Episode Transcript
[00:00:01] Speaker A: All right. Was that our theme music now?
[00:00:03] Speaker B: Yeah, we're live.
[00:00:06] Speaker A: Oh, I love it. Okay. All right, here you go. So, hey, welcome to the Greg and Paul show.
[00:00:12] Speaker B: I'm Paul.
[00:00:13] Speaker A: And I'm Gregory. And we break down the Latest in startups, SaaS, AI, whatever the Internet's debating. We aim for smart takes, but hey, we end up with a lot of dumb ones. We'll definitely hit on a few scorchers. And of course we love memes way too much.
[00:00:29] Speaker B: We're going to stream live on X and LinkedIn, then straight upload into YouTube.
[00:00:36] Speaker A: And then of course, we have a guest today. So Imran Patel is the founder of Sift Data, a Los Angeles based startup that takes the guesswork out of B2B marketing by enabling teams with people level and said signals from web traffic. So prior Sift you spent eight years at Snap. Correct. Head of engineering growth.
[00:00:58] Speaker C: Long time.
[00:00:59] Speaker B: Nice.
[00:01:00] Speaker A: Awesome. Yeah, eight years is a long time to get a company like Snap. That must have been actually pretty exciting.
[00:01:05] Speaker C: Yeah, we were like the company grew over 100x I guess in terms of its size while me and my co founder were there.
[00:01:13] Speaker A: Awesome. Oh yeah, I gotta keep going. Okay. Full bio. Okay.
You did Snap. Awesome experience. And then you're at AWS for six years in Bellevue, Seattle area.
That's exactly. And I got this from your LinkedIn. You hold a PhD in computer science from UC Santa Barbara and is the inventor of multiple CDN routing patents. That's correct, Yes.
[00:01:40] Speaker C: A bunch of them, yeah.
[00:01:42] Speaker A: Dude. What I took away from that was like. You like California?
[00:01:47] Speaker C: Yeah, I had a small break. I went to Seattle and then like I kind of, you know, I was there for, for a while, but then I, I think I, I came back south again.
[00:01:58] Speaker A: Yeah, it's. It, we had a nice couple of weeks, summer, and already it's a little, it's a little gray and then it might even rain tomorrow. And I wanted to do a nice big bike ride for, for my birthday, but I think that is now on hold. But that's okay.
[00:02:14] Speaker C: That's one thing about Seattle is just that like you have amazing days that you get, but then they don't fall on weekends.
[00:02:22] Speaker A: Yeah, exactly. I get to celebrate on the bus or something.
[00:02:25] Speaker C: Right.
[00:02:27] Speaker A: All right. All right. So we don't have Imran for the entire show today, so we're gonna change up our schedule. And nerdy news that's gonna follow our AI first marketing segment. And usually we talk about how machines do things that we're just too lazy to do, but since we have Imran here, he has a really interesting perspective on AI and marketing and the future of like let's call it even maybe vibe marketing.
[00:03:01] Speaker C: Yeah, like that's like, that's a, that's a big, you know, it's a big trend right now and I think we started noticing it.
We use AI quite a bit.
And this is like, you know, I think Paul pulled up this, this, this article, I think this was on Kyle's substack. But this article kind of captures the, the, the idea very well which is, you know, the cost of like building software is kind of really going to zero. And I think a lot of marketers, for a lot of marketers, like you know, it's becoming more accessible to kind of build out like you know, pieces of software and mostly for internal workflows is what we have seen people use an attend and just a bunch of other tools.
But I think what another thing that we are seeing and we have ourselves started doing this as somebody who has people who do founder led marketing is can you build software that's more customer facing but it's in the service of gtm, right? And so a classic example is for example, I think one of this, this article mentions ROI calculators. If you like, marketers love these like tools like ROI calculators and a bunch of other things because they know the value of those tools in bringing like, you know, leads and like building relationships with prospects. But then historically if you ask the marketer like you know, do you get any, you have all these crazy ideas about what kind of tools to build. But does your engineering team even have bandwidth or you know that they can give you to kind of build it and they haven't. Right. Like marketers don't have access to engineering, you know, the priorities like you know, kind of it gets buried someplace below some product feature or whatever it is. And I think we are kind of seeing this shift where wide coding is making you know, vibe marketing more powerful. And I think one big thing that marketers can use now with wipe coding is like build this more customer facing lead magnets and more complicated ones as well. Like don't stop just at an ROI calculator. You can actually build out like a freemium tool. And so like just give you an idea like I'm not a, I'm not a, I'm a lab scorer. Right? Like that's what I call myself. But I was able to develop a lead magnet for our own company as well. And I'll kind of share it like on like I have lost the window at this point, or my computer is kind of acting up here. Yeah. So why do you put it in the private chat right there?
[00:05:47] Speaker B: Yeah, sure, I'll pull that up.
[00:05:49] Speaker A: Yeah, okay, cool.
[00:05:50] Speaker C: If you want to put it. Paul, if you want to put it in the chat. But like, this is, this is a wipe coding tool, that wipe coded tool that we built.
And this is like a very small piece of SIF's functionality that we just are giving away for free. For example, it's a Chrome extension. It actually automates this comment for reach kind of playbook that people use on LinkedIn quite a bit nowadays. So you can kind of use to reach out to people who want to get your content.
So anyway, this was built like in like, you know, maybe a couple of days. It was like, you know, some polishing and so on. But the main idea is that you can use these kind of tools now or use white coding tools. This was built mostly with, I think, Claude code, and the website was built with Claude code as well.
And I think this is the future.
Like me, like, marketers will be able to build this, like, you know, lightweight products or tools that are adjacent to their main product and like, do these little experiments, right? Like you kind of give it away, this like free tools to your audience.
And again, they're adjacent, they're not like the main product, right?
And the reason why this can happen now is, is because software is incredibly easy to produce now.
Also, software has recurring value, right? Like, if you produce like a content, like a white paper or something like that, people read it once and then it's kind of done.
But if you kind of build out a tool like, you know, it just gives recurring value to people.
The second thing is that you don't have to commit to some like, heavyweight PLG strategy for this, right? Like, you don't have to come in and say, like, oh, we sell this big product and we have to kind of miniaturize it. And then we had to kind of price and package it and then do like the whole PLG dance. Like you could just give, like you build something very small. You don't have to kind of miniaturize your main product line or whatever it is. You don't have to come up with like pricing and packaging kind of conversations and just like, you know, just, just give this away for free. So I guess that's kind of the second benefit of this.
