अनुलिपि 🔗
WEBVTT
00:00:10.160 --> 00:00:17.359
Hello friends, it's Fraser Anderson and welcome to Edge Cases, where we explore the frontier of what's possible in AI.
00:00:17.519 --> 00:00:21.920
Our guests are the exceptional founders and investors building tomorrow, today.
00:00:22.160 --> 00:00:25.920
Like the technical limits they push against, they're edge cases too.
00:00:28.160 --> 00:00:29.920
Annalise, welcome to the show.
00:00:30.239 --> 00:00:30.640
Thank you.
00:00:30.800 --> 00:00:31.920
Thanks for having me.
00:00:32.159 --> 00:00:34.960
Is this your first time appearing before the media?
00:00:35.200 --> 00:00:36.719
Is the public getting to know you for the first time?
00:00:37.200 --> 00:00:41.679
Well, I mean, the media with a following as huge as you're creature, yes.
00:00:42.000 --> 00:00:46.000
I have said little things before, but nothing compared to this podcast.
00:00:46.320 --> 00:00:46.799
That's for sure.
00:00:46.880 --> 00:00:53.439
I mean, we have we have millions of viewers and listeners, so it's it's a real pleasure to break the story.
00:00:54.079 --> 00:01:03.359
I was doing some fairly extensive research on your background, and the thing that I'm a little bit upset that you didn't let fall when we were together in SF is that you're very fast.
00:01:04.239 --> 00:01:07.280
Um, or at least at one time you were very fast.
00:01:07.680 --> 00:01:15.840
So I'd I'd love to talk a little bit frankly about the psychology of something as painful as cross-country and track and field.
00:01:16.079 --> 00:01:29.680
And because as I as I looked kind of further into your CV, you know, I I I had to come to the conclusion that I think I probably had a lot more fun in my 20s than you did because you were so busy.
00:01:30.159 --> 00:01:43.359
And when you think about accomplishing a lot of things in a short amount of time, do you think that sports, do you think track and field has anything to do with your mentality today and the way that you think about racing through your career?
00:01:45.439 --> 00:01:46.959
I love what you did there.
00:01:47.200 --> 00:01:49.040
I've always been hyper competitive.
00:01:49.200 --> 00:01:58.959
I think like I actually only did track to get in shape for lacrosse back in like junior high or high school, I guess, but then ended up being good at track.
00:01:59.120 --> 00:02:02.079
But it's a love-hate relationship I have with running.
00:02:02.159 --> 00:02:07.599
Like, I mean I love it, but um I also loved team sports a lot.
00:02:07.840 --> 00:02:18.879
But I think for me, running is like you are in total control over yourself, over your own decisions, over both the race, then leading up to the race, your mindset after the race, right?
00:02:18.960 --> 00:02:21.120
Like, and so I like that, right?
00:02:21.199 --> 00:02:24.879
And it's like a constant competition against like trying to beat myself.
00:02:24.960 --> 00:02:31.360
So I think there's a lot of parallels to like how I think about competition today.
00:02:32.080 --> 00:02:46.000
Um I remember thinking like during during racing season, like, oh my gosh, you know, if I just get through this race and then I'll have so much time to be able to like think exclusively on like studying or whatever it might be.
00:02:46.159 --> 00:02:50.719
And it was this like overarching kind of stress that would linger over me.
00:02:51.199 --> 00:03:01.120
And uh so it did allow me to kind of like relentlessly prioritize and compartmentalize things, which I think has has helped a lot now.
00:03:01.360 --> 00:03:02.000
Yeah.
00:03:02.560 --> 00:03:07.599
It probably will be no surprise that you went from there to start a career in investment banking.
00:03:07.680 --> 00:03:12.639
Did you always think technology was the end state, or was this just like one more competition to win?
00:03:13.280 --> 00:03:16.240
Yeah, I so wish I had done.
00:03:16.479 --> 00:03:21.599
I mean, I think everyone kind of says maybe not everyone, but I I kind of looked back and went, what was I thinking?
00:03:21.759 --> 00:03:22.479
Why did I do that?
00:03:22.560 --> 00:03:27.840
I think in college I had blinders on, it was either consulting or banking.
00:03:28.080 --> 00:03:37.599
And I was an athlete, I was really competitive, and I was like, well, you know, people say that's more in line with a banking, a banker's personality, like might as well do that.
00:03:37.680 --> 00:03:46.400
I did the internship in New York, hated it, like really didn't like it, but was like, you know, what marketable skill set do I have?
00:03:46.479 --> 00:03:48.479
I'm like political science major.
00:03:48.639 --> 00:03:54.879
Um, I like know how to think and write, but I don't know, you know, I don't know the first thing about business or entrepreneurship.
00:03:54.960 --> 00:03:58.400
I kind of knew that was like the North Star that I wanted to head towards.
00:03:58.560 --> 00:04:03.680
My dad's an entrepreneur, has been my whole life, and I think I saw that and was like just in awe of it.
00:04:03.919 --> 00:04:04.879
So I got the offer.
00:04:04.960 --> 00:04:19.199
I actually applied to transfer it out to to Menlo Park, which is so grateful that I had at least that foresight because it brought me out to the Bay Area in 2013 and exposed me to kind of like everything that was happening here.
00:04:19.600 --> 00:04:25.680
And I think if I had stayed in New York, I would have kind of gotten onto this, like everybody around me was miserable in New York.
00:04:25.839 --> 00:04:28.319
Like everyone was doing kind of that thing.
00:04:28.560 --> 00:04:39.279
Um whereas you come out to the Bay Area and everyone's working on like really cool projects, doing startups, and you kind of pick your head up and be like, wow, there's so much more beyond this.
00:04:39.839 --> 00:04:44.959
Um, and so left left banking pretty pretty quickly to join uh Dropbox.
00:04:45.279 --> 00:04:55.439
Yeah, I would say that uh I mean 2013 especially, those
Load full transcript
were kind of like the Halcyon days of SF, like kind of 2013 maybe to 2016 when people were like optimistic about Uber.
00:04:55.680 --> 00:04:58.800
You know, people seem to be having a lot more fun, at least than they are today.
00:04:58.959 --> 00:05:03.040
Um and like the transition to Dropbox, did that feel natural to you?
00:05:03.199 --> 00:05:06.879
Did you have any of the your kind of your friends saying, like, are you crazy?
00:05:07.040 --> 00:05:09.920
Why get off the fast track to to join this startup?
00:05:10.079 --> 00:05:12.079
I mean, how did that kind of happen for you?
00:05:12.480 --> 00:05:23.920
Yeah, actually a friend of mine who I went to college with who actually got the offer at drop at um the bank declined that offer and I actually ended up taking his spot at the bank, and then he went to Dropbox instead.
00:05:24.079 --> 00:05:27.600
And he was there, and I remember calling him like two months in being like, help.
00:05:28.399 --> 00:05:29.519
When by done.
00:05:29.920 --> 00:05:37.600
And so I had a couple friends at Dropbox at tech companies, um, although not many, like most people were doing the consulting being route.
00:05:37.920 --> 00:05:43.120
Um, I just remember in college when he declined that offer, thinking like, wow, how crazy is that?
00:05:43.279 --> 00:05:46.160
Like he's going to like startup, what is he doing?
00:05:46.800 --> 00:05:51.600
Um, no, I mean, I honestly I don't think I like consulted many people about the decision.
00:05:51.680 --> 00:05:55.839
I I remember like my parents being like, Are you sure you want to do this?
