บทถอดความฉบับเต็มของ Breaking the To-Do List: Jean-Denis Greze on AI Agents and Reinventing How We Work

necessarily did that. I, I can tell you- Hmm ... one experience that was, [00:02:30] um... There's a part of my life where I kind of followed people places, you know? Like, like, like, m- m- my mom was not doing super well from a [00:02:40] health perspective, and that's actually one of the reasons I moved back from the United States, where I'd lived, to France, and I went there to be, you know, a good son and, and get things in order and get to [00:02:50] spend time with her. And I felt I think, I think I am, before that, that move, I think I, I s- thought of myself as a selfish person [00:03:00] Um, and, and, you know, it was a big move for me. Like, even from a career perspective, it was not the right career move to do. [00:03:10] But, um, but I, I, I'm really happy that I did it, and I have all these memories from it. And it, it ties to this theme with, with me in life is [00:03:20] you, you have to really understand how, how short life is, and you ha- you have these choices at any moment, which is like, you may not get another chance to do the thing that you want to. And that, that one [00:03:30] was really, like, stark because, you know, obviously as your parents get older and their health is or isn't the best that it could be, you kinda realize, like, you m- you don't have, like, infinite [00:03:40] shots on goal, you know? Like I'm, you know, I'm like 47 right now, and I'm, I'm doing a startup and it's great, and I, I could do a startup when I'm 55 or 65 or 75, do you know? [00:03:50] Like, there's, I have plenty of opportunities to do that, but there's other things in life where you don't get that many, that many shots to do it. And then the other one, so my, my brother, I have an older brother who's 10 years older, who's I, I [00:04:00] for, uh, most of my life have looked up to, um, tremendously, and he, he's a kind of a hopeless romantic at times. I mean, he's married and happy, blah, blah, but when he was younger. Yeah. And so [00:04:10] he would always do these... He would, he would always, like, meet a girl. He lives in Paris. He would always, like, meet a girl that's, like, on the other side of France, and then he would randomly, like, take an airplane or a train and, like, [00:04:20] surprise her on, like, a Saturday. Mm. He's like, "Hey, I'm here. Like, let's hang out." Leah: That is so romantic. Wow. Jean-Denis: It is. It is. It's a little, you know, I think- ... y- you know, in, like, in a movie, like in a [00:04:30] rom-com, it's, uh- Yeah ... these kinds of actions are, like, praised. But, like, in real life, it's kinda weird sometimes when you do it. And so, uh, and so, yeah. I mean, I, when I went to Buenos Aires to go there, [00:04:40] it, it was for, um, yeah, it was for a woman, and I just, I, you know, I just went there, and it was kind of YOLO and, and great. And, and the, the reason I mention that one also as [00:04:50] shaping me a little bit, it's like you... A- and this is silly 'cause it's like a s- like

s- I'm gonna turn this romantic thing into a startup thing. But the startup version of it is, like, you [00:05:00] can just do things, you know? You can just, you can just go and kind of uproot and go somewhere else. You can quit your job and do it, you know? It's, like, possible. There's risks, you know, there's m- financial [00:05:10] implications, and you have to make new friends and maybe learn language and all these things, but you can, um, you can do this stuff. So those are, like, two things. But the cities, I've lived in a bunch of places. Like, the [00:05:20] reality is when you move a lot, it, it seems really great, but you're less moored to your friends and the relationships that you've built, and that, that's [00:05:30] difficult. And so, you know, the last time I did a big move was in 2013 when I moved to San Francisco, and I, I moved to San Francisco when I was in my, you know, mid-30s, [00:05:40] and I didn't have any friends here. And all my friends were in New York and Boston, and they were all having kids and doing those things. And when I moved here, it kinda reset, uh, you know, my dating [00:05:50] life and, and where... And I joined this company, Dropbox, where everyone was younger. So the cool thing about it is I have all these friends that are 10 years younger from when I moved to San Francisco, 'cause San Francisco's like a kind of a [00:06:00] young city. But it also me- means that I am unmoored and on a different kind of life cadence as, as all these other people. And, uh, you know, it changes you, [00:06:10] but it also means you- I'm no longer... I'm very different actually from the person that I was before I moved here. Not 'cause I think I have changed as much as because, like, the surroundings and the people around me and what they [00:06:20] talk about, what they think about, how they approach the world is, you know, shapes me just as much as anything. Leah: Yeah. No, I love that. Absolutely. And you were t- you told me that you moved around a lot [00:06:30] as a kid as well. What was like, what was that like growing up as, like, a 10, 11, 12-year-old? Jean-Denis: Um, yeah, so [00:06:40] my, my d- so I grew up in Europe and then my dad worked for, um, an American company, a US company, and so we moved to New York and then we moved to South Florida for, for his job. And I moved to- [00:06:50] Yeah, it, it was tough for me personally. Like, I think I was, like, a little bit introverted kid. I don't think I felt really good in, like, my, you know, [00:07:00] my body and things like that. So when we moved to Florida, I moved to Florida for eighth grade. I was this French kid- Leah: Okay. Oof, that's a Jean-Denis: hard grade didn't speak English, had a French accent. Yeah. I was super skinny. Like, I was, like, for one [00:07:10] percentile, like, you know, skinny. Leah: Mm-hmm. Jean-Denis: And I was nerdy. Like, I wore my... I was like Urkel. Like, if you remember that TV show- Oh, I remember. Yeah ... like, I wore my pants high. Leah: Yes. Jean-Denis: I was like a, a real dork, you know? And, and, [00:07:20] and in, in France it's different. Like, in- Leah: Okay ... Jean-Denis: in, in, in France and Switzerland, if you're, like... If you're not good at sports but you're, like, you know, in class you get [00:07:30] good grades, not I didn't get the best grades, it was, like, good grades, it's like, there's, like, respect for that in, in school, 'cause school's about learning. There's no sports within schools, for example. So I have this, I always [00:07:40] remember this, which is, like, so I, in the eighth grade and, and, like, you know, it's been, like, a month and I can tell that it's not really working for me socially, and I'm, like, working on it, and [00:07:50] there's this guy that's, like, really popular. Like, everyone loves him. And I, I, like, look at him and I'm like, "I don't understand." Like, he- he's, like, not that interesting. He's, like, [00:08:00] not, not that, not that smart, you know? He's, he's not that good-looking, you know? I'm like, "What is going..." I couldn't understand. And then I realized he was the quarterback of the eighth grade football [00:08:10] team. And I'd, like, never... I didn't even stick around after school to, like, know that, you know, there was a football team, but it was, it was a huge deal. And I was like, "Oh." [00:08:20] There's, there's just totally different dimensions of things. Another one, the, like, really funny things is I... You don't, you don't understand this, but, like, the, the... You don't realize this at the time, but there's [00:08:30] all the... Even after I learned to speak English fluently, there's this entire pop culture thing that I don't have. So everyone would always talk about The Simpsons, do you know? And every Sunday night there was a [00:08:40] ep- new episode of The Simpsons, and then all the kids on Monday and Tuesday would talk about The Simpsons. But the thing about The Simpsons is it's self-referential. It refers to past episodes. And so [00:08:50] I was like, "I gotta fit in. I gotta be able to talk to, about The Simpsons." I would watch The Simpsons- Yeah ... every Sunday night. Yeah. But people would talk about episodes from, like, the year before, two years before. Oh, God. And I had no... You know what I mean? Yeah. I had no [00:09:00] frame of reference. Wow. So I, I'm just mentioning these things. Like, I made- Yeah ... great friends, and actually, you know, you've, you know, within a year, year and a half, like, I, I fit in. But I think the move for me from Fr- [00:09:10] 'Cause when I was in New York, when we moved to New York and went to a French school, I went to an international school, which was, yeah, there were changes, but it was like, yeah, the cultural, the, like, social cultural [00:09:20] norms were, like, very similar. Yes, I had to, like, you know, start to learn some English and, like, my math level wasn't what it was 'cause