And then the third thing which ties into AI, but from another angle from, from my perspective is me and Greg have talked about this and we see this across a lot of our customers is chatgpt and perplexity and all those LLMs are sending a lot of traffic now to websites.
And what we are seeing, and I have a few LinkedIn posts on my profile about this is that this traffic is not only growing pretty fast, but this traffic is also super high intent and bottom of the funnel. Like as in if somebody comes from chatgpt to your website they know a lot about you because they have been researching you in ChatGPT and so by the time they come to your website they're super well educated, they're pretty much bottom of the funnel, high intent. There are like studies or there are like War Cell just actually shared and Stripe shared some of their numbers like for Vercel, which is like one of the very big like you know, stack hostings providers, the creators of Next JS, 10% of their signups come from ChatGPT. Now this number was 5% two months ago and it was less than 1% six months prior to that. So there's like a curve there, like an exponential that's going on.
But the more important thing is what they say is that the people who sign up once they come from ChatGPT they're, they have high intent, they stick around longer and so, so, so now what can you do as a marketer for this new world? Like you know, like people are coming from Chat GPT, there are more of them going to come from.
If they land on your website you better convert them. Like you know, it's like sort of and, and if you have like some of these lead magnets that then there's more of a higher probability for them for you to convert them. Right to give them something that establishes that first party relationship between you and them. And then once you have that relationship then you can obviously sell them on the longer term horizon on your full product or full services.
So anyways those are the three things that I kind of, I think I'll just make this like this trend towards like white coded software as a lead magnet.
You know, software as a lead magnet strategy like be very viable going forward.
[00:10:35] Speaker A: Yeah, yeah, we, I mean it made a lot of great points.
All things that we've discussed here.
I want to go back to like the idea that resources are now, let's call them abundant and I've been pushing this with people a lot that like Lean Start and I got roasted on Edit Red all the time. Like Lean Startup is based on the idea that resources are scarce and they're not and So I think that like changes everything. Like I'm 100 with you that like vibe coded tools and marketers creating software as like lead magnets. Like, there's a whole new world out there. So given like all of that, like, what do you recommend for like, let's say early stage SAS indie stats? Like, people out there, they're trying to take advantage of stuff and maybe don't know exactly where, where to start. Like, what would you recommend that they do other than use your, your tool to do the reply for lead gen magnet, which I did like, and I did use it.
[00:11:40] Speaker B: Yeah.
[00:11:41] Speaker A: What other advice do you have for them?
[00:11:43] Speaker C: Yeah, I mean, I would say like, you know, for marketers, I would, I would.
I think there is the idea of just if you haven't done any wipe coding or whatever, right. Like, then maybe like start with internal tools first. Right. Like some internal workflows.
There are tools that are sort of, you know, somewhat like low code, no code kind of things, like some workflows that you can build out and so you can, you can go do that.
But if you want to kind of build out like lead magnets and the article that I shared, it actually talks about, talks about some of the tools there, but I would say go play with like, you know, a replit or lovable. I think those two are very accessible for marketers.
And start with something super simple. Like, you know, take something that is like, you know, an ROI calculator is a very good, you know, hello world kind of a program that you can kind of, you know, a service that you can build. You know, I know every marketer, you know, wants an ROI calculator or something on their website. So I would start there.
There is also like some other ideas. Like you might have some interesting data set that you might want to kind of share with people. But like, instead of just dumping it into like the giant report, like try to make like some sort of a nice application out of it that people can kind of, you know, play with. Right.
So I guess that would be my other kind of suggestion in terms of like, you know, where to start. But, but I would say like start with something that, that, that you are already familiar with. Yeah. Oh, that's flappy nerds. Yeah, you could build flappy nerds. That's an amazing lead magnet that definitely will give you a lot of engagement.
[00:13:23] Speaker A: That was my vibe coated game. I agree 100%. I think, I think like marketers getting in and like playing with tools and just trying to get a sense of how they work and so yeah, I made this, this fun game. I challenged myself on Sunday. I was like, dude, can I do?
Wasn't that hard. In fact, what I tell everyone is it told me even how to deploy this on Squarespace, GitHub and all of that.
[00:13:49] Speaker C: You don't even have to. You can start even maybe you have an event coming up or you have some one time campaign that you're running so you can build something in the service of that campaign as well. Like, you know, like if you're throwing like a, like that's a low risk way of doing it because sometimes people are just like nervous about hanging stuff off their website, like on their main page or whatever it is. Like, you know, just, just build something in service of like a one time, you know, promotional event that you're doing or something. So they're like, there are different ways to kind of experiment and play with it.
But I think, but I think it's like it's important to just get started and to see like where the technology is moving so fast. Right.
Which is, which is the other thing I think that people don't understand, which is this is not a thing where you tried for like, you know, on one weekend and then you kind of, you know, like I was like, no, you have to kind of keep playing with it because the things, things are moving so fast that you might try something and today it's probably not super optimal, but like, you know, in like a month or so it actually becomes much more plausible what you're trying to do. So you just have to kind of stay on that, that curve because the frontier is moving just, just, just so fast actually.
[00:15:00] Speaker A: Yeah, yeah. I mean like, I mean chat P has been out for the new 3 0, right? Is that 18 months or 20 months or something?
And now it's not that long since the like the version that really caught fire.
[00:15:13] Speaker B: Caught fire.
[00:15:14] Speaker C: O3, you mean?
[00:15:15] Speaker A: Yeah, yeah, yeah, yeah, right. It's.
[00:15:17] Speaker C: O3 is the one that came out earlier, like, you know, I think in Q1.
[00:15:22] Speaker A: Oh, two and a half. Wasn't that. The one was. Well, I can't remember. What, I don't remember.
[00:15:27] Speaker C: You mean like the, the thing, the original one? The original one? Yeah, yeah, yeah, that one. That one came out in October 2022. Yes.
[00:15:36] Speaker A: No one understands the naming.
[00:15:38] Speaker B: Yeah, yeah, the naming is terrible. But I saw a post just a couple hours ago. The stat is that nobody was coding with AI a year ago. And that's true.
[00:15:49] Speaker A: 00.
[00:15:50] Speaker B: Yeah.
[00:15:51] Speaker A: Unless you're.
[00:15:54] Speaker C: Yeah, I mean I think I saw, I think I think I saw that tweet as well. I think there has been just a progression in the coding space. I think a year ago there were some people I know, like, my CTO is very, very AI forward, but we were using GitHub Copilot, but. But it was like just a. Smarter autocomplete is essentially how people were using it.