00:05:56.160 --> 00:05:58.720
But I just knew deep down.
00:05:59.120 --> 00:06:09.439
I think it was that you know, I looked at the people above me in the role and I was like, I don't want to grow up to be like you, and that was like all I needed to know.
00:06:09.759 --> 00:06:10.560
There you go.
00:06:10.800 --> 00:06:13.199
What um what drew you into venture?
00:06:13.600 --> 00:06:20.879
Honestly, I I got into venture because I knew I didn't want to go to business school, but I also didn't know, but I also didn't really want to stay in product analytics.
00:06:20.959 --> 00:06:28.319
So basically I drop box, I done a rotational program, the first year of the rotational program there, which was really cool experience.
00:06:28.399 --> 00:06:37.519
It was kind of right before the hyperinfection grows at the company, and so getting to just like ride that wave, work with all the incredible talent that was there at the time.
00:06:37.839 --> 00:06:48.079
And so I did like sales, customer support, partnerships, user research, and then eventually helped to start a team called product analytics that was basically everything on the quantitative side of product development.
00:06:48.800 --> 00:06:54.160
And did that basically like kind of power SQL user type energy?
00:06:54.480 --> 00:06:54.800
Yeah, yeah.
00:06:55.279 --> 00:06:57.839
Like living and breathing in SQL, exactly.
00:06:58.480 --> 00:07:08.240
Um and it was essentially cross-stages, so it was like pre-product when they're ideating, like where are the pools of users, where are things that we can kind of like pull out of, like insights, right?
00:07:08.319 --> 00:07:35.759
It the team actually initially started as a it was being called insights and analytics, so it was a combination of qualitative and quantitative research, so like doing qual research kind of layered with the quantitative side, and then like depending on where we were in the product life cycle, you know, flexing one or more one of those at a single time, but then they decided to build out the user research team instead, and so that focused exclusively on qualitative, and then we built out the product analytics team.
00:07:36.079 --> 00:07:41.279
I mean, it was a great like experience in terms of just learning, learned a ton, right?
00:07:41.439 --> 00:07:47.519
Building a team, but I knew I didn't want to like stay in product analytics, and so then when I say what do I want to do after Dropbox?
00:07:47.680 --> 00:07:54.160
I was like looking at roles, but I still kind of felt like you know, what what is the role that speaks to me here?
00:07:54.319 --> 00:07:57.600
Um I'm pretty sure it's not an analytics role.
00:07:57.839 --> 00:08:21.040
Uh and I also knew I didn't want to go to business school because I was like, I'm in the heart of where I want to be, like, why take two years off, pay a lot of money, and you know, and so then just got connected to some people in venture and um and I and had a conversation with uh the team at WTI, which is known as a debt firm but does a lot of equity as well.
00:08:21.199 --> 00:08:29.600
Yeah, and for me, like that at that point in my career, it was like I just want to be exposed to as many things as possible, like I want to see as much as I can, right?
00:08:29.680 --> 00:08:36.480
And so the model WTI where you you do see just a ton sheer volume because of the debt side, right?
00:08:36.559 --> 00:08:46.399
But then you also get to kind of double click because half of the returns come from the warrants equity upside, and so you are doing kind of that that that as well, yeah.
00:08:46.480 --> 00:08:48.320
Um, like that analysis as well.
00:08:48.480 --> 00:08:55.679
And so yeah, really like a team and and and ended up there was like this will just be my replacement of business school.
00:08:55.840 --> 00:09:02.159
Um ended up staying for four years, and then sort of the rest is it goes from there.
00:09:02.480 --> 00:09:06.399
Yeah, and and and shortly after that you got the chance to work with Kevin Wil.
00:09:06.559 --> 00:09:07.519
What was that like?
00:09:08.000 --> 00:09:14.720
Yeah, so I had no I met Elizabeth like my first day in California on a spin bike at Equinox.
00:09:14.879 --> 00:09:21.919
Like I remember so clearly first day on the job at the bank, and I was like, I'm gonna start my career off on right foot.
00:09:22.000 -->
00:09:23.759
I'm gonna do like a 6 a.m.
00:09:23.919 --> 00:09:24.639
workout class.
00:09:25.039 --> 00:09:28.480
This is the most 2013 San Francisco story like of all time.
00:09:28.799 --> 00:09:29.679
This is actually in California.
00:09:31.440 --> 00:09:43.519
Um she was on the bike next to me, there was this like competition, and we like tied for first and uh got freaked smoothies after, and then we became like rhyme partners, and then I got you know her husband, Kevin, during that time.
00:09:44.080 --> 00:09:48.559
Um they had been like prolific angel investors for a while.
00:09:49.279 --> 00:09:54.799
And one day she was basically like, you know, we're thinking the two of them were like, we're thinking about starting this fund.
00:09:54.879 --> 00:09:56.000
Do you want to come start it with us?
00:09:56.080 --> 00:10:05.600
I was still at WTI at the time, and I was starting to kind of think through like, okay, I don't want to do debt forever, I don't want to like do this forever, what does like that potential look like?
00:10:05.679 --> 00:10:09.759
But I hadn't really kind of started to think about that.
00:10:10.080 --> 00:10:13.519
And so when they came to me with this, I was like, hell yeah, it's awesome.
00:10:14.000 --> 00:10:19.360
Um, yeah, at the time it was still just an idea, there was like no debt, we hadn't started fundraising.
00:10:19.600 --> 00:10:38.240
Um and so I think this was like I don't remember what month in 2020, but it was like kind of during COVID or right at the start of COVID, yeah, um, which was a like crazy time, you know, at first terrifying, and then ended up actually being like really great for us for raising the fund.
00:10:38.480 --> 00:10:38.639
Yeah.
00:10:38.879 --> 00:10:44.960
Um and yeah, so we closed a$50 million fund, you know, basically an extension of their angel strategy.
00:10:45.120 --> 00:10:45.279
Yeah.
00:10:45.360 --> 00:10:48.399
Um, most of the capital that was raised from outside LPs.
00:10:49.120 --> 00:11:09.039
And and so in terms of kind of building the strategy with them, or you trying to figure out your own strategy as an investor, you know, my presumption would be, and correct me if I'm wrong, they sort of said, look, like we like this angel thing, we get a lot of flow, we're in the mix, we see interesting ideas, we do the deals, they seem to go well, let's kind of scale.
00:11:09.279 --> 00:11:16.000
That doesn't maybe necessarily like put a lot of onus on them to change their approach because it was probably working.
00:11:16.080 --> 00:11:17.600
That's probably why they started the fund.
00:11:17.840 --> 00:11:21.200
You're coming in from a slightly different perspective, you know.
00:11:21.440 --> 00:11:24.799
My experience, at least with early stage venture, is like there's really no one to save you.
00:11:24.879 --> 00:11:30.320
You sort of have to figure out how to do it yourself, and um would love to understand that process from your perspective.
00:11:30.720 --> 00:11:33.919
Yeah, I mean, it was fascinating, right?
00:11:34.000 --> 00:11:42.799
Because like WTI is a firm that's been around for 40 plus years, like Maurice, David, Rudy, Dave, like they are they're on a tight ship.
00:11:42.960 --> 00:11:52.960
They are like, you know, very kind of very process oriented in terms of how they think about things around how deal flow deal memos around partnership meetings, right?
00:11:53.039 --> 00:11:59.759
And then you come into this basically completely unknown where okay, what are where are we gonna like implement processes?
00:11:59.919 --> 00:12:01.519
Where do should we cut stuff?