the schools I'd gone to in France weren't as good. There was, like, things like that, but you [00:09:30] adapt pretty quickly. But when I moved to, like, the American high school, it was like, oh, I actually don't speak English. Like, w- the English that I learn in, in French school is not, is not what's going on here. [00:09:40] So- Leah: Right ... Jean-Denis: and, but- Leah: It's a lot ... Jean-Denis: at the end of... Yeah, it's a lot, but you're more resilient at the end of it, and, you know, you realize afterwards, you're lu- you know, as I grew up, I got rid of my French accent, which is [00:09:50] one- I have some, some accent, but I no longer have the cool French accent, which obviously when I got older, you know, maybe could have wished for the ladies maybe that I could have some French accent. But- Leah: [00:10:00] I know. Did you actively try to get rid of it or at like what, what- Oh, Jean-Denis: 100. Yeah, totally. Leah: Oh, really? Wow. Oh my Jean-Denis: gosh. I just wanted to fit in. Oh. I just wanted to be like everyone else. Yeah, yeah. You know? Leah: Yeah. Jean-Denis: You know, [00:10:10] and when you're an adult, now I'm super, now I like don't wanna f- I don't care about fitting in. You know, when you're an adult you have the confidence so you're your own person, and people like you or don't like you and, but when, you know, [00:10:20] at that age you don't, you, you... Yeah, so I actively, you work really hard to get rid of it. You know, you're like, "I don't wanna sound like that." Leah: Oh my gosh. Well, I mean, I've heard you say too that moving around a [00:10:30] lot also gave you this ability to see all sides and, and not have things so black and white, right? And like, uh, how did, how did that affect you as far as like people's [00:10:40] perspectives and being open? Jean-Denis: Um, yeah, I mean, I, I have strong opinions like anybody else. I, I just, my perspective is I ... people are shaped so [00:10:50] much by their ... the environment where they grow up. You know? Like, just so much, so much, so much. And I am, [00:11:00] um I, I just, I, I really think of path dependence when I encounter... When I encounter someone that has a really strong opinion [00:11:10] about something, like political or religious or whatnot, I, I view it as like m- a lot of what you think comes from what you [00:11:20] were told in the environment that you grew up in and the values around you. And so we can argue about it, but I argue it much more, not like you're right or you're wrong, but like let's look at like the priors and maybe [00:11:30] we disagree on some fundamental things or we're shaped in different ways. Like, when I go to... You know, I also don't, I don't fit in when I go back to France. I now fit [00:11:40] in quite well in the US, but even the way I look at the world is probably different than, than your average American, 'cause I've just traveled a lot and have [00:11:50] French parents, and my brother

lives in Paris and, you know, we talk about different things. But when I go to France and, like, my brother is, you know, talking about this... You know, he's a soc- he's a [00:12:00] French socialist, so like Mamdani, actually. 'Cause Mamdani actually feels similar to me to, like, a French socialist. So that's where we are there. So he, he, you know, he's loves [00:12:10] Mamdani. He's, like, so excited that New York has Mamdani as a mayor. And, you know, and it's like, that's, like, the f- literally the first time that he says a positive thing about American politics. [00:12:20] You know? Every other time- Yeah ... for, like, my entire life, whenever I go back to France, I just hear it. Like, my brother, my cousins, they're just like, "Oh, there's l'Américain. The American is [00:12:30] here." Like, "Let's just drill him about everything that America politically has backwards." So, you know, you realize that and you're like, "Hey, if I... My brother ha- hadn't s-..." 'Cause he's 10 years older, so when we moved to the [00:12:40] US, he stayed in France. Mm-hmm. If, if he'd come here- Yeah ... then he'd probably have a very different set of values. It would've been different. Yeah. And so, like, obviously some- Mm ... things are better than others. Mm-hmm. Like, I don't [00:12:50] believe that everything's just equal, right? Mm-hmm. There's- Mm-hmm. I believe there's good and bad and things worth striving for, but I also understand that on a lot of these things, people are as much influenced by, by their [00:13:00] background. And yeah. Leah: So you started coding in 1999, and, uh, like you and I are actually, we're, we're similar generation. We are. We are. And I, I started, [00:13:10] uh, coding around the same time and learning code, C++ and a little bit of Java at the time. Um, you were writing backend infrastructure. [00:13:20] What was that like as far as, like, getting into coding? Did that make you feel like a creator, not just a student? Did it help you feel [00:13:30] like you were fitting in in a different way? Or what was the coding thing for you? Jean-Denis: Yeah, I mean, I actually started coding in, um, like, in the ninth grade, [00:13:40] mostly to automate, uh- video games for myself. So there was a- Leah: It's a good reason. Yeah ... Jean-Denis: there was something called BBSs. Yeah. Before the internet, you had a modem on your [00:13:50] laptop. It would call a phone num- it would call a phone number, and other people would call the same dial-up, would call the same phone number, and you'd all be in, there'd be, like, 12 of you in this BBS with a chat [00:14:00] room. And you chat with random strangers, but from your, like, area codes. Because if they weren't in your phone area code, you had to do long distance charges, which are very expensive I'm a [00:14:10] dinosaur. So I used to do BBSs, and one of the BBSs had, had a game. There was a couple of games. One game was called Crossroads. Another one was called Trade Wars. And what I did [00:14:20] is I wrote... It's all text. And so I wrote, my first programs were programs that, when I was at school, would dial into the BBS and play the games for me, because the more you played, the stronger your [00:14:30] characters would get. So that's, that's how I learned to code. And then I went to summer school one summer to learn to code a little bit better, and then eventually studied it as an undergrad. [00:14:40] And, and I studied it as an undergrad without asking myself why I love coding, actually. I just did it because it had been the thing that made me a little different than most of my classmates, that I, that I did the more and [00:14:50] enjoyed. I think my career... So at that time, in my first job, as I said, like, I was just writing back-end code. I think I really, it was just a thing I was good at, and it seemed [00:15:00] like the future was code, right? 'Cause the internet bubble, the, the internet bubble was happening when I was in college. And right as I graduated, it crashed, which was super, super fun to go in the job [00:15:10] market when things are crashing. Leah: Same. I was right there with you. Yeah. Yeah. Yeah. Jean-Denis: But, but it wasn't until much later, actually, in my career, like maybe a decade later, after I'd been a lawyer for a bit, that I asked [00:15:20] myself, like, "What do I love to do?" And, and then I realized what I love about coding is you can, it's a creative... Coding is creative. Now, it's not, not... The coding itself is creative, [00:15:30] like you're solving problems with the code, but it's also there's a thing that doesn't exist in the world or a problem that needs to be solved in the world. You, you implement something. You put it in front of people. You see if it solves the problem. They give you [00:15:40] feedback, and you can iterate on it really quickly. And a lot of things in the world, you can't iterate on it quickly, so the feedback cycle. So the, the, it's the mix of, like, you can be creative and get feedback and change the [00:15:50] thing so quickly. Like, that's what I enjoy the most about building software. And what's cool now is with vibe coding, I think a lot more people can, um, if they want to, I don't think everyone should code, like, [00:16:00] you know. But if you would like to, you can experience the magic of, you know, creating something and, and honing it better. And the thing now, it doesn't even have... You know, it could be like a digital nighttime book for your kids [00:16:10] or, or whatnot. It's super fun. Yeah. So I, I love that. Yeah. I Leah: love that about the- Yeah. I love that too. It's cool. That, that's what drew me into it too, was like- Yeah ... whatever was in my head, I could, I could kinda put out there, and, [00:16:20] like, people would use it. And they would give me feedback, and it was alwa- like, it just made it really exciting. I thought it was fascinating that you ended up at Harvard Law