And then now where we are like, which is what more agentic coding workflows, which are like, just like, you just tell it like, I want to do this and then it just goes and does it. I think that's like pretty much like a 2025 phenomena. Right? Like, it's just not.
And so it's, but it's just moving. It's, it's, it's. It's moving so fast. Now people are. The latest thing is like, now you can do sub agents, so you just don't give it one task. You can kind of say, can you do these five things at the same time? And so it's just like. And it can go and do five things at the same time as well.
[00:16:43] Speaker B: So like it's moving so fast. I just hooked up my copilot or my cursor with Slack.
So now I can essentially pin my cursor from my Slack with a error message that I get from anywhere inside of my code base and just tell it to fix the code directly in Slack. So I don't even have to tell from Slack. Yeah, exactly.
[00:17:10] Speaker A: It's like having an agent working for you.
[00:17:12] Speaker B: I. Exactly. So my next move is I'm going to pick a couple of clients of mine and I just be like, here, here you go.
[00:17:20] Speaker C: Yeah. There was a tweet yesterday. Somebody was saying that, like, I think it was Hiten, who's like one of these, you know, entrepreneurs. He writes very insightful tweets. He was saying, like, he mistakenly opened Slack when he wants to open ChatGPT.
And. And I replied to him. It's like. I think that it's like the reason why that happens, and it has happened with me as well, is because I think there is a convergence between those two platforms. Right, right. Because we use Slack to talk to humans. But I think increasingly the other end of Slack is not going to be a human. It's going to be some agent. And that's why ChatGPT and Slack, if you think about it, are very similar under the surface in a way. Exactly.
Although they look different to us, maybe.
[00:18:06] Speaker B: Or you talk to humans on Slack and then once you make a decision collaboratively, you just tell the agent to now go do it.
[00:18:16] Speaker C: Yeah, my mental model always is that. And we have some stuff coming out in our product. We have a very strong Slack integration in Sift.
But our mental model has always been like, imagine that you hired an internal, you know, in, in, in, in. In. In France.
How would you interact with that intern? You usually would just chat with them on Slack and they would do stuff and they would come back and whatever. Like you hired some remote person. And I think that's the, the mental model that you should use with, with AI agents is like, you know, you, they are at your back and call over Slack and they do their stuff and that's it.
[00:18:58] Speaker A: Except they were in France, they'd be on vacation. A lot more than.
[00:19:01] Speaker C: Yes.
[00:19:02] Speaker A: Than my, than my agent. But yeah, I think it's all mind blowing. The idea that like it's gotta be commonplace to have agents that you interact with on Slack and they're gonna just feel like regular employees is just. It breaks my brain. Like it hurts. But what's interesting, I think for like young people, it's gonna be very normal. They're gonna just like log into these tools and they'll have friends that are, are agents and they'll have co workers who are real. I mean it's going to be. It's like a whole new world. It's. It's hard to even like really wrap your head around it.
Interesting.
Paul, anything else you want to add to this discussion or should we maybe move on to some news items? There's some really great stuff to talk about.
[00:19:43] Speaker B: Let's move on to the news item.
[00:19:45] Speaker A: Yeah. So do we want to start with the Jassy stuff or.
Oh, sure. No, we should do the, the. What was the post?
[00:19:56] Speaker B: There's a couple great posts.
[00:19:58] Speaker A: So the future of work with AI agents.
[00:20:02] Speaker B: Yes.
[00:20:02] Speaker A: Let me pull that thing was intense.
[00:20:05] Speaker B: It's a survey post.
[00:20:09] Speaker A: I don't know if you looked at this Emran, but there was like a new survey that came out from Stanford and what they interviewed 1500 workers and AI experts about which jobs AI will replace and which ones will automate.
And so at least the way that the. This, let's say the results of the survey or the analysis showed is that a lot of people perhaps have the wrong assumptions about how AI is going to play out. That was my interpretation because Paul, you and I, we looked at this. Right. Some of the charts were complicated to understand, but that's how I would frame it is that there's some things that are really obvious and we've Been talking about them that AI is automating right now and we're all getting a lot of value out of it. It's basically some very extended version of Autocomplete or I can go into ChatGPT and write some copy, we all know that.
But there's people that are building all kinds of startups and this paper was asking people what kind of things do you want automated? And for the most part if you ask a worker or an employee, they want things that are boring and repetitive and not fun to do. But that's not what like a lot of people are working on. It runs the gamut. But people work in a lot of areas that perhaps aren't necessarily great for automation. So anything that would involve like, yeah, that involves like critical decision making. I mean you start getting like the legal space, right? Like maybe can automate it, but maybe people should have some input in those decisions.
What do you want to add, Paul?
[00:21:45] Speaker B: Yeah, you're absolutely right. So there's a disconnect between what types of things AI companies are building versus what people actually want.
So just to summarize this paper is that the researchers discovered about only about 46% of tasks.
People want it to be completely taken over by AI.
These tasks are very, very repetitive tasks. So think of like analytical work where you're just sifting through large quantities of data, doing the same calculations over and over again.
AI is great at doing that. But if you look at all of the venture funded AI startups, it's disproportionately being money is being invested into things like creative work, right? AI producing music, AI producing movies, AI writing content, or even a very high level things like AI making decisions for you.
So the paper is kind of calling out that there is currently a major disconnect between what people want to adopt and want to use right now, which is I just want AI to be a helpful, a partner in my day to day work.
Right.
People are saying that workers are saying automate my boring stuff, not the creative work.
[00:23:22] Speaker A: So I guess like maybe, maybe the question for you, Emory, is like what do you think about this?
Perhaps Stanford researchers who have found what they think is a disconnect between like what Silicon Valley is focusing on and what workers would like them to focus on.
[00:23:37] Speaker C: Yeah, I mean I haven't read the paper. I think when I saw it in the feed and then I kind of quickly glanced through.
[00:23:46] Speaker A: You can just use AI to summarize it though, you know.
[00:23:48] Speaker C: No, I mean that's one thing that your PhD trains you very well, which is like, you know, just look, you look at the, you look at the abstract and you look at the conclusions and then like, you look at the three charts and say, okay, I got it.
I think, I think, I think, I think there is a.
I like the way they framed it, which is that, like, okay, well, there are tasks that are very repetitive and they're low complexity and humans want them automated. And I think, I think a lot of people are working on that front.
And I think that's where. I think that. I don't think there's a lot of divergence there in terms of like, what people are trying to, you know, all the startups that are coming in, obviously in the GTM space, like, we are one of them.
Like, we are looking for those kind of tasks, right? Like, like sort of low complexity happen very often and humans don't want to do them.