00:12:01.600 --> 00:12:03.679
Where can we just like move move quickly?
00:12:04.240 --> 00:12:16.159
Um, and and I think what works for an angel strategy when you're deploying your own funds doesn't necessarily translate to what when you're deploying$50 million of other people's capital, right?
00:12:16.240 --> 00:12:24.879
And like there's a lot that can be taken from that, but there's a lot of stuff that also has to start, like you start to have to create processes and guardrails around, right?
00:12:24.960 --> 00:12:30.559
And so I think there was you know the first like a lot of figuring out what that looked like, right?
00:12:30.639 --> 00:12:36.080
And Elizabeth had come, Elizabeth had worked in adventure firms before, so she also knew um knew a lot.
00:12:36.240 --> 00:12:56.879
And I think the combination of you know the the networks, the distinct networks we all had, especially Kevin's, which was incredible, you know, combined with kind of like our our venture experience, our net and and like being able to see things through these different lenses was um was great, and it was really uh really cool opportunity.
00:12:57.279 --> 00:12:57.519
Yeah.
00:12:57.840 --> 00:13:02.720
And so I mean today I read a lot of your writing, you're very prolific, it's very good.
00:13:02.879 --> 00:13:08.240
You seem to be quite thesis-driven in terms of your approach to application software investing.
00:13:08.399 --> 00:13:09.360
Was that always there?
00:13:09.600 --> 00:13:15.840
You know, you were coming from a pretty different environment by all accounts into one that probably had more of an angel structure.
00:13:15.919 --> 00:13:23.120
So, how did you think about kind of developing as an investor when you know you probably had a fair amount of on top autonomy in terms of where you could write checks?
00:13:23.440 --> 00:13:23.840
Yeah.
00:13:24.159 --> 00:13:31.039
So the the strategy with Scribble was very much like high volume, lower check size, lower ownership percentages.
00:13:31.120 --> 00:13:35.120
And so we were writing almost like a check a month, more or less.
00:13:35.360 --> 00:13:36.720
Like um that's fun.
00:13:36.879 --> 00:13:53.120
Yeah, I mean, it was it was again great from an exposure persp perspective, from a network building perspective, but it definitely one thing that I like really felt was lacking was
like that sort of intellectual stimulation, right?
00:13:53.200 --> 00:13:57.279
Like the really kind of diving super deep, and then also post-investment, right?
00:13:57.360 --> 00:14:07.120
Like, because you're not as deep, because you don't know as much about that space, like your ability to really help and dig in is is is less, right?
00:14:07.360 --> 00:14:18.240
Yeah, and so I think when I was at Scribble, I started to have these feelings of I don't know, it definitely not boredom, but like like not intellectually stimulated, right?
00:14:18.320 --> 00:14:25.039
Um and so like I would go off on my own and kind of do these like deep dives trying to understand different markets.
00:14:25.279 --> 00:14:42.559
That's actually what led me to go start Prima, which was a company I started after Scribble, was basically I started to kind of like do a deep dive on everything that was happening in Asia from well, not everything that was happening in Asia, but a lot of what was happening a lot of what was happening with like supply chain manufacturing in India.
00:14:42.799 --> 00:14:45.519
Um I had invested in whatnot.
00:14:45.600 --> 00:14:58.399
I'd led the the seed and whatnot while I was at Scribble, which was again a similar kind of situation where I was like looking at what was happening with live shopping, live commerce in Asia, specifically in China, companies like Pinduo Duo.
00:14:58.960 --> 00:15:12.399
And I think there was, you know, I think in in some ways Asia, China, even India in some cases, is like some of their trends are are especially China, is like five, ten years ahead where US is.
00:15:12.639 --> 00:15:18.080
And so was looking at that and kind of that definitely informed some of my investment decisions.
00:15:18.320 --> 00:15:20.240
But I wasn't like writing about it per se.
00:15:20.320 --> 00:15:26.320
You know, I was like here and there writing about it, but not not with like as much kind of cadence as I as I do today.
00:15:26.639 --> 00:15:28.879
Yeah, just like see the thing and then start looking for it.
00:15:29.120 --> 00:15:29.679
Yeah, exactly.
00:15:29.759 --> 00:15:37.759
And it and and it did help me because it did help me because it ultimately is what led me to say like, okay, there's an opportunity to start a company here.
00:15:38.000 --> 00:15:48.960
Yeah, you know, talk to my now my future co-founder um who was also exploring a similar space, and and that that's you know, it was it was great timing from that perspective.
00:15:49.039 --> 00:15:51.600
But it was not necessarily part of like the scribble strategy.
00:15:52.000 --> 00:15:53.519
Yeah, I mean, talk about that leap.
00:15:53.759 --> 00:16:08.559
I mean, from my own perspective, like the risk of me starting any non-investment business is quite low because the thing I like about Venture, what it gives to me anyway, is like I'm kind of endlessly curious, but I'm kind of lazy, so I sort of require founders to do all this research to keep me entertained.
00:16:08.720 --> 00:16:12.399
But um, so starting an operating business just sounds like the worst thing in the world to me.
00:16:12.559 --> 00:16:16.639
Um but uh obviously it didn't sound like the worst thing in the world to you.
00:16:16.720 --> 00:16:18.000
So so talk me through that.
00:16:18.399 --> 00:16:22.320
Yeah, I mean, I think there were like again multiple facets of of it.
00:16:22.480 --> 00:16:33.759
I think I was on the board of a couple companies, and I, you know, I was on the board of whatnot, and I was like thinking like, what how what the hell am I supposed to tell Grant about how to run his company?
00:16:33.840 --> 00:16:35.440
Like I've never been in his shoes.
00:16:35.519 --> 00:16:39.919
I don't, you know, I and to be clear, I think there's amazing investors who have never been founders or operators.
00:16:40.000 --> 00:16:52.480
But I think like for me personally, it just felt like this thing that was missing, especially as like a really early stage investor, because so much of what you're doing is like in the weeds, brute force, rolling up your sleeves with these founders.
00:16:52.559 --> 00:16:57.440
And and so it kind of felt this like, you know, how can I really help?
00:16:57.600 --> 00:17:15.759
And so so there was that thing that was happening, and then it was like again, this this idea that I had been starting to think about, and I like read the the S1 for Zometry over a weekend, which I don't like typically read S1s over the weekend, but like that Monday I had- I suppose little little flex there.
00:17:16.000 --> 00:17:32.559
No, I that it was like honestly the first thing, like it was just such a random coincidence because like that Monday I was catching up with this this founder I'd previously invested in just to catch up because he was also thinking about stepping down from his company, he was ready to kind of start something new.
00:17:32.720 --> 00:17:38.400
I had kind of told him where I was at in terms of like, you know, kind of starting to explore, maybe starting something.
00:17:38.559 --> 00:17:43.599
And so we had this call to talk through kind of where we were both at.
00:17:43.920 --> 00:17:47.440
And he's Mexican, his family comes from the steel industry.
00:17:47.680 --> 00:18:03.200
He was also seeing kind of the same analogies of what was happening in the India ecosystem from a manufacturing supply chain perspective, but he had the insight of like, hey, we can apply this to Lat Lat Dam to specifically to Mexico and and sort of build this near shoring business.
00:18:03.359 --> 00:18:12.240
I came at it from the lens of like, you know, the US side, like US business US buyers, you know, looking to kind of diversify their supply chains.
00:18:12.480 --> 00:18:20.720
We ended up talking for a long time on that phone call, kind of hung up and being like, huh, like maybe there's something here.