[00:16:30] and- Yeah became a lawyer first. So, like, w- what was that path like into becoming a lawyer? And Harvard Law, that's, like, you know, that's a big deal school. That's not like [00:16:40] I'm just gonna go get a law degree at some random place. Jean-Denis: Well, I am, my... Yeah. I mean, to that last comment, I'm, I'm, I'm, I'm just, my brain [00:16:50] works pretty well, and I, I don't know why, but, um, yeah. I, uh, things like LSATs, SATs, like tests- Yeah ... standardized tests, I, I don't know [00:17:00] why, but I always do well on them. Leah: You crush them. You just crush them. Yeah. Yeah. Jean-Denis: But look, the- It's Leah: amazing ... Jean-Denis: you know, you, you were talking about traveling before, right? Leah: Like- Jean-Denis: Think about a person that travels a lot in their 20s and [00:17:10] early 30s. They probably haven't figured things out. And the reality for me is just it took me a really long time to figure out what I like to do. Like well, as I said, into my mid-30s. So, you know, I, I worked as [00:17:20] an in- you know, I worked as an engineer out of college, and then the internet crash happened, and, you know, the first two companies that I worked at basically didn't need me for that long. And, and [00:17:30] then I was like, "Ooh, this doesn't feel good." And so then I went to g- I went, I went back to what I did before, which was school. So I just went back to grad school for CS, and I was like, "Well, I don't enjoy doing this either. I don't wanna be a [00:17:40] PhD." Like, I don't wanna... You know? And I was like, "Maybe, maybe this isn't the right thing for me." And in America, when you don't know what you wanna do, sometimes the two things you can [00:17:50] get, you can go to business school or you can go to law Leah: school. Business school or law school. That's right. Jean-Denis: Yeah. Leah: Yep. Jean-Denis: And it turn, it turns out it's easier, if you're just good at taking standardized tests, s- [00:18:00] law school is better. It's easier to get in. Because bu- business school, they look at what you've done from a business perspective. Right. But I just, I had two jobs that I got- More subjective. Yeah, I got two jobs that I sucked [00:18:10] at. None of tho- those bosses are gonna write references about me. And then I've been a grad student in CS. So yeah, law school was the... It's like, how do you generalize your resume? At the time, for me, it was [00:18:20] going to law school. And I love s- I'm good at school. And I told you, I'm good at tests. So I, I crushed law school. Amazing. And I loved it, 'cause it was fun. Leah: It was fun. A lot of smart people. You, yeah, you did [00:18:30] well. You came out. You became a lawyer for- Yeah ... like few years. And then, and then Jean-Denis: you- Leah: And I hated it again ... you hated it again. Okay. So, so what was that [00:18:40] decision point of like, "Okay, this actually, I need to get out, and I need to go find what I love doing"? Jean-Denis: Yeah, I mean, so I, you know, I told you, so I, I became a lawyer and [00:18:50] then m- the m- the thing with my mom, and I was like, "Hey." I asked... The firm was great. Solv & Cromwell, great. They, like, were like, "Yes, you can go to our Paris office." And I went to the Paris office, [00:19:00] and I d- and it was... I did, I worked really hard, but I had what I called the, the terrible week. The terrible week was a week where I went to Amsterdam to work [00:19:10] on a deal, and I went i- in a basement room of this building, okay? And I looked at documents. No windows. I wa- looked at documents. [00:19:20] This is during the financial crisis. This is 2008, okay? I, I'm looking at, uh, investment vehicles who- whose credit [00:19:30] rating, when it gets bad, requires the parent company to infuse capital. There's thousands of these. You have to read all the contracts, I worked for the parent [00:19:40] company, to identify any of the vehicles where if the credit ratings went bad, capital from the parent company would have to be put in, 'cause if that happened too much, the parent [00:19:50] company wouldn't have enough money to put in. So I'm in this room. I'm doing that all day, okay? And I'm like, every time I find a document, I have to get on the phone at the end of the day with a partner in Paris and [00:20:00] explain the clauses, and I'm ema- we're, like, talking about it for a week. But my second job is at the same time, the third largest merger i- in the history of France is [00:20:10] happening at the same time. And so when I get back to my hotel at 10:00 PM, I work for like six hours on this merger. And the reason why there's a divestiture, there's... [00:20:20] As part of the merger, there's a split that's happening and, and, uh, my law firm was also representing the board of, uh, the acquirer, and they have to wr- they have to write an opinion letter. So I'm working on that, [00:20:30] and then I sleep for like three hours, and I do the same thing Monday, Tuesday, Wednesday, Thursday, Friday. Friday, I f- Now, that, okay. I'm okay long hours. It's one of my [00:20:40] superpowers. I get on an airplane. I get on the airplane back to Paris. I, like, land at Paris at, like, 4:00 PM or something. I, I go to the office. [00:20:50] I, like, finish this important work for the opinion letter, send the email out to the other law firm. It's a major deal. There's like 17 law firms in... [00:21:00] Every EU country has two law firms for each side because these, this company is all over Europe. And, and I go, uh, and I go... Finally, it's all, I'm, [00:21:10] like, done. Like, my, I've crushed my week. I'm gonna have to work on the weekend, but I've crushed my Monday through Friday. So I go to sleep at, like, 2:00 in the morning on Friday. I'm like a shell of a human being, right? And my [00:21:20] phone rings at, like, 6:00 AM, and I pick up, and it's a partner from Goldman Sachs and just, just laying

into me. Like, just, [00:21:30] just laying into me on the phone. Like, just not being a nice human being on the phone And because what had happened is another law firm had not sent [00:21:40] the opinion letter that we had finished, that I had sent at, like, 2:00 in the morning. Somewhere in the chain, someone forgot to send the thing so that then the partner was afraid [00:21:50] that it wasn't gonna be translated in whatever the seven languages that need to be trans- And so, okay, so that happens. But I'm like, "Okay, cool." And I wake up, I fix it, blah, blah, blah. On Sunday we do [00:22:00] more work. I come into work on Monday and the, like, partner that I worked for on both of these deals He's like, he's like, "Hey, I hear, [00:22:10] I hear something, you know, with Goldman, like didn't go well." And I like, I explain it, and he's like, "Oh, okay. Good." And that's it. There's no like, no thank you. No like- Ugh ... that [00:22:20] wasn't your fault and you shouldn't have leaned into... You know, g- Yeah ... it was just like the expe- Right. Leah: Ugh. Jean-Denis: I was just like- Gosh ... I just needed a little human touch. Yes. And so after the d- After that- I was just like, you know what? I'm just not doing it you're like, [00:22:30] "I'm done." I gave notice like the next day. You Leah: did. 'Cause Jean-Denis: I, it... Leah: Yeah, yeah. You know, good for you. Good for you, honestly. And then did you- And after the f- ... know you'd go back to [00:22:40] coding or at that point, or what were you thinking? Jean-Denis: No, it, it took me like a w- that's when I, you know, followed the girl to like Buenos Aires and stuff. B- yeah, it took, it took a little bit of time, but [00:22:50] effectively, you know, I started asking myself the question, like, what do I love doing, for the first time in my life, not feeling like there was a next thing that I had to do immediately. Um, [00:23:00] and I'd always been coding the whole time. I was a lawyer in law school on the side, and yeah, I just realized I liked the, that creative process. So, you know, [00:23:10] I, I went back in it. I had a friend that was starting a, like a finance company that needed someone to write some code, so I did that for a few, for a few years. And then someone that I knew had- was working at [00:23:20] Dropbox in California and invited me to... They had something called Hack Week, and they invited me there, and I was like, "Oh, these are the coolest, smartest people. Silicon Valley is incredible." Yeah, I was [00:23:30] like, "I gotta move here." And somehow I s- swindled Dropbox into giving me a job somehow. Um, I got a job there. That's amazing. Leah: I wanna go back to the moment you're [00:23:40] in the basement with no windows and you're looking through these documents. Can you imagine putting those docs in Town now and having it like do it for you? Jean-Denis: Oh. Like- Yeah, that job's d- Leah: yeah, Jean-Denis: yeah, totally ... like Leah:

30 seconds, right? [00:23:50] Like done. Yeah. Yeah. Jean-Denis: Yeah, that kind of a Leah: review is- It's, Leah: it, it's crazy. Yeah. Jean-Denis: It's incredible. Yeah. Yeah. Yeah. I mean, think- Leah: It's amazing ... Jean-Denis: yeah, all the, all the associates- So many things ... that are saving time. Yeah. Leah: I know. So many things. I gotta ask you [00:24:00] this, though. So I have this theory that when people take these detours and like go deep and become experts in a lot of different areas, like you got a law degree from Harvard, you practiced law, [00:24:10] um, my theory is that it actually does help with that cross-functional thinking when they're in, you know, you're coding, you're building, you have this new [00:24:20] startup. Do you find that the law degree or that way of thinking has helped as you have been building companies, building, you know, helping to scale [00:24:30] infrastructure, maybe in regulated spaces? Or, or how do you... Do you find yourself using it at all now? You're like, "No." Jean-Denis: I j- When- whenever- Yeah ... no, whenever [00:24:40] I talk to the lawyers, I, I warn them- Oh that unfortunately I was once a lawyer and so- Right, yeah ... I may be a slightly more annoying client. Leah: You're like the worst client for them. Jean-Denis: Exactly. Yeah. Yeah. I do sometimes read a little too [00:24:50] closely, but the- Leah: Yeah ... Jean-Denis: you know, the, the... It's, I don't know the out- I don't know what I would be like if I hadn't done all the things. Like, there are... I can write. I'm pretty good at [00:25:00] writing. I used to be terrible at writing, but- Mm ... you know, being a lawyer gets you to be- All with a lawyer ... a good writer. Yeah. Yeah. Yeah. I'm very... I can, I can be a very structured thinker, but also a very nuanced thinker [00:25:10] because most law, people think of it's like just rules, but actually most interesting law is gray zone. 'Cause if it was clearly, if the law... It was clear under the law that it was A, then no [00:25:20] one would go to court, you know? Or if, like you're writing, if you're drafting M&A, you're, you're, you have to think about what are all the things that could go wrong? What are all the things that are ambiguous and then try to use [00:25:30] language to make it less ambiguous. So the really nice thing about law is it teaches you to look at every situation from different sides, and it, it teaches you to find gaps. And in the [00:25:40] office, day to day, like when people are disagreeing about things, being able to see it from everyone's point of view, tremendously helpful. Being able to identify gaps in like situations, in [00:25:50] language, is very helpful to try to understand a market, what your competitors are, wanna do. So I, I feel tremendously lucky from all the experiences I've had. [00:26:00] Like, I think I've, I've learned, I learned so much from being a lawyer. Is it the straight line path? Like, you know, if, if I meet a 25-year-old and he's like, "Hey, I wanna go to law school to learn some [00:26:10] people skills, and like- How to write understanding ambiguous situation-" Leah: Yeah. Jean-Denis: Like, it's like, it's like, friend- You're like, "There are better Leah: ways." ... Jean-Denis: there's a cheaper- Yeah ... better way to, to get to that. But you- Yeah ... you, you are [00:26:20] shaped, you know, we said at the beginning, like you're shaped by your experiences, and they're amazing experiences. I also know, you know, I know a lot of people in DC, you know? I have friends who've worked for, for Trump and Biden and [00:26:30] Hillary, and you know, like all the, all the people. So it's, it's, it's nice that I've, there's this alternate career path and universe of folks that I know that are doing very different, interesting [00:26:40] things in the world. Leah: Yeah, different network. So. It's very cool. Yeah. Yeah. So you mentioned Dropbox and then Plaid. You spent seven years, and then you left for the first time to become a [00:26:50] founder at 45. Mm-hmm. And you've said publicly, "If I didn't become a founder now, I don't know if I ever would." Is being a founder something you had in the back of [00:27:00] your mind, like you wanted to build and create from scratch? Jean-Denis: I mean, I love building things. And so then there's ego, which is, you know, do [00:27:10] you th- do I think I could build something that wouldn't be as good or wouldn't exist in the world if I weren't there doing it? And, you know, d- different people in life are motivated by different things. Like, I am motivated [00:27:20] by pushing myself and, and seeing kind of what are the, what are the limits of the impact that I can have on, on the world. [00:27:30] And I think, I think I felt like I, I was a really good exec and, um, and, and, and [00:27:40] I, I was like, "Well, is that what I want to do forever?" And I was like, "Well, I haven't really put myself in the position to fail in the biggest way possible or succeed in the biggest way possible." And I [00:27:50] think, you know, I think being a founder for me was, like, the way to, to really test, uh, how much impact I could have in the world. And, and an- [00:28:00] and you're... I was always curious, like, could I do it? You know, you You meet other founders, and I think there's, there's two... There, over time, my emotion, one [00:28:10] of 'em, one first emotion like, "These, this person's incredible. This thing would exist and that would not exist and all without this human being. That's awesome." But then sometimes when you've worked closely with other founders, and I, I've w- you know, [00:28:20] Drew or Arash at Dropbox, or William and Zach at Plaid are incredible human beings. You realize, like, they're better than me at a lot of things, [00:28:30] but they're not that much better, and actually there's some things I might be just as good, or I might do it differently. And so

once you start to realize that- Bring a different perspective. Leah: Yeah ... Jean-Denis: yeah, yeah. Why not? Why not? [00:28:40] Why not try it? So, and then the age thing, yeah, I mean, we... You know, one of the questions you had is, like, things that shaped you and, and so, you know, for me it's like [00:28:50] my father's death shaped me and, and all these mo- But I had, like, a heart thing actually. Hmm. Uh, my sixth year at Plaid, I had, like, a heart Leah: surgery. Oh, really? Okay. Jean-Denis: Yeah. Wow. But [00:29:00] it's like in the realm of heart surgeries, it's like, you know, it's like the doctor tells you, like- Uh, I think, I think heart surgery is pretty- ... "It's never a good time for heart surgery." Yeah. It's pretty serious. But if, if you're gonna have o- Yeah. But l- it's not going to... The point [00:29:10] is, it's like, it's totally fine. But before it, I definitely was like, "Let me write a note to my son," you know? Like, "Let me do some things." [00:29:20] And, and it just, when I came out of it, I realized, like, well, life is not infinite, right? And I'd known that forever. And I was like, "When... I may not get forever [00:29:30] to do the next thing." So that got me to think about how much do I love Plaid, and I loved it so much. It was a very close call, but how much do I wanna try something else? And so, you know, it was, like, kind of a forcing function. I think if that [00:29:40] hadn't happened, I probably would've stayed longer because I wouldn't have forced myself to step back and ask myself some big questions about, you know, like, it's like money, family, happiness, like, [00:29:50] where I spend my time. And yeah, there's no right answer, but I felt like I wanted to, I wanted to see what it would be like to spend my time on something that I'm, I'm building. And then the final part of it is my [00:30:00] co-founder is, like, one of, is, like, awesome and a great friend, and getting to work with him at that point in time was like it was a... You know, he was leaving Google. There was like... [00:30:10] There was a point in time to work with this incredible human being. I was really excited to do that too. And that's just a selfish, you know, it has nothing to do with impact. It's just like, "I wanna have fun day to day. Who are fun people to [00:30:20] work with?" Leah: And that's so important, particularly, you know, not, not doing it alone is so important. What do you know after 20-plus years of [00:30:30] building inside other people's companies that you believe maybe a first-time founder in their 20s might not know yet? Jean-Denis: I know, I know lots of dumb stuff that saves me time. Like, I know how to recruit. I know how to, [00:30:40] like- Leah: Yes ... Jean-Denis: do comp. I know how to fundraise. Leah: That's not dumb though. You know what I mean? The, these are key things. No. Yeah. These are key things. Yeah. Yeah. Yeah. Jean-Denis: And they sa- they save you a ton, a ton of

time. Yeah. I w- the, I think the most [00:30:50] interesting perspective actually on what you said is what are the things that I know that stop me from being as effective as the 20-year-old? It's, it's e- the, the more experience you have, [00:31:00] the easier it is to say no to things. You know that? It's, like, easier in your brain, you can rationalize why something not gonna work. Sometimes you can also find ways to make things work that others could, but I think when you have [00:31:10] more... When you're younger, you, you think more things are possible because even if you don't know how to do it, you have less reason to believe it's not doable, and you have less things in your head [00:31:20] that tell you, like, "Oh, that company tried that and it didn't work," or, "This other person tried it and..." And so I ask myself a lot, h- how could I be the 20-year-old version of myself [00:31:30] and have the kind of naiveté and energy that I had, that I had then? Um, that actually I would say is the, the thing that I, that I try to capture. [00:31:40] In the other direction, you know, the, the advantage of being older, it's like my network. I've worked for a long time with lots of great people who I like and they [00:31:50] like me, and so I can recruit them if they're excited about the mission that we're doing. You know, I know a lot of venture capitalists. I have experience launching things. I [00:32:00] can run one-on-ones. I k- know how to set a plan. L- and all, each of these little things are, like, a little skill, but if you add it up, I think I can build an organization that [00:32:10] operates really well Without nee- you know, like faster and without, I don't need to like hire the head, the next head of engineering or- Yeah, right 'cause you've done it because we know enough about that. Yeah. [00:32:20] Exactly. Leah: So- Well, it's kinda interesting 'cause I think most of the time, right, you start as a founder and then maybe you like exit, maybe the exit's like an acqui-hire, right? You end up in [00:32:30] like a big organization, a big company, and then you're kinda like, "Oh my God. How do I operate in this big company?" But you're like the opposite. You're like, "I know how to operate at scale. I know how to build [00:32:40] teams, organizations, fundraise." And so you have like, uh, sort of like this, this growth hack, you know, almost from the beginning built in which I think is pretty [00:32:50] powerful. Jean-Denis: Yeah, but I think if I'd only worked at scale it would be less helpful. So I think what, where I got lucky is I, you know, I saw Dropbox from 200 people [00:33:00] to 2,500 and I saw Plaid from 50 people to 1,200. And so I've seen, I've seen like a lot of the post product market fit journey [00:33:10] which is where we are at now with Town. Like we're a- almost 30 people and, and we are definitely post product market fit. And so now, now I'm seeing it for the third time in a way. Yep, yep. And [00:33:20]

