I don't know. It is tough. Any advice on picking companies or like on finding those people? Is it a feeling? Can you interview for it? How can you tell in advance, if at all?

Gagan Biyani (42:13)
You should ask yourself, if this works, how big can it possibly be? And then you have to start thinking about like, okay, what are the chances of it working? And you sort of like, in your head, you kind of multiply these two things together. But actually what you should do is you should evaluate them relatively independently. So what you should do is like,

You know, Hey, if this AI thing works, it's going to be this big. if Anthropic work is going to be this big. Okay. That's a whole different category than other things working. And the biggest problem with exponentials is it's, you know, it's not just exponential growth rate. It's also exponential differential between the top percentile and then the next decile.

And then the next, you know, and then the top quintile and then the top, I actually don't know the word after that, but the top 20%, the top 50%, et cetera.

each of these is materially different from each other in terms of outcomes in startups and tech. And so you want to be shooting for top decile, at least outcomes.

And if you're in the top decile outcomes, then you just have to ask yourself whether you believe it has a chance of success or not. And that should be a binary question, not a percent question. If you percent multiply and you try to create an expected value, you will fuck up the math because of the exponential curve.

Cecilia Ziniti (43:42)
it's

like that, like, I always say, like, you know, three quarters of a percent of nothing is still nothing, you know? And so it's like that binary, that's exactly right. Yeah, I love that advice.

Amazing. All right. So, so we're getting close on time. Let's do a lightning round. So a myth about startups that, ⁓ that you think people have that you'd like to dispel.

Gagan Biyani (44:03)
I think a lot of people think it's more risky than it really is. I mean, at a macro level, most of my friends who started companies in 2008, they all pretty much failed except for maybe one or two of them. I'm thinking about like my founder Institute class and my first incubator that I was in. They all failed with their startups. As far as I know, they all had amazing careers and are financially quite successful and happy and you

Et cetera, not all of them, but most, the vast majority. And that is telling, you know, that tells everything. ⁓ Startups are really not that risky anymore. And I don't think they were back then.

Cecilia Ziniti (44:42)
love that.

I kind of agree. mean, so certainly

for employees, I would agree with that. ⁓ What's a founder or a book that has shaped how you think?

Gagan Biyani (44:52)
I just read Ben Horowitz's book, What You Do Is Who You Are, and I've been thinking a lot about it. mean, this idea of how you create cultural actions rather than just cultural values that remind people of the cultural values is super interesting. And honestly, I think he kind of undersells it in the book a little bit.

⁓ he just gives a lot of really great examples, but I think the core message is, look, if you want a high ownership culture, create like little, create lore within the company about moments where people had high ownership and it really worked out for them. And then, and he does that, he does explain that really well. and then the other, the other thing that, you know, is create like specific.

recommended actions that occur that people can never forget. The classic example is at Amazon, Jeff Bezos famously wanted to create a culture of frugality, which he did. He gets railed on it for it today, but it's been extremely successful. And one of the main things he did was everyone starts with a door as their desk.

So when you have a desk, you could pay like whatever $250 at IKEA to buy a desk. And instead they'd go to the hardware store and buy a door and nail four pegs to it. that would be, that would cost like, I don't know, 50 bucks or something. I'm just making up the numbers, but it'd be a lot cheaper. so Amazon for a long, long time, everybody had a door as their desk.

Cecilia Ziniti (46:29)
Yeah.

I love that. If you could leave listeners with one thought about building after both success and surviving regulatory and now in this age of AI, what's your one thought you want to leave us with?

Gagan Biyani (46:47)
do it, go do it, do something about it. Like, it doesn't matter what it is, just start doing something about it and allow yourself to gain some early wins and then just let it snowball. Like it's, it's an incredibly powerful thing and it'll change your life for the better if you can, if you can use it, you know, to your advantage. And if you don't, it'll change your life for the worse. So you kind of have a clear choice there.

Cecilia Ziniti (47:12)
Love that. Topics of agency as well. Amazing. Thank you so much for being here,

Gagan,

Amazing.

Gagan Biyani (47:17)
Thanks for having me,

Cecilia. And thanks for all your continued support and partnership.

Cecilia Ziniti (47:21)
Amazing, I love Maven, so check it out, maven.com / Cecilia Z or you can go to gc.ai / classes, it'll redirect to Maven. That was my conversation with Gagan Biyani, the CEO and Co-founder of Maven. We use Maven to run our cohort-based classes, and it has changed at least how I connect with folks. It's been incredibly valuable for my own career as well as for many of our users, and really ties back to.

some of the themes of today's episode around you have agency to learn, you have agency to take this next step. If today's conversation got you thinking about AI, of course, visit us at gc.ai, connect with us or follow CZ and Friends wherever you get your podcasts. Thanks for listening and we'll see you next time.