Right. A classic one is like, you know, inbound marketing or inbound, like, you know, inbound sdr, for example, or any of those. Like, you know, like, very repetitive, very common, low complexity. Like, you know, like somebody raises their hand, you want to kind of, you know, schedule a meeting. Like an AI agent can do that very well. Now.
I think, I think there is a. There is, there is, there was this other kind of.
And I guess I, I just. So that I can answer your question, Greg. I think your question kind of prem.
It says that, like, there's a disconnect between what is being innovated upon and what the study is kind of telling us. Right.
[00:25:17] Speaker A: And that's what they're trying to say. Yeah, for sure.
[00:25:20] Speaker C: Yeah. I, I don't know. I don't know if I kind of really, I don't. At least in my bubble, I don't see that as much where I still think that majority of stuff is actually in that green light zone of like, very repetitive, you know, and so on.
I think, I think the, I think the fundamental issue is just that the technology is kind of just moving so fast that there are more things that are being brought in, into the sort of the complexity. Like something that was high complexity is now doable by AI.
But there is, like, because it is high complexity, humans don't do it often, in my opinion.
But then, but then if it becomes, you know, automatable, maybe we would do more of it.
And I think some of these papers just don't kind of, you know, they don't capture that loop very well, in my opinion.
[00:26:32] Speaker B: Do you have an example.
[00:26:35] Speaker C: Like, what would be like, super.
[00:26:37] Speaker B: Yeah, this is A good, I think this is a good, good thought process. Do you have like any kind of example?
[00:26:42] Speaker C: I mean I, I think, I think everything that, that that has transpired so far is like that like you know, like writing was sucked for a lot of people, including myself. Right. Like you say like I want to write an article which is like super SEO friendly, blah, blah.
Any, any kind of creative writing was like, you know, was difficult before and we didn't do a lot of it and now we can. And I'm not saying it's completely automated away, but nobody starts with a blank slate. Right? Nobody gets stuck generally. Right. Like, I mean it's just like there's less friction and so like we write more because of AI. I think we write better on the, on average with AI.
And so I think, and I'm not talking about just like garbage fluff that people just completely auto generate with AI. I'm talking about like people who kind of really are like using AI, thought partner and so on. So I think that's one example.
It looks very trivial now, but it was an issue like two years ago.
Same with software. We were just talking about software earlier. Like building complicated software was just like that was a lot of work to kind of get ship a feature like, you know, you have to kind of get involved like you know, three people like product design, blah, blah, engineering to kind of ship something. It would take like you know, 15 meetings and like, you know, God knows how many strategy sessions and now people just move much faster because you know, what was a complicated thing like is not super complicated and we'll write way more software because of it.
And so anyway, so I think these studies are like interesting but they just don't capture that.
[00:28:22] Speaker A: Yeah, my take is that this study, it's very academic in a sense that it's good because I think they're trying to counterbalance some of the over enthusiasm that executives in Silicon Valley has for AI. So I appreciate that. But on the other hand, I think it just shows that they don't understand venture capital and investing at all. And I wrote something a long time ago about this. It was basically titled why Smart Investors Fund Dumb Ideas.
And I think that like yeah, they don't, they don't get Silicon Valley right. Like Silicon Valley needs a bet in every category. So of course like some of them are because I chart there is dispersion across everything and like maybe there's emphasis in this one quadrant that they claim are the dumb ones or something dumb investments.
But like Silicon Valley always invests in Every category. That's how like the venture firms at like the, they see like an investment class. Like we need a legal option, we need a creative option, we need a investment in this class. Right. And they, they put it into everything.
But I think for me, what's more fundamental about venture capital that very few people understand is that big companies are working on smart innovations.
So obvious innovations are going to be captured by big companies.
And that could be maybe the CD ROM or the iPhone. There's so many examples. Like a big company produced an incredible innovation, bought its market, is very successful. Those are all the obvious ones. What Silicon Valley does is find the non obvious ones. And the way to find the non obvious one is you have to look at like the stupid ideas. And some of them are just not that stupid and they work. And in my opinion, the ultimate example is the, the website that we love, Twitter. That was like not a smart idea. It was kind of stupid. It was like, oh My God, a 30 character micro blog that I can only read on text. When it came out, like it was like text message based. It was the craziest idea.
[00:30:21] Speaker B: Weird. Weird. That's not Facebook.
[00:30:24] Speaker A: It's silly. But it was so different and so weird and so new that it took off. Right. And for me it's an excellent example of like an idea that's like, it's kind of dumb but like just interesting enough that it created a whole new category and a whole new way for people to communicate. And so I think that that's what this like report doesn't get. I think it's important to point it out. But I do appreciate they've tried 10% of the like maybe over enthusiasm with people claiming they're going to automate everything. Like it was very obvious to me and I feel comfortable saying like Klarna went too far. And it was very clear that like they were making these statements perhaps to juice their IPO or something and convince people that they were more profitable than maybe they were. And so they were like, oh, we're going to replace all these positions with AI and made all these claims. And then it kind of, I think significantly backpedaled. Actually if they went from hey, we're going to get rid of like all these positions to now we're actually rehiring customer service because the results weren't very good.
Is pretty interesting.
Or it's a good example of like, let's call it the over enthusiasm for, for AI with some people.
[00:31:35] Speaker B: Yeah.
[00:31:36] Speaker C: There's also like a structural issue with studies like this as well, sorry Paul, I spoke over you.
[00:31:40] Speaker B: No, go for it.
[00:31:41] Speaker C: Yeah, I think, I think one thing that's like a lot of these studies, like obviously you are asking the workers directly and so there is like I think two things with it. One is like obviously there is a, the sort of the self preservation aspect of it or like. But I think the more important one also is that like they might not be aware of a lot of AI is just moving so fast. They just don't know what are the capabilities of the models. Right. And so like there's just a lot of like biases that kind of come into play with like some of these things like with just like the way the study is kind of structured. Right.
But so, so I think, I think you have to kind of discount for some of those facts as well. Like it might be better to kind of sometimes ask for, ask these questions to somebody who's like less in a, like you know, it's a little bit more outside of the system in general. Right.
Like go ask like the decision makers perspective. It's like okay, where are we kind of seeing inefficiencies that you know like so like there's just like that aspect of it as well. But I don't know, I mean it's, it's, it's a, I don't think I agreed with a lot of what the paper was trying to say. Like it's just that I think, I think startups are working on stuff that is, I don't know, I, I, that is very, very, that's very repeatable I think is like what, where a lot of, of research, where a lot of productionization is going on. I think there is a lot of like the other kind of stuff that's happening which is like high complexity, doesn't happen often kind of stuff and I think foundation model companies are doing a lot of that stuff, right? Like, like cracking like exams and like you know, hard mathematical problems and like, like that kind of stuff is happening.