that's, that's the, that's what's been really helpful. Leah: You're Jean-Denis: like, "I know how to do this." Yeah. Yeah. I think the problem if you've ... And by the way, sometimes I need to remember like, oh, I don't, I shouldn't compare to Plaid at 500 people, I need to [00:33:30] compare to Plaid at 40 when I joined, right? That's like the thing you gotta remember in the experience. 'Cause I think if you only worked at scale places, it, they're great. Scale places are great but like [00:33:40] early is such a, you know, you're like in the tumbler of the, you know, the, the dryer- Leah: Yeah ... Jean-Denis: like just Leah: like surviving. Right. You don't have all the people. You don't have the layers. Yeah. You [00:33:50] don't have the, yeah, the, the structure. So I was, I didn't know this. I didn't know that Town started as this SMB tax automation company. Oh, yeah. So okay, [00:34:00] you tell me about that because you were initially raised from First Round. Love the guys at First Round. They did TaskRabbit as well. And the technology worked. You had early traction. But then, [00:34:10] but then you killed it. So tell me- Yeah ... that story and what it's like to walk away from something that technically is working. Jean-Denis: Yeah, we, we had, I would say we had, [00:34:20] like, good product market fit and, you know, getting close to a million in revenue and the feeling that we could get [00:34:30] 2.5 to 3 million of revenue the year after. So if you think of those things, you might think it's, like, pretty good product market fit. So yeah, so here's ... [00:34:40] I think we got lucky. I always think luck has a lot to do with life. You just gotta listen to it when it comes your way. Um, so my, my co-founder and I, Tony, we don't ... We knew that [00:34:50] one of the problems with the tax business is that you cannot pivot throughout the year, 'cause it's not a good situation for your clients who [00:35:00] expect you to help them prepare taxes, or the firms to whom you're providing software. So what we had always done is we had in, like, September, right before the end of the [00:35:10] tax season, like September of the year, we had a week where we said we would think about how the business is going. And so we did, we did a full year, we did a full [00:35:20] tax season, and right before it ended, but by then you know, you've got your partnerships for next year laid out and all those things. He and I just spent a week talking about the company, [00:35:30] and in the build-up to that, for the month, I was starting to put metrics together and looking at things. And the way I would tell you is the feeling was not as good as we thought it should be at that point in [00:35:40] time. And so, so then we had the week, and we were like, "Well, we gotta decide now," 'cause we're starting to commit f- to our partners who are bringing us- Leah: Right you Jean-Denis: know, businesses. Who are Leah: coming on for another cycle. Yeah. [00:35:50] Right? Jean-Denis: Yeah. Cycle. Exactly. Leah: Yeah. Jean-Denis: And the more we talked about it, the more we realized that we would spend five, seven years of our life working on this [00:36:00] business, and it would be just, like, an okay outcome. And, like, and an okay outcome in a way that I think for most people on the team, their, their opportunity [00:36:10] cost, like going to work somewhere else, made the okay outcome not, like a, like a bad outcome for them. You know? It's like okay outcome for the founders, bad [00:36:20] outcome for everyone else on the team. And I, and I think once ... And then we asked ourselves, "Well, are there ways to change the trajectory?" And we just didn't find a way that we thought made sense. Leah: And did you [00:36:30] feel like it's like a market size thing with this SMB tax market? Jean-Denis: I think m- I think there was that a little bit, but it was more [00:36:40] We didn't have a path to hyper-grow it given tax is like an annual purchase. And then the technological edge that we had, we were using AI to help prepare returns, which was going really well from a tech [00:36:50] perspective. That edge felt like at the rate that the models were getting better, like it would go away faster than we could capture a big market. And you know, and so we, we sat [00:37:00] there and, you know, meanwhile, you know, it's like to my right I have like, you know, Cursor is crushing it. I have like my friends at this, other... You know, there's like you're surrounded by people that you [00:37:10] know. Some do, some do, do well, don't well. And one of the lessons that I've learned in Silicon Valley is like you either want a business that fails or one that crushes. The worst [00:37:20] ones are the in between. B- A- And if you ever m- If you meet someone who's like spent 10 years on a b- on a, on a zombie, it's, it is the worst. You can't hire great [00:37:30] people. You can't leave. There's no s- no... You can't get acquired by somebody. Can't raise more money. You can't get more funding. Leah: Yeah. Jean-Denis: Yeah. You can't take- It's, it's the Leah: worst ... Jean-Denis: you can't take money out. And so it kind of felt like there was a [00:37:40] pretty big risk that that's where we Leah: would go. And, and by the way, the majority of companies either fail, I mean, the majority fail. Yeah. And then you've got this like [00:37:50] big chunk that end up in that area. Like I almost wish more founders that get on this venture train, because this is the business, right, would take that moment to [00:38:00] reflect and say like, "Okay, this is, this is not gonna be the rocket ship. Like what do we do?" Jean-Denis: Well, I mean, I w- 100% what you just said. It's just, it's [00:38:10] hard. It's hard 'cause, 'cause for... The, the thing that made it easier for us, why I say I got lucky is we didn't have a choice. Because if we committed to the next year... So [00:38:20] I wasn't... I

never asked the question like, "Oh, can we make a couple tweaks and get better product market fit?" 'Cause that's where most companies are. Most companies are in the zone where you always feel like you're a few product things away [00:38:30] from it, it potentially working. We weren't like that. It was like, "We're committing for 12 months," and at our burn, that was gonna be enough of our money that we could never pivot, right? And then, [00:38:40] and it was gonna work well enough that I was gonna find a way to raise the series A then, and we would s- Leah: Yeah, right ... you Jean-Denis: know, you Leah: would do that. Then you end up on that track, yeah. Jean-Denis: Exactly. So, you know, so and I, we, we chatted, we made [00:38:50] the decision, and then very quickly we, you know, we told the team, told the customers, and went back into a, an R&D motion. And, um, and clearly it's, it's worked [00:39:00] out now, and that's like amazing, and I feel very lucky about it. But I think, you know, what you said is like you wish founders would put the event on the calendar and spend the time to evaluate. And I, that, I think that's great [00:39:10] advice that you just gave. Like, if, if... You need to be honest with yourself about whether you're just climbing the wrong hill. Now, once you've decided... It's tough, by the way. Like, you [00:39:20] know, for me, I always say egg on face. Like, you know, for me, I am like, uh, never failed since I've been in Silicon Va- I failed plenty of times before, thankfully. So, you know, you haven't failed, and then maybe you have to like [00:39:30] let parts of your team go. Yeah. You have to admit to pe- You know, everyone when you're a founder- It's the worst everyone asks you, "How's it going?" Leah: Feels awful. Yeah. Jean-Denis: Yeah. But you gotta do it. You know- Yeah ... this is the thing. Life is, um, [00:39:40] if you want one more story later, I'll tell you the wine story, but life is short, and you gotta be really aware of that. And the most important thing in life, once you have [00:39:50] basic sustenance and the ability to always feel like you can get your next job, is like time. Time is the thing that is... And it just, I think you're not being true to [00:40:00] yourself as a founder if you are working on something that you don't think you can make like a great outcome, not just for yourself, but for your team. That just, it started to feel not, like not the right thing. Like I'm not a good person in the [00:40:10] world because it, it's like, you know, you go in the office, you look at someone, they're working their ass off on this tax thing that you've told them about, and you no longer truly believe you can make it a home run. Like, [00:40:20] so, you know, it's, it's easy to, it's easy to have suspension of disbelief in a positive way, but sometimes you just need to change path. And one of the nice things about Silicon Valley, which you don't get in France, [00:40:30] is people