00:18:21.279 --> 00:18:26.319
And then within a few weeks, I was like in Mexico spending
time working alongside them.
00:18:26.400 --> 00:18:30.559
Um, it was him and two guys that he had known since business school.
00:18:31.200 --> 00:18:35.359
And yeah, one thing led to another, and I was like, this is if if not now, like when am I gonna do it?
00:18:35.440 --> 00:18:36.400
I I just cease the shop.
00:18:36.720 --> 00:18:37.519
It's still gonna get easier.
00:18:37.759 --> 00:18:38.000
Yeah.
00:18:38.160 --> 00:18:39.359
So that's exciting.
00:18:39.519 --> 00:18:39.680
Yeah.
00:18:39.920 --> 00:18:41.680
So how how long did you live in Mexico for?
00:18:41.759 --> 00:18:42.960
Or was were you back and forth?
00:18:43.200 --> 00:18:45.119
So I was back and forth for about a year and a half.
00:18:45.279 --> 00:18:46.400
Well, no, less than that.
00:18:46.480 --> 00:18:48.000
I was probably back and forth for a year.
00:18:48.160 --> 00:19:01.599
Um the plan when we started the company was for me to be based in the Bay Area running like the US business essentially, and then they would be based in in Mexico running, you know, the supply, the the Mexican side of the business, which was the operations.
00:19:01.839 --> 00:19:07.200
In in short, we were helping like large multinational companies move their supply chains to North America.
00:19:07.440 --> 00:19:19.920
Um and so most of our customers had manufacturing lines in in Asia and were either looking to fully move those over to Mexico or like set up parallel manufacturing.
00:19:20.160 --> 00:19:32.880
And so, unlike kind of the fictives and the zometries that were dealing with, like, I mean, at the time, 5K AOVs, engineer credit cards, we were really going after like heads of procurement, very large buyers, six-figure AOVs.
00:19:33.440 --> 00:19:41.839
Um within like a few weeks of starting the business, we were like, well, we can't really start selling to US customers until we have like the Mexican operations set up, right?
00:19:41.920 --> 00:19:45.119
Until we have like this, we didn't want to actually own the factories.
00:19:45.200 --> 00:19:50.640
We wanted to kind of tap into spare capacity across Mexico, although we were the tier one supplier.
00:19:51.279 --> 00:20:02.079
And so I was working on setting up the engineering team, the software engineering team, and then kind of spending like one week a month, then that became like two, three weeks a month.
00:20:02.160 --> 00:20:10.480
Because after that, I started setting up the industrial engineering team, then the supply team, and probably like six months in, the plan was for us to run the company in English.
00:20:10.559 --> 00:20:12.799
I didn't speak a word of Spanish before starting the company.
00:20:12.960 --> 00:20:19.599
I'd spoken, I had taken like French all throughout call like high school college, which was freaking a total joke.
00:20:19.839 --> 00:20:26.799
I still don't speak any French, but like six months in we would decided to kind of make the switch to speak it to run the company in Spanish.
00:20:27.039 --> 00:20:33.440
A lot of our suppliers, like their English wasn't great, and it was just like this need to kind of do more in Spanish.
00:20:33.680 --> 00:20:42.160
Um, and so I had to get fluent really fast, and so I ended up like spending a ton of time there, deciding to just like move fully down there.
00:20:42.400 --> 00:20:46.000
Um, and so I was down there for three years, I think, in total.
00:20:46.720 --> 00:20:47.039
Yeah.
00:20:47.359 --> 00:20:49.119
And what what was the technology play?
00:20:49.200 --> 00:20:54.480
And then also, uh, pardon me, but like what do you know about setting up an engineering team or an industrial engineering team?
00:20:54.799 --> 00:20:58.799
Nothing, like it was all learning on the fly.
00:20:59.839 --> 00:21:02.640
So, and having really great people to help me.
00:21:02.880 --> 00:21:04.640
Yeah, what was the technology side of the business?
00:21:04.799 --> 00:21:08.319
Yeah, so basically, like the the URL is prima.ai.
00:21:08.640 --> 00:21:21.440
So, like we had this, we had this notion of like um, you know, can you start to layer in like a this was back in 2022, I think, 2023, yeah, 2022.
00:21:21.519 --> 00:21:27.039
So it was it was still early before like really uh chat had launched when we started the company.
00:21:27.359 --> 00:21:32.960
But this idea of can you start to put in layer in technology across each leg of this supply chain, right?
00:21:33.039 --> 00:21:43.599
So essentially you have to get the RFQ, the request for quotation from the customer, you have to quote that, then you have to kind of there's a negotiation back and forth, right?
00:21:43.680 --> 00:21:45.519
That's all part of the procurement process.
00:21:45.759 --> 00:21:48.559
Then you have to source the raw materials, right?
00:21:48.640 --> 00:21:55.599
And then you have to manufacture the then there's like I'm simplifying this a lot, but like there's some sort of design for manufacturing, right?
00:21:55.680 --> 00:21:59.759
Where a lot of the drawings that they were using in Asia don't necessarily translate.
00:22:00.079 --> 00:22:04.480
Translate over to Mexican manufacturing for a lot of reasons, right?
00:22:04.559 --> 00:22:08.559
Like steel tubing diameters are slightly different in Mexico versus in Asia.
00:22:08.720 --> 00:22:10.400
Machines are slightly different.
00:22:10.559 --> 00:22:13.759
Um, and so like you have to kind of do this like DFM step.
00:22:14.000 --> 00:22:24.480
Then you manufacture the product, then you have to do kind of like process updates during that, then you have to do the quality check, then you have to do packaging, then you have to do the logistics, freight, customs, all that.
00:22:24.640 --> 00:22:26.319
So like there's a there's a lot of steps.
00:22:26.480 --> 00:22:34.160
And now in AI, there's like each leg of that has like a has a company kind of getting started to build that, right?
00:22:34.319 --> 00:22:41.440
And in India, like back in 2019, 2020, there
were companies that were starting within each leg of these supply chains, right?
00:22:41.519 --> 00:22:45.920
You had like Moglix off business that work, right, to to kind of do that pre-AI.
00:22:46.480 --> 00:22:52.480
Our idea was like, can we create this network using technology and using kind of spare capacity in Mexico?
00:22:52.720 --> 00:22:55.839
So, for example, like can we build our own quoting engine?
00:22:56.000 --> 00:23:09.680
If we know kind of the inputs and if we know what the labor costs of our suppliers are, you know, in like each each factory has different kind of inputs, can we more effectively quote such that we can save them a lot of time, right?
00:23:09.759 --> 00:23:14.480
Because a lot of these factories are quoting a hundred projects, maybe to win one.
00:23:14.720 --> 00:23:18.880
Um, they don't the the raw material procurement, right?
00:23:18.960 --> 00:23:23.279
They don't get to benefit from like order like really high volume orders.
00:23:23.359 --> 00:23:24.640
Like, can we do that, right?
00:23:24.799 --> 00:23:26.880
Working capital is a huge issue for them.
00:23:27.200 --> 00:23:30.240
Quality inspections is really hard for them to do at scale.
00:23:30.480 --> 00:23:34.799
So there's like a lot of pieces where you can start to insert technology to kind of help with those processes.
00:23:35.200 --> 00:23:37.519
But I think honestly, it's a lot.
00:23:37.839 --> 00:23:38.480
It's a lot of technology.
00:23:38.880 --> 00:23:41.119
I think we probably bit off way more than we could show.
00:23:41.200 --> 00:23:42.720
I mean, the company's still going.