But I think that's kind of more happening on the sort of the, like the open AIs and the anthropics of the world are kind of spending money on that kind of stuff rather than sort of like, you know, venture back startups generally.
[00:33:35] Speaker A: Yeah, I mean you're dead on with the bias, right? It's like if you go to worker and say like do you guys want to be replaced by AI? They go like not really, but can it do some of these like really boring things they don't like to do? Yeah, it's Very. It's a. It's very obvious.
Okay, Paul, should we move on to the next one?
[00:33:52] Speaker B: Let's do it.
[00:33:52] Speaker A: We have a related news item.
[00:33:54] Speaker B: Very related, actually.
Also a little bit of paradoxical. Right. I'll share the news article here, which is the headline is AI will shrink Amazon's workforce in the coming years, says the CEO Jassy.
Following up to everything that we just discussed, where Klarna hired back some of the 700 or so humans that they fired for AI.
So I guess the real question here is, do you think that this is.
Is there.
Did Jassy say, first of all, like, would you believe that this is true? Where Amazon's main reason for laying off all of these employees is because of AI and will they hire them back in half a year or year or so?
What's your takes?
[00:34:48] Speaker A: I mean, as a former Amazon employee. Right. Like, what do you think about this?
[00:34:54] Speaker C: It's. For me, right?
[00:34:56] Speaker A: Yeah.
[00:34:56] Speaker B: Yes.
[00:34:59] Speaker C: I mean, I think Amazon is one of those.
I worked there. It was the first place where I worked in my career.
And it's a company that is very much invested in operational excellence and efficiency. Right. Like, people forget, like, Amazon's background is in like, you know, it's essentially an E commerce, like a, you know, company.
And so they really care about their margins and their whole, like their whole strategy in a way is like, you know, there's a Bezos quote, which is, your margin is my opportunity.
And so I wouldn't be surprised if they are.
Of all the big companies, they are the ones that kind of go and try to get as much out of AI as possible. Because I think their business model is just very much tied with that efficiency philosophy.
Unlike, say, Google, which pretty much has a monopoly on certain businesses.
I don't think Amazon operates in that kind of a business in general. Right.
Not on the commerce side even. Not even on the cloud side. It's very competitive now.
So I do agree with what they are trying to say, which is that they will kind of use AI to kind of shrink their workforce as much as possible.
But again, I think it's just like a lot of it is just going to get.
I don't know. I. I think again, it's like, what is it like that phenomena that we were talking about, right? Like, which is. I think the term for it is Jeevon's Paradox, which is like, you know, if something is like, you know, it's like if you go talk to Sam Altman, he'll say, just say, like, yeah, we'll Lose the jobs that we have, but we'll make better jobs, other kind of jobs.
And so I think that's kind of the, that's, that's the big unknown, right? Like just sort of like, okay, we are just thinking in terms of current jobs because that's all we can envision right now. It's like, okay, job, yeah, we'll replace it. But a lot of people just don't know what other resources, what other kind of jobs kind of come out of this glut of resources that we'll have and the glut of tokens that we'll have at our disposal. So I don't know. My guess would be like, companies probably stay at flathead count, but it just gets redistributed to job functions that are different than how we kind of imagine. Like, you know, that that would have been my, my guess.
[00:37:36] Speaker A: Yeah, I think he basically hits on all those points in his own memo. I think that, you know, the media likes to frame it as like the sky is falling.
I, I think that like, I find it refreshing that a CEO is like being balanced and let's even say nuanced and how he's communicating and trying to like, tell people about like maybe something they don't want to hear. I think that's great. I think he should be clear about like where he thinks the business is going. And he's speaking to both, let's say shareholders and, and employees. And he hit on all the points that you hit on that, which is like, we don't know exactly how it's going to work. Today the mood seems to be that everyone thinks it's going to swing in this direction.
Even says, like, educate yourself, AI. It's obvious some jobs that go away and new jobs will emerge at the highest level. I agree 100% with you that like new jobs will emerge. I think that all of this, my personal opinion, I guess is that, yeah, then you can chime in. Is that all of the, like this like, like labor force slack in the white collar job area is just over hiring from 2021 and the end of Zurb. I don't think it's related to AI at all. I think AI is a great excuse and there's a lot of efficiency and there's a lot of like desire or there's like a mood that people see that they need to be efficient and there's a threat to motivate employees. But I think that most companies are just going through this transition where you went from the ZIRP era where money was free. And so because of that, it was very real that people were hiring as rapidly as they could and spending as rapidly as they could because there was a lot of market share that or at least perception that they needed to grow fast. Right. And that was like WeWork or Uber or DoorDash or any of these companies. And so we're just in a very different business environment where there's not the same kind of competitive pressures. Like we're not having these like unicorns minted once a week that need to hire people. Right. So there's just not the same competitive pressures to drive a lot of different factors in the job market. Right. And so I don't really think that it's AI.
There's some on the margins things happening with, with AI where there's a reduction in some perhaps job needs, I think.
But I agree with you that ultimately all the slack will get used up. It's like when you increase the capacity of a freeway, it ultimately just accommodates more cars. And I think it's exactly what will happen here. There's, like I said, maybe even though even the truck drivers are going to go away. You guys remember that narrative that's from like nine years ago.
That's a long time ago. Like we still have truck drivers. And the more that I dug into it, so I was very curious because I didn't really buy that narrative at all.
You need the truck driver to like unload the truck.
So unless like we have Optimus driving these things, like there's a lot that has to happen to get to a world where like truck driving is completely automated 100% and maybe someday we'll get there, but we're not anywhere near as close as perhaps some people wanted us to think.
Even, even eight years ago.
[00:40:46] Speaker C: Yeah, I think, I think that, I think that another, another sort of parallel is that like the business process or out offshoring industry? Right. Like just sort of like that. That purely works on economics as well. Like, you know, like, you know, you just cheaper labor offshore, like, you know, contractors and all of that stuff.
And those are like in a way kind of tokens as well. They're a little bit more expensive, but they have been available. And I mean that hasn't really gone. And yes it has over a period of time. But I think people overestimate the time horizon on which these things play out. It just takes a while.
Just sort of even things that right now, obviously in the news, the hollowing out of American manufacturing process hours like that happened over a long Period of time as well. So, like, just. I think these things just kind of play out over a longer time period.