here are very supportive of failure once you decide to take the failure. You know? But you just gotta decide to take the failure, you gotta be honest about it. You do... Or, or you just like, you wind it [00:40:40] down, it didn't work, this is what we learned. And if you do that well, people will continue to s- well, they give you a second shot, which is, which is- Leah: Absolutely ... Jean-Denis: you know. Leah: So Town's [00:40:50] current product is a personal AI assistant. I'm so happy you pivoted into this. It learns your patterns, and it starts proactively doing the work for you. I, like, [00:41:00] literally installed it. 24 hours later, like, it started drafting emails, and I was like, "What is happening in my inbox right now? Like, this is amazing." And so [00:41:10] we're gonna get a little technical for a moment, but I want you to explain this 'cause I think it's important. So you've described the approach as anti-FDE, forward deployed engineers, which is kind of the [00:41:20] opposite, right, of how most AI companies right now are deploying code and building. Can you talk about that FDE, anti-FDE, and, like, what is that? What does [00:41:30] that mean- Yeah ... um, and why is it important for Town? Jean-Denis: Yeah, so I think a lot of... The way a lot of larger companies especially, uh, [00:41:40] incorporate AI, which is the right model for them, is, you know, they will find one of the, like, big labs or a, a big consulting firm that goes [00:41:50] and uses AI to start to automate internal processes and things that are happening internally. And so what the FDE, the forward deployed engineer, they go in, they learn how the organization works and what it does, and then they [00:42:00] use AI to, like, customize and, and take the toil that humans are doing and kind of try to, like, give humans more time for the thinking work and have AI do more of the, you know, [00:42:10] scanning documents and, and all those kinds of things. And so you have, unlike in SaaS you always have products, in AI now you have AI products for things that are [00:42:20] common enough across companies. Then you have FDEs building, like, custom workflows for things that are cust- that are company specific. Like, those are, like, the two, the two main ways that software gets deployed. [00:42:30] Um, our, our thinking is there are a lot of people in the economy where they cannot hire an FDE to do and customize the [00:42:40] AI to their business, and I think there's, like, different shapes to this. Like, one of them obviously is, like, your SMB or mid-market company where FDEs are very expensive, you know? So, you know, if you can't sign [00:42:50] a $4 million or $5 million contract with OpenAI's consulting arm or another, you know, to do it, you just don't, you know, you just... You, you can only use the product version of AI. [00:43:00] So, so, so there's lots of smaller companies that are not touched. And then the other thing that's interesting is even in bigger companies actually, if there's, there's [00:43:10] a lot of functions that don't get the FDE love, you know? So the way to think about it is, like, the FDEs are deployed on the things that are seen as the most important for the business, but there's lots of other work happening [00:43:20] often in these companies that where AI could make it, the work, like, just much more efficient and, and make people much more effective where it's not touched. So one way I think about Town is- [00:43:30] We are trying to use-- We're trying to learn what you do. So we look at your calendar and your inbox and your Slack and the docs that you write. And yes, we [00:43:40] help you draft emails and schedule meetings, and we're like a general assistant. We do that well. But the second part of the product is we learn what you do, and then we tell you, "Given AI's capabilities today, it could do this for you."[00:43:50] Right? You're organizing this event. Do you want us to draft-- Like, do you want us to go through the RSVPs and all of the, like, food requirements and draft an email [00:44:00] right now to the caterer so they know what everyone needs? That's boring work. You gotta do it, but it'll just notice that you have these things dangling, and it'll just do it for you. You, you could prompt it to do [00:44:10] that, but it just n- it figured out that it can do these things. And so there, what's happening is it's basically building a one-off workflow, or sometimes it's an automation, something that's on a repeat basis that's specific to who [00:44:20] you are. You're a real estate agent. Do you want me every morning to look for new properties in your market and tell you which ones we think might apply to your current clients? You do that anywa- anyways. Maybe you can do a first pass that saves [00:44:30] you time. So it tries to notice these things, and then it gives you a one-click way to get the automation. And that's kind of what an FD does. But it's for things where the ROI of a human [00:44:40] doing that work isn't there. And our thesis is as the models get better, this kind of personalization will get better. So over time, like right now, an FD can build much more [00:44:50] complex things than Town can. But over time, as AI gets better, we will catch up, and we'll provide automations for, for users that are of, you know- [00:45:00] Almost the same level of, of subtlety and power. So we're not anti it, like it's not that I think that FD is silly. I think it's a great idea, but I think there's a lot of places where [00:45:10] it, it's doesn't have the right ROI, and that's where we can play and really, really win. Leah: Well, it is like this bottoms-up approach, I think- Yes ... to AI, right? And [00:45:20] it's like democratizing being able to have access to this type of technology and, and get into the flows of how you work. Jean-Denis: Totally. There was a TikTok, y- yesterday, someone TikToked [00:45:30] about our product, and

I was reading the comments in the TikTok, and this, uh, this person, the comment was basically like, like the reality is like I just don't have the time to learn, you know, AI. Like, I want it to be helpful to [00:45:40] me, I just don't have time to learn it. And so we're just trying to, to, to take that, that learning away. The other thing for me that I'm noticing more and more, which wasn't part of the initial thesis, but [00:45:50] we're noticing that people are doing roles that have more variety today than like pre-AI. Meaning like in tech, we always say like, oh [00:46:00] now like the engineer does more PM work and more design work, or the designer all does engineering work. But actually like what we are noticing in our like s- SMB mid-market tech companies, like tech forward [00:46:10] companies that use us, is that more and more even like, like a marketer now does some paid marketing and, you know, they do more of the things. And the reason why is [00:46:20] because AI more and more enables them to not have to hand work to someone else with a different specialty 'cause the AI can kind of do it good enough for, you know? [00:46:30] So people are, are doing a broader set of things. And so what's interesting about that is I think it means like vertical software, like software that's built for a salesperson or for a marketer [00:46:40] It responds less to people's needs because pe- people now do, "Oh, I do a bit of marketing and sales and this." Um, or, "I'm, like, do sales and, like, customer support." Like, you do more things. And, [00:46:50] and so if you can build a product where the h- it matches to the human, you're not matching the product to the capabilities, but to the human, and then the product allows you to [00:47:00] do different facets of your life, it, it's, it's tremendous. So we have a bunch of users use Town for, like, as parents with their kids, and then to help manage their vacation rental, and then [00:47:10] also for their main job as a salesperson in one product. There's no SaaS. Like, no one's like, no one goes to, like, Sequoia or Index or, you know, like, First [00:47:20] Round and is like... You know, or like Menlo or 16 or like any of the funds, and is like, "Hey, I wanna build vertical SaaS software that's for, like, parents, salespeople," and- Right ... you know, they're like- Right ... "That makes no sense." Leah: [00:47:30] Yes. But Jean-Denis: that's what- Leah: But it's about the Jean-Denis: person ... Leah: that's what Jean-Denis: AI allows you to do. It's Leah: about the human. Jean-Denis: Exactly. Leah: Yeah. Jean-Denis: That's right. Leah: Yeah. I love that. It's, it- The other thing I love about it being about the human is that you have made the brand of [00:47:40] Town so whimsical and sorta quirky, and I know this was, like, a deliberate competitive bet. Why is it so important to you that the brand matter [00:47:50] so much? Because I think a lot of other companies are just going for,