00:23:42.799 --> 00:23:53.519
It's just I think I think there's also a tr um in Lat Dam, maybe I won't speak broadly for Lat Dam, although I think I I think this is a general trend there, but I'll speak speak to Mexico.
00:23:53.599 --> 00:23:59.599
I think there's a l uh a desire to do stuff with brute force as opposed to like scale it with operations with technology.
00:23:59.839 --> 00:24:02.720
And the reason for that is like pretty obvious, right?
00:24:02.880 --> 00:24:06.720
Because like human labor is a lot cheaper there than it is here.
00:24:06.960 --> 00:24:09.680
And so investing in technology.
00:24:10.160 --> 00:24:11.279
Yeah, and rates are higher.
00:24:11.440 --> 00:24:13.759
Like just so people want cash today.
00:24:14.000 --> 00:24:21.759
And so I think we ended up defaulting to starting to trying to like brute force our way through a lot of things with human labor as opposed to with real technology.
00:24:21.839 --> 00:24:26.400
I mean, I think I mean we are building real technology around that, but in the early days that's what we were doing.
00:24:27.039 --> 00:24:27.359
Yeah.
00:24:27.599 --> 00:24:31.680
And then engineering and industrial engineering team, building those.
00:24:31.759 --> 00:24:34.720
I mean, yeah, I mean, I'm not an engineer by training.
00:24:34.799 --> 00:24:51.440
I mean, I I've taken coding classes, I kind of know some basics, but I like had to rely on a lot of people around how do you structure a team, what do you look for, how do you like interview engineers, um, and then getting help for that in the early days of with advisors and things and people like that to help me.
00:24:52.000 --> 00:24:52.640
That's awesome.
00:24:52.799 --> 00:24:54.240
Well, you're nothing if not resourceful.
00:24:54.480 --> 00:24:55.519
Yeah, exactly.
00:24:55.920 --> 00:24:59.920
And then tell me about tell me about coming back to the States, starting at Zeta.
00:25:00.240 --> 00:25:11.039
You know, like actually, it was one of my first introductions to technology investing, like early stage, was actually um a podcast with uh with Ash Fontana when he was at Zeta.
00:25:11.200 --> 00:25:18.720
So, like like Zeta's like sort of like from day zero for me, always been in my mind as an interesting place to invest.
00:25:18.960 --> 00:25:28.960
I mean, you'd started a company, you'd started a venture fund, you know, going to work at someone else's venture fund with an interesting history, like maybe it was not the most obvious next step.
00:25:29.119 --> 00:25:36.400
I mean, you meet Mark, you meet Dylan, and like who wouldn't want to work with them, but um just curious to understand kind of that process and what drew you back.
00:25:36.640 --> 00:25:36.960
Yeah.
00:25:37.200 --> 00:25:53.119
So I was taking a little time off traveling, and an LP actually from Scribble connected because my plan was take some time off, travel, kind of like restore myself, and then come back to San Francisco in January and like start to think through what I wanted to do next.
00:25:53.279 --> 00:26:04.079
I was like having some conversations with folks, and then some of those conversations started to pick up, and actually an LP in Scribble connected me to was like you have to meet this guy a Porv.
00:26:04.400 --> 00:26:12.400
So I had a call with a Porv actually when I was in Argentina, and I don't just like really bought like really liked him.
00:26:12.480 --> 00:26:15.680
Yeah, um, I don't you've met a Porv, so you know how he's actually haven't met the Porv.
00:26:15.839 --> 00:26:21.359
If you haven't, he's amazing and yeah, just like really clicked.
00:26:21.599 --> 00:26:27.359
You know, it was like a I don't know, 45-minute call probably, and but liked him a lot.
00:26:28.000 --> 00:26:29.279
And then I met Jocelyn.
00:26:29.519 --> 00:26:33.440
Um, again, another call, really liked her, kind of liked the story.
00:26:33.519 --> 00:26:37.519
Um, and then flew back to San Francisco and spent some time with the team, got to know them.
00:26:37.839 --> 00:26:45.279
And I think what drew me to them was like how different their backgrounds all were, not only from from mine, but from each other's, right?
00:26:45.359 --> 00:26:48.480
Like they all brought such a unique, different lens.
00:26:48.880 --> 00:26:51.359
You know, Mark is like the OG, right?
00:26:51.440 --> 00:26:57.920
He started
this firm in 2013 with the thesis that software is getting commoditized and the next great wave of companies are gonna be AI companies.
00:26:58.000 --> 00:27:05.119
And like, I mean, that was in 2013, which is like mind-blowing, you know, and you know, his career has just been incredible, right?
00:27:05.279 --> 00:27:09.759
Like he did, he was worked under Obama administration for AI policy.
00:27:10.000 --> 00:27:12.160
He's chairman of the board of MIT now.
00:27:12.319 --> 00:27:16.079
Uh, and he's still like that's his full-time job is still Zeta.
00:27:16.319 --> 00:27:18.559
Like, I don't know how he does it all, and he's in his 70s.
00:27:18.640 --> 00:27:19.039
It's crazy.
00:27:19.599 --> 00:27:20.960
And he ran a marathon the other day.
00:27:21.119 --> 00:27:25.200
Oh, and he's run like I don't even know how many like hundreds of marathons by this point.
00:27:25.440 --> 00:27:25.680
Yeah.
00:27:26.240 --> 00:27:37.680
Um, and then Jocelyn, right, she was like the first engineer at this tiny startup called VMware back in the day, and then became GM of that business, then was like head of engineering at Meta.
00:27:38.000 --> 00:27:40.480
She was the one that implemented ML on the news feed.
00:27:40.799 --> 00:27:43.200
Diane Green was one that connected her to Mark.
00:27:43.279 --> 00:27:51.119
So, like they just everybody has like these very fascinating, unique perspectives in terms of like technology, in terms of AI.
00:27:51.279 --> 00:27:56.400
You know, Dylan had started a company and actually also in the manufacturing space before joining Zeta.
00:27:56.720 --> 00:28:06.960
Kaporv had worn a ton of different hats, started companies, been in venture, been like Ted's like senior at sort of on go-to-market teams across a number of companies, engineer by training.
00:28:07.200 --> 00:28:11.119
So I just thought like they complement kind of my skill sets really well.
00:28:11.200 --> 00:28:12.880
I think I can learn a ton from them, right?
00:28:12.960 --> 00:28:21.759
Like I every time I talked to them, I felt like so stupid after I hung up, like, oh my gosh, what like writing notes down frantically to be like, okay, I have to look this up afterwards.
00:28:22.079 --> 00:28:25.440
I'm very, very familiar with that feeling every time I get together with Dylan.
00:28:25.680 --> 00:28:27.200
Yeah, it's crazy.
00:28:27.359 --> 00:28:29.920
Um I mean, Dylan is just like brilliant.
00:28:30.079 --> 00:28:31.599
I learn something from him every day.
00:28:31.759 --> 00:28:37.599
And so yeah, I think just and then also the fact that there's like zero ego, right?
00:28:37.680 --> 00:28:43.039
There's like it's very rare to find a firm where there's just no drama at all.
00:28:43.200 --> 00:28:47.119
And everyone just kind of like everyone has their majors and minors, focuses.
00:28:47.440 --> 00:28:49.279
It is very thesis driven, right?
00:28:49.359 --> 00:28:56.960
Um so each partner does probably two to three deals a year, so low volume, higher check sizes, higher ownership percentages.