I think counterintuitively, what this will do is, and I know Amazon has a lot of, like, a big, like, labor workforce that is not permanent. Like, they do, like, a lot of, you know, like, contracting. Yeah. But I think what's going to happen is, is this will.
This will depress or this will set back emerging economies that were going to use services as a way to kind of, you know, bring their standards up very, very quickly. Like, I grew up in India and, like, India has a huge services industry. Right.
In software and, like, call, you know, calling centers and all that stuff. Like, same with, like, you know, like Philippines and Thailand and all of those countries, and they use, like, these services things to kind of really bring people up.
And I think I'm just somewhat worried that, like, this kind of, you know, kind of dampens their growth a little bit going forward because, like, a lot of what was going to be, like, outsourced now kind of can be done, you know, like, a bit more efficiently, like, probably just in house or something like that.
So I think those are some of the, like, side effects that people just don't know. Don't think, like, same thing with software. Right. Like, you could just like, write like, some.
If you wanted to write some simple, like, you know, internal software, a fair amount of companies, like, they will contract it out to, like, you know, they'll offshore it, but now they can just build it in house with.
[00:43:07] Speaker B: Right.
[00:43:08] Speaker C: So I think that are, like, there are those effects that, like, I think probably are not as well understood.
[00:43:13] Speaker B: Yeah. Even. Even in those areas. So a good example is qa.
In software, there's a ton of QA that's offshored.
And the chatter or the wish is that AI will first of all completely take over that entire industry, automate it.
It's still left to be seen if that is doable, because even with qa, there's so much nuance that you need the human mind. A good QA engineer pays for itself. That's like the, the.
The phrase that everyone says in software.
So even if that's, like, if QA cannot be outsourced to AI, then very, very few things in the software could be completely outsourced to AI.
That's kind of paradox.
[00:44:07] Speaker A: What is it, Steel, man?
[00:44:08] Speaker C: It.
[00:44:08] Speaker A: I would take the other side to it.
[00:44:10] Speaker B: Yeah.
[00:44:10] Speaker A: Because I think you made some good points, Emren in particular. And you did too, Paul. But. So here's the other side is that Because I, I worked for companies headquartered in India and I've been to India many times, right. And worked a lot of people over there.
And like, so when I first started working with people in India, that was like 2009, it was a long time ago. And so the primary market was, or at least with the perception was like it's this outsourced market, right. But it's come a long way and now it's like excellent market on its own for high tech startups and it's really become very different than it perhaps was perceived. And I don't know if like everyone in let's say America has caught up with like the fact that a lot of these emerging markets have come a long way since, since they were like the outsourced markets. And India's like demographics mean that like it's going to be a leader in the future in a way that China's demographics aren't going in that direction. Right. And so I think the, the long term like relationship and economic, I don't know, value of the United States and India working together is, is, I'm very excited about, I think it's gonna be really cool. There's gonna be all kinds of stuff that happening. I don't think people are necessarily see that. And I think that's, that's happening like in Indonesia and Grab and that company se. Right. There's a whole bunch of stuff happening. It wasn't happening. So, so it's the other side to it. It's going to have an impact, just like it's going to have an impact here. But.
[00:45:40] Speaker C: Yeah, yeah, and I think that's a great point. I think that's the counterbalancing argument. Right? And I think that to a certain extent that happened with China as well where like China, like, you know, like domestic consumption increases and so their own internal markets are huge because like you know, India and China and everybody like south, like those countries are like huge populations and so their internal market can support their growth in a way. Right. So I think that's a fair point. Like you know, that shift like you know, like where China was like just like now they have a very robust internal economy and they can just produce and sell for themselves. And I think that the same trend happens, like you said, like in services with India as well. There are Indian software startups that are not just like it started with offshoring obviously and we had like the big three in India that like, just like we have the big four here, they're like big three in offshoring in India. But now, like, there are lots of innovative Indian startups that kind of build software for the local economy as well as, like, obviously they export it out as well. Like, you know, to like Zoho, for example, is a very famous example based out of India.
But. And it's a SaaS that is like global now.
[00:46:54] Speaker A: Yeah, I mean, I have a. I've been. Paul. I don't know if you work with a lot of Oshko people, but I've been working with people and somehow I stumbled into like a. A group of people in Eastern Europe that I've worked with out of like Serbia and Hungary. And they're excellent to work with. There's even one, like, he's a kid, he's like, he's like in college, goes to engineering program in like Hungary, but he's from the Middle east and he's working on this interesting startup. And one of the primary value propositions is that he supports a bunch of languages that like an American startup wouldn't necessarily support. And I was like, dude, I think you're like really onto something. And then we were talking, his English is very good. And I was like, is Arabic your first language? Like, oh, of course, Poland. I was like, dude, you should go to Dubai and get funded and like be a player. Like if you can support Arabic and all these European languages and like have this kind of different kind of approach and like typical American company would do, there's big market. So at a high level, I'm really excited. I think that, like, I think that there's all kinds of stuff that's happening and AI is going to enable a really exciting future with all kinds of startups and they'll be able to focus on these markets. Because the core for me here was that AI and translation are.
Yes, it's a problem that AI can solve very easily. So the idea that you could have these startups now that the language and like localization piece isn't anywhere near as expensive as it used to be. And so when I did a lot of global marketing, like that was always this giant blind item that we got nothing for it. I was always so frustrated. I'm like, gotta spend all this money, translate it. And then I had to do different versions of Spanish. Like, this is just so frustrating. So now that I don't have to have, like, have this giant line item, I could just focus on, you know, spending more on marketing or whatever and getting the product in front of people and the translation can happen. Like there's just all kinds of Interesting avenues I think AI will unlock coupled with globalization.
[00:48:49] Speaker B: I see.
Yeah. That's actually like, a really interesting takeaway, I think.
So the direct translation jobs will be completely gone, but I don't think AI can replace the cultural nuance that comes with the translation.
[00:49:06] Speaker C: Right.
[00:49:07] Speaker B: Yeah, yeah, but it will be rolled into, let's say, design now. Right. You can't. Let's say you're building a web page or a web application.
You can translate it in all the languages now very, very quickly, but doesn't mean that the same design will work across all countries.
[00:49:26] Speaker A: Yeah, you can put more effort into, like, doing the localization.
You can put more effort into localization rather than just translation. Right. There's a lot more that needs to happen. Yeah, yeah, absolutely.
[00:49:38] Speaker B: I think that's what AI is helping us to do right now. Right. Like, our bottleneck, the goalpost of the bottleneck will move down somewhere else.