like, raw capability here, right? Jean-Denis: I mean, a lot of that is, like, Tony, actually, my, my co-founder. But the, y- [00:48:00] well, um, one of the lessons from, uh- I hate, I'm, I hate what I'm about to do. I'm, I'm gonna say Steve Jobs. Like, look, [00:48:10] mainstream, mainstream users don't care about capabilities. They just really just don't care. They just want something that's, that works, easy and simple. [00:48:20] Powerful, yes, but easy and simple for them to reason about. So, you know, like, the, the, how this plays into the brand is we think the e- easy and simple mental [00:48:30] model for a person with an AI, it's like, it's my AI, has a name, it has a look and feel, it knows me, it's, it's my, it's like a relationship. So we think [00:48:40] the brand has to represent this idea of a relationship, and for a lot of people, they're just gonna want this piece of software that they feel like really knows them, understands [00:48:50] them, has their best interest at heart, and they've built trust with that AI agent, that Townie, over time. So the brand needs to speak to that trust, and the way to, the way humans trust things, that they can see [00:49:00] them and touch them and give them names. That's how we do it. That's how... Right? We name things ri- So that, that's why the brand is really big around this concept of the, of the Townie character. [00:49:10] Now, the s- that's, that's, like, the real philosophical reason behind it. The second reason is just from a market perspective, there's a huge counter-positioning opportunity. I mean, we've made, [00:49:20] we've made AI so, like, sterile and scary, and you have to retool, and all the... It's, like, scary. You know? It's like all the companies, everything's, like, black and white, and, and it just, [00:49:30] there's just a huge opportunity in the market to be- Yeah ... like, "Hi, I'm the more approachable- Leah: It's fun Jean-Denis: like- It's Leah: whimsical. Jean-Denis: Yeah ... yeah, more fun and whimsical. Yeah. It's not for everybody, but for- Yeah ... a lot of people, they're like, "Oh, I, I [00:49:40] did need a little more delight in my life- Yes ... it turns out." So. Leah: Yes. Jean-Denis: But we Leah: are- And you mentioned... Yeah. Jean-Denis: We're early. Go ahead. Like, there's so much- Leah: Yeah ... we are- Oh, I can't wait. Yeah. Jean-Denis: We are... I think the thing, we [00:49:50] did it well in the launch, like, the, the brand came out, but I think I wanna tie the brand more to the product experience. Like, even silly things like I want you to be able to buy a T-shirt with your Townie on it. I want my Townie on it. I want Leah: you, [00:50:00] Jean-Denis: I want your Townie to listen to music, you know? And, like, it's, it's, I know it's a little silly, but, like, the more we can do to build the connection, I think the more, [00:50:10] the more of an opportunity there, there is for something fun. Leah: Well, you mentioned an important word there, which is trust, and one thing I noticed about Town when I started using it [00:50:20] is that i- in my opinion, I, you gotta tell me if this is right or wrong, I don't know, but it, it, it takes a more conservative approach in its [00:50:30] responses. Like, sometimes you put things into ChatGPT, and it just, like, hallucinates, and it doesn't tell you, and it, like, is wrong, and, like, that breaks trust with users, right? And so I started using Town, [00:50:40] and I was like, "Okay, one of the things I love about it is I know that it's not gonna try to fleece me and just, like, make s**t up. Like, thank you. That is really [00:50:50] important to me, actually, as a user." So was that something built in that was imp- an important piece of the product as well? Jean-Denis: Yeah, we, we try really, really hard. [00:51:00] We would rather have towns say, "Can't do this," or tell you, like, "Hey, I made s**t up here because I wasn't sure what the answer was." Like, we're really trying to get it there, 'cause what we [00:51:10] found, what we found from users from, like, feedback and talking to them is we're not always right. Like, I, I... You know, it's like the thing about AI is it's not deterministic [00:51:20] technology, so, you know, it doesn't always do quite what you want it to necessarily. Sometimes it gets it wrong. But what we've noticed is if the, if the LLM, if the AI, or if our [00:51:30] product tells you, like, "Hey-" I- you asked me to schedule this meeting next Tuesday, and, uh, at this time, but I, you know, [00:51:40] it looked like there was a conflict, and I didn't know what to do, so I canceled the other meeting and scheduled over it. If it tells you that, you're like, "What? [00:51:50] Why would you cancel the other meeting?" What? Right. But if it says... But, but if it then says like, "Hey, I don't know what to do. I'm not sure what to do, but I'm gonna ask you, like, would you like me to cancel the other meeting?" That's better. And [00:52:00] even if it makes the mistake of canceling the meeting, if it tells you why. So say it canceled the meeting and it said, "I canceled the meeting because I see it's a meeting with this person that I believe will be on a flight at [00:52:10] that time, and that's why I'm canceling it, and here's a, a button for you, me to un-cancel it and undo," then you feel like much better 'cause you, you understand the rationale. And if the rationale is wrong, you can go in town, you can [00:52:20] say like, "Hey, look, like it may look to you like the person is traveling, but blah, blah, blah, blah, blah," right? And then it'll behave better next time. So we're... What we're trying to do is have it, one, [00:52:30] explain what it's doing, and then you make the cost. Either it doesn't do the action and gives you a button to approve it, or if it does the action, makes it easy to unwind it is, is the goal that we have. But it's, [00:52:40] it's very hard, and we are more

conservative. I was talking to a user who uses like a... I don't like to think of competing products, but a competing product, and they were saying like, "Hey, Town writes suggests fewer [00:52:50] emails than the competing product." And I was like, "Yeah." And they were like, "Well, I want Town to suggest more emails." And I was like, "Well, the problem is, the way we've built it so far is it doesn't write more emails [00:53:00] because it's, it's not as confident about the other emails, and we think it's better not to put a low confidence draft, right, than..." But it's not what everyone wants. So the question is like, could you [00:53:10] ask, could you say, "Hey, Townie, like I want you to be more aggressive now and-" Yeah, Leah: be more Jean-Denis: aggressive ..."draft more emails." Be more risky. Yeah, Leah: right. Jean-Denis: Risky. Leah: Yeah. Jean-Denis: But it, it depends on the person, right? And it depends- Mm-hmm ... on the role. Like if [00:53:20] your, if your role is primarily like answering support emails, and there is no- enough knowledge over time from looking how you've [00:53:30] responded to support in the past to write responses, probably the Townie is more often than not correct in most of the respo- And also maybe the like if you're wrong [00:53:40] about it, it's not as big of a deal. But if it's scheduling meetings for you and including the wrong people on the invites or, you know, like not realizing there's a time zone [00:53:50] change or something like that, then it's like it makes... We call it egg on face. It makes you look bad with the other party. That's the other thing. Like if it makes an action wrong that's just for you- [00:54:00] That's bad. But at least it's just you. If it does something wrong and it makes you look bad with someone else, that's when it's not just lose trust, like egg on face. You don't want egg [00:54:10] on face. You're trying to look better, not worse. So those are like some of the things we, we wrestle with. And i- it's funny, the tech I mean, the product in those respects makes so [00:54:20] many fewer mistakes than it did three, four months ago. Like, we've- It's amazing ... been able to improve it tremendously. Leah: Moves so quickly. Wow. Jean-Denis: Yeah. Leah: So you've said that your [00:54:30] competitive beliefs about AI are secrets you keep because you view them as a competitive advantage, and I'm not gonna ask you to reveal what they are. But I will ask this: What is a broadly held [00:54:40] belief about AI that you just think is simply wrong? Jean-Denis: Yeah. And you know, I was thinking about this question a lot. I'm, [00:54:50] I'm like an optimistic person about humanity. So it's like in general, I think if m- a lot of things that are broadly like, you know, like common wisdom [00:55:00] more often than not, much more often than not, I think is, is correct. I don't even know what most people think about AI, 'cause all I do is I, I read some newspaper or

read some stuff on X, [00:55:10] and you know, it's hard to tell whether that stuff is actually representative of, of, of w- what humanity believes. I am constantly shocked [00:55:20] And I'm, 'cause I'm a child of my age. I am shocked by the amount of negativity and pessimism towards this technology.[00:55:30] Um, and I think we will look back 20 years from now and AI will have been the, like, biggest boon [00:55:40] to humans ever. Along the way, there will be tough things and, you know, like always, we will have to use government and la- and laws and, and to [00:55:50] protect humans from the negative impacts of technology. Just like chemicals, we used to put chemicals in rivers and then we were like, "Nah, that's not a great idea. Let us pass EPA [00:56:00] legislation to stop companies from putting chemicals in rivers." Like we, you know, made those realizations. Asbestos is bad. Maybe we shouldn't use it as a building material. You, like, realize those things over time. AI will [00:56:10] be no different. But I think today we should approach it... We should be optimistic about our ability to regulate AI along the way and take its benefits and apply them to society versus, I [00:56:20] think what is happening is we are so projecting the worst scenarios forward in, like, a scary way. I don't understand. It's like a, it's like as a society I think we have become [00:56:30] pessimistic in, in general. Like, Leah: I, I'm- And I think people embrace fear, right? Like, and, and it's click- clickbait too, right? Like it's, it's- It is ... it's all propaganda and clickbait. Yeah. [00:56:40] Jean-Denis: Yeah, so, uh, it, you know, if you... It, it's the thing for me that I just, I have, I have trouble with because I, I love people and I, I love humans [00:56:50] and, and I think by and large we are mostly all trying to do good and, and lead us to a better place and it is... I have just trouble with the, with the vibe generally. Uh, you know, [00:57:00] things that people... Like, it, it's a new technology. Mm-hmm. So I, yeah, I don't have a big... There's not a big thing where I feel like I, I think about it very differently than, than others. I think the one [00:57:10] unlock, I think the, the part of... AI's still by and large a single user product It's like we're, we're like the internet before Facebook and Instagram right [00:57:20] now, like equivalent of AI. Yeah, absolutely. Leah: Yeah. Right. Jean-Denis: It's, it's still... And I do think, like my wife and I, um, we have some features not everybody [00:57:30] has, but we ha- we, our, our like Townies can like work closely together to do things. Oh, Leah: yeah. I tell my husband- It- I'm like, "If you need something from me, you go through my Townie assistant." Like- Did you- I, [00:57:40] I need you in my Jean-Denis: flow ... so you enabled Ate? Y- did you enable... 'Cause there's a, a thing you can do where he can have ac- like he can ask your Townie the same things that you can,