00:28:57.279 --> 00:29:04.000
But the theory is like when you meet a company, you you know a lot about that company, a lot about that space already, right?
00:29:04.160 --> 00:29:09.359
Um it just allows you to go so much deeper, it's more enjoyable.
00:29:09.440 --> 00:29:15.599
Um, and then like your ability to kind of help and work with these teams post-investment is also, I think, a lot higher.
00:29:15.920 --> 00:29:16.400
Yeah.
00:29:16.720 --> 00:29:23.920
And so talk about now, like we're sort of in the we're in Annalise Prime, we're in the like the apex state.
00:29:24.079 --> 00:29:31.200
You're doing the thing you're doing now, building a thesis, focusing on application software, which historically is not something Zeta had done.
00:29:31.359 --> 00:29:35.920
So, like how how have you thought about carving out your niche or your domain?
00:29:36.000 --> 00:29:38.400
And uh, you know, how do you think about investing today?
00:29:38.799 --> 00:29:43.039
Yeah, I I mean actually I think I think Zeta has done a fair amount of application stuff.
00:29:43.119 --> 00:29:46.400
I think they they really started to do that a lot in fund three.
00:29:46.559 --> 00:29:49.200
They have a number of sort of application layer investments.
00:29:49.359 --> 00:29:58.000
I think though there was like a really good fit with me and them because there's you know, I was going to kind of solely look at vertical AI application layer stuff.
00:29:58.160 --> 00:30:03.440
And then obviously my lens comes at it through like this I this legacy industries, right?
00:30:03.519 --> 00:30:10.160
And yeah, you know, I don't just cover supply chain, manufacturing, logistics, I cover like lots of different verticals.
00:30:10.240 --> 00:30:14.799
I mean insurance, uh any kind of regulated industry as well.
00:30:15.039 --> 00:30:23.359
Yeah, but they all have like parallels and similarities, and I think it's helpful to have like been there, like seen these problems firsthand.
00:30:23.440 --> 00:30:27.440
And even though I haven't, you know, I've never worked in insurance, right?
00:30:27.519 --> 00:30:36.400
But I've started to do a lot there, I've made an insurance investment, like there's there's a lot of um overlapping kind of trends, right?
00:30:36.720 --> 00:30:44.160
Um and so yeah, that's kind of like where I'm I'm focused right now, where I see a lot of opportunity.
00:30:44.319 --> 00:30:58.240
I mean, you look at like the top 50, 60 public companies, the first five are like hyperscalers, but the long like the next 50, 55 are are kind of like SAS era companies.
00:30:58.400 --> 00:31:04.079
And I think that those companies are gonna start to get replaced with like AI application layer companies.
00:31:04.160 --> 00:31:11.599
And I I I'm really bullish on like vertical AI application layer being sort of the next where the value is gonna start to accrue.
00:31:12.079 --> 00:31:27.839
And and when you think
about what that means practically, are you like I've got the Fortune 50, like take out the hyperscalers and be like, okay, number eight is Broadcom or whatever, and and and start, and start like from there, or how do you how should I think about it practically?
00:31:28.160 --> 00:31:30.960
Yeah, no, I I I it's not I don't think it's like that.
00:31:31.119 --> 00:31:42.319
I think it's more I mean I I think in venture there's this constant push and pull of like I'm interested in a space, so I'll go like research it, go really deep, and then try and find companies in and around that space.
00:31:42.559 --> 00:31:48.160
And that might be because of like a personal connection to the space, something I've read recently, heard recently, right?
00:31:48.400 --> 00:31:53.839
Or I might meet a company, I might be like, oh, that's a really interesting, I've never I thought about it from that angle.
00:31:53.920 --> 00:31:56.000
Let me go really, really deep into that, right?
00:31:56.160 --> 00:31:58.240
That's actually sort of what happened with the insurance.
00:31:58.319 --> 00:32:05.200
You know, I met a company building like an insurance claims like AI native TPA, and then that sent me down this like very long rabbit hole.
00:32:05.839 --> 00:32:17.279
But I, you know, I I think we're like at this point right now where companies that haven't that kind of haven't been able to innovate for a very, very long time have been using these kind of static systems of record, right?
00:32:17.440 --> 00:32:22.319
Are now all starting to kind of like be turned on their head in terms of how they do operations, right?
00:32:22.559 --> 00:32:29.440
Bestemer, I think, came out this really great piece a few months ago around like systems of action to replace systems of record.
00:32:29.839 --> 00:32:44.480
Um so if you think about what the sort of what that means, like it's a lot of industries that might have been like skipped in the sort of first SAS wave of the last like 10 years, right?
00:32:44.559 --> 00:32:49.119
And maybe still using software from like the 90s, let's say, right?
00:32:49.359 --> 00:32:50.720
Yeah, it's all it's all kind of crud.
00:32:51.119 --> 00:32:51.839
Yeah, exactly.
00:32:52.000 --> 00:32:57.759
They're they're now having this ability, this, there's this opportunity to kind of build for them, right?
00:32:57.839 --> 00:33:03.519
Whether that be building up an entirely new stack from the ground up or building sort of an orchestration layer on top.
00:33:03.759 --> 00:33:14.880
Um, and so for me, I'm trying to find like markets that, you know, they might be very different, but they have those same underlying kind of similarities, right?
00:33:14.960 --> 00:33:28.960
And so again, like insurance, for example, whether that be underwriting, whether that be claims uh or freight forwarding or you know, logistics fleets, trucking fleets, or things like that.
00:33:29.440 --> 00:33:29.839
Yeah.
00:33:30.160 --> 00:33:37.119
And when you think about this kind of like AI cycle, there's obviously a lot of hype, but sort of ludicrous amounts of substance.
00:33:37.279 --> 00:33:50.400
Um I was listening to an interview with Martin from A16Z the other day, and I think he was being provocative and also just being like an intense infrastructure investor, but saying, like, you know, there is no technology in a vertical software company, like it's all go-to-market, it's all understanding the workflow.
00:33:50.559 --> 00:33:54.799
Um, you know, I'm sure you've been pitched as I have, you know, 25.
00:33:55.119 --> 00:34:01.839
Oh, we're doing we're doing this thing for AI, and the founder has like no background in AI, completely unable to recruit.
00:34:02.000 --> 00:34:03.920
I mean, maybe the business will succeed, who knows?
00:34:04.000 --> 00:34:06.240
But like there's certainly not a genuine AI angle.
00:34:06.400 --> 00:34:09.199
How do you like how do you think about that?
00:34:09.360 --> 00:34:11.360
Like, what's what does your process end up looking like?
00:34:11.519 --> 00:34:16.320
How oriented are you on kind of proprietary data, proprietary models?
00:34:16.480 --> 00:34:18.079
Like, you know, what what's your process there?
00:34:18.400 --> 00:34:20.480
Yeah, oh, so many stories about this.
00:34:20.639 --> 00:34:25.199
I I went to um an insurance conference called Target Markets a couple months ago in Dallas.
00:34:25.440 --> 00:34:27.840
It's like mostly focused on underwriting.
00:34:28.000 --> 00:34:32.159
But yeah, there was a room of like 30 startups, 30 AI startups, right?
00:34:32.239 --> 00:34:36.960
And and you meet these companies in there, and like they're not AI startups, right?
00:34:37.039 --> 00:34:39.280
They're like consulting services businesses, right?
00:34:39.360 --> 00:34:47.920
They're just yeah, and and then you talk to like buyers at these conferences and they kind of loop like kind of think all of them are the same, right?