So, like, you're. Initially, you wouldn't even take on this project because translating into 150 languages is impossible, but now you can, and now you need to aggregate all of that effort into, like, maybe localized customization for 10 different regions.
That's still a ton of work.
[00:50:08] Speaker A: Yeah.
[00:50:09] Speaker C: Awesome.
[00:50:10] Speaker A: Great discussion. Okay, I should move on to Reddit.
[00:50:14] Speaker B: My favorite Reddit thread.
[00:50:17] Speaker C: I'm going to duck out here, but thanks for having me.
[00:50:20] Speaker B: Awesome.
[00:50:21] Speaker A: Of course. Thank you.
[00:50:23] Speaker B: Thank you for coming.
[00:50:24] Speaker C: All right, see ya.
[00:50:28] Speaker A: Yeah, we can't not do the Reddit segment.
[00:50:31] Speaker B: Yeah, it's perfect time.
[00:50:32] Speaker A: Perfect timing. Perfect timing. All right.
Our favorite Reddit threads where we doom. Scroll Reddit so you don't have to.
All right.
Is this the one?
[00:50:46] Speaker B: Yeah.
Let me share this. The building public isn't a good idea.
[00:50:52] Speaker A: Oh, yes.
So the guy wrote the thread about building in public. Right. Not a good idea. Here's my experience.
I built a product that made $18,000 and someone copied it.
Do we really need to read the rest of this or is like.
[00:51:07] Speaker B: No, I think that tells the entire story. Yeah.
[00:51:11] Speaker A: That's really funny. You and I started actually chatting a lot about build in public and some of the downfall of it.
[00:51:20] Speaker B: Yeah. I mean, especially after our discussion today where, you know, building is not no longer a mold. Maybe this is a good thing. I'm gonna steal, man, this. Right? Okay.
[00:51:33] Speaker A: Okay.
[00:51:34] Speaker B: Building public isn't a good idea. I think that's scarcity mindset in where you still think that your moat is the fact that you built something.
I think, no, there's no unique ideas out there anymore.
I think execution still matters, but it's really about the distribution.
So if distribution is the most important thing and getting people's attention, then building public is a good way to get distribution and people's attention.
It's the easiest way to market your product is to build it in public.
[00:52:19] Speaker A: I mean, I. I think, I think. So what do I think? I feel like, okay, a couple of things. I feel this trend has started to just kind of decline and we're in the. It's in. It's the, it's the. Everything works in the Gartner life cycle curve, right?
[00:52:33] Speaker B: Yeah.
[00:52:33] Speaker A: Like, there's all this excitement, enthusiasm, building public, and then we reach the peak of inflated expectations and now we're headed into the trough of disillusionment. Or if you're like, oh, my God, I built public and it doesn't work anymore.
[00:52:45] Speaker B: Yeah.
[00:52:45] Speaker A: So, yeah, I just think it's part of the cycle and that people just aren't as excited about building public. I even think, like, this year will probably mark the low point and just general enthusiasm for startups. I mean, in Silicon Valley, AI startups are really hot. But beyond that, it's not something that everyone is as enthusiastic about as they were perhaps in the past. And I've seen this kind of cycle come and go. Right, right. And so I think that's part of that. There's just this, like, general kind of malaise. It's hard to raise money. Startups are hard to build.
They're not that cool anymore.
And so, yeah, build in public.
The other point I want to make was like, it was on this week in Search, Jason Calipanis, maybe yesterday was talking about maybe the company that we could like, perhaps say that they popularized. I don't know if they invented building public, but Buffer, they have that social media management tool.
[00:53:42] Speaker B: Yeah.
[00:53:42] Speaker A: And they went really extreme with this transparency approach, like postpartum.
[00:53:48] Speaker B: Right, Right.
[00:53:48] Speaker A: Yeah, that's the term I use.
[00:53:49] Speaker C: Right.
[00:53:50] Speaker A: They posted all the salaries and equity structure that people like.
They did some crazy stuff that, like, I just would not be comfortable with, like posting people's salaries and stuff. It's just.
[00:54:01] Speaker B: Yeah, I just.
[00:54:02] Speaker A: I just don't get comfortable with it.
[00:54:04] Speaker B: Yeah, I think so. Sorry to cut you off, Peter. Theo actually got this exactly correct.
You want to build in public when you're small.
Right.
You can be transparent. You can market yourself, you can show success, you want to build yourself up. If you have momentum, you want to tell everybody that looks how well I'm doing, show all the momentum. But then after a while, when you're big, when you're let's say a monopoly.
It's actually the complete opposite. You should hide all your secrets.
[00:54:43] Speaker A: Yeah, it's like a hack. It's a hack when you're small and you, you and you share transparency and then you can show traction.
[00:54:48] Speaker B: You're like momentum traction. Exactly. Yeah. Oh, I'm growing 100 month over month.
Look at like I'm doubling my revenue count, customers retention, you know, employee satisfaction. All of these things roll into the story that you want to tell. You're super successful, but then eventually you should maybe step back a little bit. Right.
Because then it's scrutiny.
[00:55:14] Speaker A: Yeah, I, I, I mean I agree with all of that. I think that like, I mean and like look, even like with Buffer, I think it's a good example where they, they got a lot of traction for it and, and they kept being radically transparent and they, and they've just take, that company's just gone in a different direction. A lot of the companies go too. Like, I don't think they raised a lot of money. They raised some money and then they were like, they just kind of ran it as a successful high tech company and just, you know, didn't let's say play the venture game or try to raise more and try to go public or something like that. So they kind of took a different approach too where that worked for them. And so I guess the high level, I like the idea that like, you know, whatever works for you, do right. Build in public is working for you. Great. And then don't be dogmatic. Like if there's a point where it doesn't work for you, don't do it anymore. Right. Flex advice people get very absolute, like if you make this decision, you have to like keep doing it and stuff. But that's, that's ridiculous.
[00:56:15] Speaker B: Correct.
[00:56:16] Speaker A: Correct.
[00:56:17] Speaker B: I think people lose sight of what's the most important thing. And also I think it's just like once people start adopting something like building public, the definition changes. So there's so many different definitions now. Right. Especially maybe it started in the software industry and now I've heard financial advisors are coming on and jumping on this.
E Commerce founders are jumping on building public factory owners or some of the VC backed factory owners are building public. You know, they're building a factory in public. There's A, there's a YouTube series of, I think a guy that makes, I don't know, something and he's building in public on how he runs his entire factory.