you know? And, and- Leah: [00:57:50] Yes. Yes ... and Jean-Denis: it's... Okay. Leah: Yes. Jean-Denis: So we- Y- yes ... we use that all the time. It is, like honestly, it's... Like the other day, just like a, a silly... You know, I get every month her [00:58:00] Townie asks mine what is the balance on my credit card bills. Why? Because we need to put money in our checking account from our, you know, from our savings account. [00:58:10] And it's like, it's silly, but like my Townie knows how much money is in all my bank accounts 'cause it gets all the emails, and it knows what the due dates are, and then it just tells her. A- and it's not like we spent a bunch [00:58:20] of time on that before, but we would, you know, I w- she would text me and she would be like, "Hey," like, "when is your Visa coming in? How much?" And I was like, "Oh, I have like 3,000 coming out this month." Then we look at the bank account and we're like, "Oh, we [00:58:30] need to transfer 2K in there," you know? It just, it's a silly little th- or, you know, she's like, uh, uh, she's like, "Did you sign our son up for the, the s- the summer camp that I asked you [00:58:40] to?" And she can tell in my email if I've get the confirmation email from the, from the summer camp, right? So d- like, so- Leah: It's, it's- ... little things like that ... I, yeah. It's in... But yeah, and they, they add up, and I think, [00:58:50] I mean, that's, that's what changes the way people work. It changes their lives. It changes their interactions. If you look at the whole AI landscape right now, not just at [00:59:00] Town, but like the whole field of what's going on, where do you see the single biggest opportunity that you feel like is being missed right now, that someone's not moving on [00:59:10] yet? Is there something that you're seeing? Jean-Denis: People are mo- it's moving so fast. I, I think people are moving on all the, all the, all the things. I'm, like I would say [00:59:20] what I'm surprised by is the existing large players not do- Like the Apple with the phone not getting voice and the assistant right on the phone to [00:59:30] me is just like, it's, I do not comprehend it. Like as a matter of fact, we're building Town. It's like super successful and everything. If- If Apple wanted to provide this functionality [00:59:40] to their users in a well-packaged way, it would, you know, it, it would be like Town Leah: for- But it's so hard for these big companies, right- Yeah ... to really do from, like, bottoms up. I mean, it is. [00:59:50] It's like- It is ... the startup thing over and over again. But yes- It is ... you're right. Yeah. Jean-Denis: Like, but I think they're, you know, they're in the, in, in, in the best unique, like, in the best place to do it. Like, [01:00:00] the, they have the right to win, honestly, with, with this tech, but not able to capture it. Otherwise, you know, everything is happening. Like- This, this is why- Leah: Yeah ... Jean-Denis:

as a [01:00:10] f- as a founder, there's this... I think there's a real danger right now that we're all building companies that will be very successful. But that there's so much competition that the businesses won't be nearly as [01:00:20] good maybe as previous generations. I don't believe that about Town. But like, for example, like, it's, you know, Plaid had, like, 85% gross margins. [01:00:30] I don't know if this speaks to your audience, but, like, very high gross margins. That means it's just a, it's a great business. Town's a great business, but it's really hard for me to imagine being 85% gross margin just 'cause the inference [01:00:40] is so expensive. And if, if I try to have those margins, then a competitor will just, like, you know, go, go cheaper. And obviously, as much as I like to think that my product is quite differentiated, which [01:00:50] it is, I also live in a world of, of competition. And then there's so many startups. You know, there's people doing voice agents, and there's compu- there's, like, a lot going on. So any part of AI that you look [01:01:00] at, there's a lot of more companies and more really smart people, I think, looking at it than almost any, any at least version of Silicon Valley that [01:01:10] I remember. So the good thing for users is I think we're gonna deliver value to users faster, faster. I think for founders, the winners can be [01:01:20] Very large, high, high-impact companies for sure. But as opposed to competing with five people on day one, you're kind of competing with 12 people, and that means everything. You know, that means, like, [01:01:30] almost everything is being explored right now. The excitement, the ex- outside of our, uh, you know, I... People ask me, like, "What's the vision for your product?" Well, I think fundamentally [01:01:40] we're building, like, a device or a browser. Like, the way people interact with most information in a few years, mostly mediated through their main assistant, right? I hope they're Townie. [01:01:50] Uh, but it serves you, right? It's not like a search engine or like a feed that's trying to, like, monetize you with ads. It's like, for you, you're paying it. It's trying to do always the right [01:02:00] thing for you. Um, I think that's the biggest opportunity in, in, in, in software. But hardware is where, you know, if, if some of the [01:02:10] things happen around vision that, and, and world models that people expect, you could, you could start to see real change to how we build and, and, you [01:02:20] know, we are, um... You're gonna get a rant. You can, you can edit this out. But, you know, I love San Francisco and the Bay Area. It's beautiful. [01:02:30] It's wonderful. But you're, like, surrounded by houses from 1950 and 1960 and 1970. And in San Francisco itself, you know, uh, after they built the Transamerica Tower, they passed all [01:02:40] these laws to stop in a lot of the city from building really tall buildings, and so on. And so,

you know, if you go to China, China, everything's new and modern. You may not like China for many reasons, you may love [01:02:50] China for many reasons, but you have to respect the amount of building and, like, their malls and their restaurants and the public transportation and how safe it is and people in the streets and all this. And then you look [01:03:00] at our infrastructure here, and it's like it's frozen in time. And, and it's frozen in time for many reasons. One of them is we're not optimistic, meaning as a [01:03:10] society we don't believe we can build better cities. We do not believe we can improve neighborhoods, so we protect, right? Also, for tax reasons, especially in California, really disincentivize people from [01:03:20] building new houses and all that. But also because it's, it's expensive to build and it's tough to build better, and I think I'm very [01:03:30] excited about seeing some of the companies in the physical space enable us to improve how we live, like, physically as humans in the world. And, you know, we've had so much digital innovation [01:03:40] for 20 years that, like, a lot of people think, like, that's innovation. But if you look at medicine and the ap- application of AI to, like, drug discovery, if you look at AI in the physical world, those [01:03:50] things, we're flesh and bones and atoms. And, like, in the world of atoms, I think these things are gonna be really, really wonderful for us. At least, I'm hopeful. Leah: I love it. I love it. Well, [01:04:00] my last question was going to be, and I think you just answered it, if you could hand the next generation of founders one precedent to break that's something [01:04:10] bigger than maybe, you know, a company or a product category, what would it be? Would it be this ver- like, in the real world AI physical [01:04:20] applications? Jean-Denis: Yeah, I mean, the, the two places I'm very excited for is, like, um, the entire drug discovery and testing part of the universe, and the second one is, is, [01:04:30] is, yeah, the, the physical spaces. Or, you know, even things like companies like Zipline. You know, it's like making delivery, like, much cheaper. Also, it, it would allow us to populate parts, [01:04:40] like some space that are from a logistical perspective a little bit difficult. There's-- I don't know what it looks like in the end, but there's, there's just a ton of, of, uh, good opportunity there. But we have to be-- I [01:04:50] say this, we have to be optimistic. It's hard. You can say it. I think people have to believe it, but we, we have to get to a place where w- we think we can build a better future. [01:05:00] For most of human history, well, that's not true. See, in the, in the a- dark ages, in the Middle Ages, people actually thought that in the past, Socrates and, and, and [01:05:10] Plato and Aristotle had had all the great ideas, and there was this period of human history where we didn't think we could do better, and life didn't change for quite a long time. And in the more [01:05:20] recent time, we've been optimistic about our ability to build better societies, better government, better things, and I think that's-- for me, that's like more the scary thing that we miss than our ability to create [01:05:30] technology. Leah: Yeah. Jean-Denis: Wow. Yeah. This did not go where you wanted it to be as sort of a Leah: humanist. Jean-Denis: No, I Leah: love it. No, it's exactly where I want it to be. I love your optimism. Thank you for sharing it with us. The whole [01:05:40] conversation was so much fun. So, um, thank you for joining Breaking Precedent. It was Jean-Denis: a- Yeah ... pleasure to have you on. Leah, thank you so much for having me. It was awesome. Leah: That was Jean-Denis Grutz, a founder who didn't just [01:05:50] build one of the most important pieces of fintech infrastructure. He challenged the assumption that the people who build the rails should spend their whole careers driving someone else's train. [01:06:00] From a childhood spent moving between cultures, to a year as a Parisian lawyer, to scaling Plaid's engineering team to 350 people, [01:06:10] Jean-Denis' story is really about one thing, the refusal to accept that expertise earns you only a seat at a table that someone else [01:06:20] designed. He reminds us that the most valuable builders, the most valuable founders, are the ones who can see the systems. Not just code, but legal systems, cultural [01:06:30] systems, trust systems, and who are willing to step out of the one they've mastered to build a new one from the first line. Why does AI have to be sterile? [01:06:40] Who decided that a tool powerful enough to do your work for you shouldn't feel like something you actually wanna spend time with? And if the bottleneck for agents is trust, what [01:06:50] does it mean to earn it rather than demand it? If you know someone who's rewriting the rules in their own industry, I wanna hear their story. You can reach out to me on my website, [01:07:00] breakingprecedent.com. Until next time, I'm Leah Solivan. Displaying BP_Jean-Denis Greze_Basecut.md.