00:34:48.079 --> 00:34:50.800
Even the ones that are really building true technology.
00:34:51.039 --> 00:34:56.159
Um, and I talked to a founder the other day who's also said something that really worried me.
00:34:56.320 --> 00:35:03.920
He was basically like, oh, you know, like the technology is somewhat commoditized right now, so it's really around like having that industry wedge and the go-to-market wedge.
00:35:04.000 --> 00:35:15.599
And it's like, no, no, not at all, you know, like we're just at the very beginning, and you know, I I know it's trite to say at this point, I think, but you do have to constantly like be building at the edge of these model capabilities, right?
00:35:15.679 --> 00:35:17.920
Because like they're continually getting better.
00:35:18.079 --> 00:35:25.440
You know, all of these frontier labs, open AI is building up their forward-deployed engineering team, like they're all gonna start going after these like verticals.
00:35:25.519 --> 00:35:33.360
And so I think it's you know, when I think about what
differentiates you know, vertical teams, it's really kind of like that feedback loop, right?
00:35:33.440 --> 00:35:35.360
This continuous learning feedback loop.
00:35:35.519 --> 00:35:41.920
I talked to Rebecca Caden from USV yesterday, and she she puts it as like network effects, which I I love that, right?
00:35:42.000 --> 00:35:49.440
But it's like the more you the more you use the product, the better it gets, which is like the way I think about it is you know that continuous feedback loop, right?
00:35:49.519 --> 00:36:01.440
Where you, you know, whether that be through RL or um supervised fine-tuning, probably some combination of both, but like you're not necessarily building your own models, but you're tweaking them, you're refining them on top of that.
00:36:01.599 --> 00:36:04.880
And you and I think also the unique data piece of it too, right?
00:36:04.960 --> 00:36:15.440
Like um, that's you know, I think there's a big opportunity there, but it's yeah, it's not, I don't think, commoditized by any means.
00:36:15.760 --> 00:36:16.960
The technology piece of it.
00:36:17.679 --> 00:36:29.519
How do you characterize the things that you think will beat open AI with its like in anthropic and open AI with its inevitable team of um sort of ex-McKinsey consultant turned forward-deployed engineers?
00:36:29.840 --> 00:36:30.480
Yeah.
00:36:30.719 --> 00:36:50.800
I mean, I think like the they are massive, they have massive amounts of budgets, and I think like if they turn their attention on something like watch out, because if they want to, they can destroy, they can c they can like win in whatever vertical they really want to win in, is is sort of how I think about it.
00:36:50.960 --> 00:36:57.360
But I also think about like a company of that scale is going to where are they gonna be wanting to dedicate their resources?
00:36:57.440 --> 00:37:01.039
It's going to be on like the the really big customers, right?
00:37:01.199 --> 00:37:05.920
Yeah, and it's gonna be on solving 80% of the problems for 80% of the population, right?
00:37:06.079 --> 00:37:21.760
And so that leaves open a big opportunity where like startups can find these wedges, find these kind of oper these openings to kind of insert themselves in, build this, build modules that you know might not seem obvious at first, right?
00:37:21.920 --> 00:37:23.840
And then and then grow from in there, right?
00:37:23.920 --> 00:37:32.320
And so thinking through like what is that right entry point, what data is that gonna give me, and what does that unlock beyond like this initial wedge, right?
00:37:32.400 --> 00:37:37.760
And so I think it's fine if the initial wedge doesn't seem particularly like massive, right?
00:37:38.000 --> 00:37:42.719
But what's key is understanding like what does that wedge wedge you into, right?
00:37:42.800 --> 00:37:56.320
And that has to be massive, and that also has that wedge has to give you a leg up from like a data perspective or from this feedback perspective, or from like really sort of ingraining yourself into these workflows, right?
00:37:56.559 --> 00:37:59.360
Um, that will make it harder for others to catch up.
00:37:59.599 --> 00:38:13.039
I mean, that's another thing is like you know, you look at the four deployed teams at at OpenAI, they're probably not gonna be sitting next to, you know, some mid-market freight broke freight forwarder, right?
00:38:13.119 --> 00:38:14.880
They're probably not gonna That feels unlikely.
00:38:15.039 --> 00:38:15.599
Yeah, right.
00:38:15.679 --> 00:38:20.159
They're gonna be going after like I don't know what size, but that's just my sense, right?
00:38:20.239 --> 00:38:30.880
But and that's not to say like, okay, it's just startup opportunities at the mid-market level, but I think it does that's just an example of like figuring out where you can wedge your way in.
00:38:33.599 --> 00:38:34.159
That was me.
00:38:34.400 --> 00:38:35.840
That was me wedging my way in.
00:38:36.239 --> 00:38:36.480
Okay.
00:38:36.559 --> 00:38:37.440
Oh, there you go.
00:38:37.679 --> 00:38:42.239
Um tremendous, uh, tremendous theatrical ability on display here.
00:38:42.559 --> 00:38:46.159
Um what's uh what are the what are the things about teams that stand out to you?
00:38:46.320 --> 00:38:47.360
What do you look for?
00:38:47.920 --> 00:38:55.039
I love when a team someone comes from industry and someone comes from technology.
00:38:55.440 --> 00:39:00.559
However, I think that there's like a balance.
00:39:00.800 --> 00:39:37.360
Like I've talked to some teams where like the industry person is 25 years in the industry, they were you know chief ex officer at five different companies, they are like they know this like the back of their hand, and then they're partnered with like some very senior AI engineer, and like that does fit the build of like you have an industry expert and you have this like really great tech tech person, but I think you know, sometimes you then run into the fact of like that person is probably like wed to a lot of processes and a lot of ways of doing things that isn't like novel.
00:39:37.440 --> 00:39:44.159
And so I think that the balance for me is constantly like what does that right experience level look like, right?
00:39:44.239 --> 00:39:55.039
Where they're gonna come into this with a fresh perspective of like this is the possibility, as opposed to being like wed to kind of how systems have been done previously like historically.
00:39:55.519 --> 00:39:56.960
What's the right level of context?
00:39:57.199 --> 00:39:58.320
Yeah, yeah, yeah.
00:40:00.079 --> 00:40:08.880
Beyond like technology and ways to make money, um, what things about AI kind of or to extent they do keep you up at night um in terms of societal impacts?
00:40:09.360 --> 00:40:22.639
I think with every technology, there is like you can point to kind of how this could be bad for society, how this could like really hurt society.
00:40:23.440 --> 00:40:26.800
And I
guess I think less about that with AI.
00:40:26.880 --> 00:40:30.480
I probably should think about it more, but I'm more excited about like all of the opportunity.
00:40:30.559 --> 00:40:40.880
And I'm honestly, you know, not as worried about like job replacement because I think it's just gonna level up people and they're gonna figure out like new roles, you know.
00:40:41.199 --> 00:40:45.599
I one thing I think about though is like, am I getting dumber, right?
00:40:45.760 --> 00:40:48.639
Like, because of how much I rely on AI now.
00:40:48.800 --> 00:40:57.280
I think there was a study done where it's like even just six months of using AI, you know, like they show your brain gray matter actually decreases.
00:40:57.440 --> 00:40:58.000
Yeah.
00:40:58.480 --> 00:41:02.480
It's like shit, what is my how is my gray matter doing?
00:41:02.639 --> 00:41:06.559
Um, you must have been pretty smart six months ago when we hung out a few weeks ago.
00:41:06.880 --> 00:41:11.440
Like if uh So I mean that is something I think about probably a lot.