It's interesting, it gets a lot of views, but I think the definition just completely changed over, over the time to whatever.
[00:57:19] Speaker A: So, I mean, Jason, Jason Calcan is pointing out a point that it's hard to prove, but he felt that like, if Buffer wasn't so vocal about how much money they're making, they would have had less competitors.
[00:57:29] Speaker B: And that is what Right, right.
[00:57:31] Speaker A: Identifies in this thread as well. He says that, like, if you post, I think this is probably some of the revenue numbers people post that people can copy. If they think you're onto something, they will copy you, which could hurt your business.
Because that was, even Jason was saying, let's say they still have a great business, but if maybe they hadn't been so vocal about how much money they're making, they might make more, that might have pushed the price down for that product and put a lot of people into it. So I, I think I would generally agree. I, maybe I'm just old to like, sharing financial information is just very taboo and it seems very, like, unsophisticated or it's just something that is the world that I grew up in. Like, just not, it's just, it was, it was just not something you did. You just didn't really talk about these things. And I know young people, like, they share their bank accounts, screen capture each other and everything. And like, I don't do any of that at all.
[00:58:28] Speaker B: I'm just, yeah, especially the hackers act, where every single one of your revenue numbers over a period of time can be tracked back. I've seen it work really well. Right. Like, I've seen slowly raising your pricing public is a good strategy. So you could be like, hey, right now I'm only offering a 500 plan for the next 10 people and then I raise it to a thousand, two thousand. I think that's probably the best way to signal and building public without revealing how much you're making. People can do the mental gymnastics if they want to. Right? Like, okay, you close 10 spots at 500, we're probably making $5,000 a month.
I think that's the proper way to quote, unquote, building public.
[00:59:19] Speaker A: Yeah, I, I, I guess so. So, yeah, I'm very, I'm very careful about what I share publicly. There's just some information and just don't want to share. I just don't think it's appropriate. I don't, and it's, it might not be logical, but like, I think it's perfect.
[00:59:35] Speaker B: You don't, you also don't know about it. Who's watching, who's listening. You know, I just, I just like.
[00:59:40] Speaker A: A Conference world where, like, it seems just impolite. It's very.
It's very, like.
Like, it's very, like, new money versus old money. I don't know if this makes sense anymore, but, like, yeah, west coast versus.
[00:59:53] Speaker B: It's a little bit of west coast versus East Coast.
[00:59:55] Speaker C: Yeah.
[00:59:55] Speaker A: Like, where I grew up in New York too, like, old money tended to be understated, and they were. They wanted to just be viewed as, like, regular people or. Or understated is a better term than regular people.
[01:00:08] Speaker B: Yeah.
[01:00:08] Speaker A: And then new money, very flashy and showy, and that was always been the kind of, like, yeah. Stereotype.
[01:00:14] Speaker B: Right.
[01:00:15] Speaker A: And they looked down upon new money. Right. And so maybe that's where some of my, like, perception comes from. Like, is that if you're too flashy and.
And vocal about your startup and how much money you're making, all that. It just. It just seems very. It just seems very new money.
[01:00:31] Speaker B: Yeah, I agree.
[01:00:35] Speaker A: All right, awesome. Are we where's. It's authority. We're past the hour. Okay, so what do we have left on our. On our schedule?
[01:00:43] Speaker B: I think we're. I think we're right on time.
[01:00:46] Speaker A: Do we need to do our meme. Meme of the.
[01:00:50] Speaker B: Let's do it.
[01:00:50] Speaker A: All right.
[01:00:55] Speaker B: Oh, I love the meme of the week.
[01:00:57] Speaker A: Oh, yeah. This was good.
Are you gonna share it?
[01:01:01] Speaker B: Yeah, I'm pulling that up.
[01:01:03] Speaker C: All right. All right.
Okay.
[01:01:12] Speaker A: Yeah. There you go.
[01:01:12] Speaker B: Hilarious.
[01:01:13] Speaker A: Oh, my God.
The Iranian memes were pretty funny. I was having a good time with these ones. That account was pretty good too. There were a couple other that were like, yeah, I thought this was. This was good. Right. If you aren't aware, like, Circle launched a stablecoin, or they went public, they launched one, and then the regulation changed around stablecoins too, and a bunch of crypto stocks went. Went to the moon. And so I found this one quite, quite funny.
I retweeted this. I said, oh, so that's what all the fuss is about.
[01:01:53] Speaker B: Yeah. I actually really like the. Your latest one.
You had another one? That's really okay.
[01:02:00] Speaker A: This one's good. This one's good.
Can you make that one bigger?
[01:02:05] Speaker B: Yeah, this is the Baywatch one, but a Baywatch one.
[01:02:10] Speaker A: Yeah, that was good too. Yeah. Everyone was sharing all of those, like, pre Iranian revolution photos of, like, people on the beach and stuff.
[01:02:17] Speaker B: Yeah. This is hilarious.
[01:02:20] Speaker A: Yeah, that was why I thought that was funny. Basically, Payman is in Baywatch, right?
[01:02:23] Speaker B: Yeah.
Awesome. I love these meme formats where you're just posting something completely wrong. I've seen you post a few ones where it's like, I love.
This is like, Sicily, Italy or something. It's a photo of a photo of some, like, American town.
[01:02:43] Speaker A: Yeah, those. Those have been good. I. I did one.
Yeah, it was Cinque Terre in Italy.
[01:02:49] Speaker B: Yeah.
[01:02:50] Speaker A: I posted it and I said, oh, Burlington, Vermont is beautiful @ this time of year.
Where my mother lives there. And so I know Burlington quite well. Like, a little tiny town of Vermont.
And it was really funny. That was like, the usage was different on Twitter. There was all these people replying with, like, that's not Burlington, Vermont.
Like, totally look closely. And then I would, like, point out something about Vermont. And then I had a bunch of people actually reply with, like, oh, my God. They figured out that, like, I knew the town.
[01:03:21] Speaker B: Yeah, yeah, yeah, yep, exactly.
[01:03:23] Speaker A: Quite well. And actually I actually met a few people from that thread. Like, I actually met a guy that's like, an investor.
So the lesson there is, like, if you do some hilarious posting like that, you can actually, like, post your way to seed 100.
I didn't race Seed, but, like, it's totally possible. I believe in it. All right, that was fun.
[01:03:48] Speaker B: Love it.
[01:03:50] Speaker A: Thanks, Paul.
[01:03:50] Speaker B: Yeah, we'll see you next week.
[01:03:52] Speaker A: Yeah, take care.