00:41:11.519 --> 00:41:15.519
It's just like the intelligence level of society as a whole.
00:41:15.920 --> 00:41:16.320
Yeah.
00:41:16.559 --> 00:41:24.159
Yeah, I think like, I mean, I probably should worry more about societal displacement because I mean Salesforce did acts like half of its workforce last week or something.
00:41:24.239 --> 00:41:25.519
It's like pretty wild.
00:41:25.679 --> 00:41:33.840
But um I I do frankly worry more that like we're gonna just have be like a society of Philistines because people will like never read anything challenging.
00:41:34.000 --> 00:41:34.320
Yeah.
00:41:34.480 --> 00:41:50.159
Um, because they'll just get the kind of open AI version of Virgil, and that's like that's not really that like doesn't create enough kind of cultural context for us to I think generate more art and like be stimulated by creativity, and I think anyway, it's sort of sad.
00:41:50.480 --> 00:41:54.960
On that note, let's move to uh rapid fire questions if you're ready.
00:41:55.280 --> 00:41:56.079
Perfect.
00:41:56.400 --> 00:41:56.719
Okay.
00:41:57.039 --> 00:41:58.559
Who impresses you?
00:41:59.360 --> 00:42:01.039
I don't know, a lot of people impress me.
00:42:01.119 --> 00:42:06.639
I think people who like have really mastered a skill set or mastered an industry.
00:42:06.800 --> 00:42:13.360
You know, I actually listened to an interview with um Todd Combs, CEO of Geico recently, and was like very impressed.
00:42:13.519 --> 00:42:20.000
I mean, he started out as a financial analyst, I think at Progressive, worked his way up, worked at um Berkshire Hathaways.
00:42:20.079 --> 00:42:23.599
I think he cold cold emailed them, which is how he got the job.
00:42:23.760 --> 00:42:25.039
Now he's CEO of Geico.
00:42:25.280 --> 00:42:30.480
He's just like brilliant and he knows the problem of insurance from literally all angles.
00:42:30.719 --> 00:42:34.239
And I think that's yeah, I find that really impressive.
00:42:34.559 --> 00:42:38.880
Yeah, mastery I think is extremely, yeah, extremely charismatic just in and of itself.
00:42:39.039 --> 00:42:39.280
Yeah.
00:42:40.000 --> 00:42:45.440
What piece of media, book, podcast, movie, article, essay have you gifted or shared the most?
00:42:46.159 --> 00:42:48.000
I love Lenny's podcast.
00:42:48.960 --> 00:42:53.039
And Pieces Every Step by Tiknyat Khan.
00:42:53.519 --> 00:42:54.159
Amazing book.
00:42:54.480 --> 00:42:55.599
Um, I've gifted that a lot.
00:42:56.639 --> 00:43:03.760
It's just very short vignettes of like how to incorporate mindfulness in like your day-to-day routine.
00:43:03.840 --> 00:43:10.400
And so you can literally read the book front to back, or you can just read one little vignette for five minutes.
00:43:10.639 --> 00:43:16.719
And it's like how to practice mindfulness when you're washing the dishes, when you're in a subway station.
00:43:16.880 --> 00:43:22.000
Like, I remember reading it, I think, my uh summer in college, and just like it really having an impact on me.
00:43:22.159 --> 00:43:26.159
And I still like reference it a lot, think about it a lot.
00:43:26.400 --> 00:43:26.639
Yeah.
00:43:27.199 --> 00:43:27.519
Yeah.
00:43:27.760 --> 00:43:35.840
I I love books that you can kind of like plunge into for a I don't know, 10 minutes, five minutes, like um uh All Trivia by Logan Pirschell Smith.
00:43:36.000 --> 00:43:36.880
I feel like is one of those.
00:43:36.960 --> 00:43:48.880
It's um it's it's such a gift when like an author is so um they distill something interesting or profound or charming in like such a dense amount of text, and you can kind of jump in and get so nice.
00:43:49.280 --> 00:43:52.880
What skill do you spend the most time developing, work or non work related?
00:43:53.599 --> 00:43:54.239
Spanish.
00:43:54.400 --> 00:43:57.360
So I'm still like actively studying Spanish.
00:43:57.519 --> 00:43:59.760
I think I'm like right on the brink of being like fluent.
00:44:00.320 --> 00:44:10.239
But like every night we'll I'll watch a show in Spanish with Spanish subtitles, and like I'm now at the point where I just really love learning Spanish, you know, like I'm reading books in Spanish.
00:44:10.719 --> 00:44:13.920
But it's just like a constant requires work.
00:44:14.159 --> 00:44:19.119
And then I think coding is something too spending a lot of time on.
00:44:20.320 --> 00:44:22.960
We'll have to exchange vibe coding stories next week.
00:44:23.840 --> 00:44:25.920
Who's influenced your career the most?
00:44:27.599 --> 00:44:30.320
Probably probably my dad, to be honest.
00:44:30.480 --> 00:44:31.840
I mean, he he's an entrepreneur.
00:44:32.000 --> 00:44:38.880
He's like he's just been a sounding board for me throughout my career and helps helped me through through like really big decisions.
00:44:39.280 --> 00:44:47.360
And he's just like really supportive and like believes in me, which you know he's never like told me.
00:44:47.519 --> 00:44:49.760
Yeah, he's just very supportive.
00:44:50.079 --> 00:44:50.800
That's awesome.
00:44:50.880 --> 00:44:50.960
Yeah.
00:44:51.119 --> 00:44:54.079
That's I think it's one of the kindest things you can do for somebody to believe in them.
00:44:54.480 --> 00:44:56.159
Um it's wonderful, that's your dad.
00:44:56.480 --> 00:44:56.960
I love that.
00:44:57.039 --> 00:44:58.159
That's a great place to end.
00:44:58.320 --> 00:45:03.280
How can how can my millions of followers follow your writing and doings?
00:45:04.079 --> 00:45:05.360
So I'm on Substack.
00:45:05.440 --> 00:45:08.719
I write a Substack every week called Vertical AI Opportunities.
00:45:09.039 --> 00:45:13.760
Usually it's deep dives on specific verticals on where I see opportunities for for startups.
00:45:13.920 --> 00:45:20.320
Um sometimes it's on just like, you know, that continuous feedback learning or like how to think about design partners.
00:45:20.559 --> 00:45:24.400
So that's one way, and then Twitter on Elise Gamble.
00:45:25.039 --> 00:45:25.440
Love it.
00:45:25.599 --> 00:45:27.119
I heartily commend both to you.
00:45:27.280 --> 00:45:29.280
Thanks so much for being with me today.
00:45:29.760 --> 00:45:34.159
And um, I will look forward to seeing you on location in SF very soon.
00:45:34.400 --> 00:45:34.559
Yes.
00:45:34.880 --> 00:45:35.280
Very cool.
00:45:35.440 --> 00:45:35.599
Yes.
00:45:35.840 --> 00:45:36.079
All right.
00:45:36.239 --> 00:45:37.199
Thanks for having me.
00:45:39.119 --> 00:45:42.480
Hello again, friends, and thank you very much for listening till the end.
00:45:42.639 --> 00:45:44.480
I hope you enjoyed the conversation.
00:45:44.639 --> 00:45:46.159
And don't be a stranger.
00:45:46.320 --> 00:45:49.039
Let me know what you thought or if you have any follow up questions.
00:45:49.119 --> 00:45:51.119
I'll be happy to try and get them answered.
00:45:51.360 --> 00:45:54.960
See you down the road for our next episode of Edge Cases.