SaaStr 806: How The Smartest Founders Are Already Winning with AI: SaaStr CEO Jason Lemkin on My First Million

0:01Welcome to the official Sastra podcast where you can hear some of the best Sastra speakers. This is where the cloud meets. Up today on the Saster Podcast. But did you use like Delphi or you built your own tool or what'd you do? I tried a couple of things and then I used Delphi. And then at first we had so much content, it was a little difficult to ingest it. So there is a little bit of work. And then I spend about 10 minutes a day training it, which we could talk about. So there is a little bit of maintenance involved, but it's better than me. And the thing I didn't get until I saw something that was hyper-trained, Like, this is a lot of content, 20 million words, right?

0:38Very few mistakes, very few hallucinations, and the ability to connect things that I can't. I just don't remember. I don't remember what Sam and I said on The Hustle on stage years ago versus when we did an MFM a while versus today. But the AI knows it all. Hey, are you tired of listening to hours and hours of sales calls? Recording is yesterday's game, folks. Yesterday's game. Attention.com unleashes an army of AI sales agents that auto-update your CRM, build custom sales decks, spot cross-sell signals, and score calls even before the coffee's cold, even before the coffee's cold. Teams like Bamboo HR and Scale AI already automate their sales and rev ops using customer conversations.

1:21Step into the future at attention.com slash saster.

1:31Hey, everybody. This episode is brought to you by Adio, the AI native CRM. Just connect your email and Adio instantly builds a powerful CRM. With every company contact and interaction you've ever had, get 15 % off your first year. That's 15 % off your first year at attio.com slash saster. That's adio.com slash saster.

1:58So for this episode, what I'm thinking this is about is this is, and guys, this is going to be generous to us, but what is the smart money doing with AI? And, you know, in betting and sports betting in Vegas, you always kind of want to watch, what are the sharps doing? They tend to be just, you know, maybe 10 to 15 % earlier to understanding things or to recognizing things. And so similarly, Jason, I wanted you on because you've been doing really interesting stuff with AI, both from investing, you're talking to startups that are using them, you're reading all the reports, but also you yourself built this AI tool.

2:30And so I want to kind of get your reaction to a few of the different things. Can you first explain this AI experiment you did where you basically made a body double of yourself out of AI? And like, what was the experiment and what'd you learn? Well, yeah. So Sastr has a lot of content, as you guys know. We have 12 years of content. Some of it's out of date, for sure. Saster being, you're a previous founder who's exited multiple software companies. Now you're like the godfather of like Sass because you have this blog and media company and events business called Saster and you have a huge fund. So that's kind of the...

3:01But probably most interestingly for like seeing the future a little bit is I have about 10 ,000 pieces of content I've written over the years. A lot of it, which is not out of date. And I update it personally. So about 10 ,000 pieces of content. And then we have several thousand interviews like you guys do, but with so many great folks, early ones with Dharmesh, middle ones from Dharmesh. Dharmesh, when HubSpot had just IPO'd, Dharmesh the other day, Yamini last week, Brian Halligan over the years. And we have all of it. And our AI took everything I've written, every tweet I've ever done, every LinkedIn post, every SASTR video from every speaker for 12 years.

3:37Aaron Levy, the week he went IPO for the first SASTR annual to last week when we were talking about AI. all of it ingested in this AI. And so what I instantly saw was as soon as I would talk to this AI, it was much better than me. It was much better than me because I forget stuff. How did you get that information in there? You just hook up RSS, right? The feed for blog. It can scrape the web, which is pretty easy. Soon we'll be able to do other things like MCB, but it's just scraping APIs and RSS. It's really just scraping. You just give a URL. But did you use like Delphi or you built your own tool or what'd you do?

4:12I tried a couple things and then I used Delphi. And then at first we had so much content, it was a little difficult to ingest it. So there is a little bit of work. And then I spend about 10 minutes a day training it, which we could talk about. So there is a little bit of maintenance involved, but it's better than me. And the thing I didn't get until I saw something that was hyper-trained, like this is a lot of content, 20 million words, right? Very few mistakes, very few hallucinations, and the ability to connect things that I can't. I just don't remember. I don't remember what Sam and I said on The Hustle on stage years ago versus when we did an MFM a while versus today.

4:50But the AI knows it all. By the way, 20 million words is like 200 books worth of content. An average book, I think, is like 100 ,000 words. So 20 million words you're saying is you basically gave it 200 books of your own brain and the brains of other guests that have come on your show. Yes. And your goal was to build, was it to build basically an advisor, an AI version of you that could be an advisor to a founder who wants to come and talk to you, ask questions, but you don't have enough time of the day to book calls? Was that the big idea here? I just wanted to learn like you guys did. I just wanted to learn.

5:21I didn't think it would be as good as it was because I used other products and I've used other things and it wasn't as good as my digital AI was. The other interesting thing is the use cases with AI are much broader than you'd expect. Okay. So I didn't expect people to upload their board decks before board meetings and ask for feedback before board meeting. That's an issue case. I didn't expect hundreds and hundreds of founders to use it as their therapist for their company where the growth has slowed. I didn't expect that. I didn't expect so many folks to use it to review their sales scripts, their SDI scripts when there's specific tools to do it, right?

5:58It's so good at reviewing VC pitch decks. It's so good at it, better than anything I've seen. Can you get tactical for one second? So you go to Delphi.ai and you just uploaded the Sassers RSS feed and it did all of the work? YouTube, RSS, Twitter, blog. But could I go to a health guru or a business guru who I subscribe to, who I love? Yes. Can I say YouTube.com slash Alex Hermosi, upload that and it automatically does it? Is this that simple? Well, it's an interesting question. It is prohibited, right? It is prohibited by their website. their terms of use. They prohibit it. But honestly, this is an AI risk.

6:38You could do it. Absolutely, you could do it. You could redo what I did. It might take you a little while longer, but you could just clone exact myself, yourself. And then you could do it just as well as I did. Sam, I've never felt so used and abused when Jess Ma came on the podcast and she was like, yeah, actually, I have my team training in AI of you so that if I'm brainstorming ideas, I don't have to ever call you. I'll just ask the AI what, you know, basically, you'll just be in the meeting with us yeah it's just like we yeah we just took all the podcasts and we're training it on how you think the questions you would ask what you would say reactions you would give and now you're just there in the meeting so she basically hired me without ever paying me right because i have this corpus of content out there i've actually put my brain out there in the public how do i do that like how can't can i just upload a youtube channel and say this youtube creator has great advice and i could upload 800 videos and it will transcribe all of them.

7:30I could go to Delphi. But yeah, you can build all of this. And here's the interesting learning. So we could build a digital Sam and we could do this for next week or whatever. You can do it. My learning from my AI, it's okay that it's not you. This is why, for example, you can train your voice, 11 Labs. And 11 Labs is epic, right? But you can choose, do you want it to be exactly like you, which is what Brian Halligan did? And then I realized that was creepy. I don't want digital Jason to be me. I want digital Jason to be a prime version of me that 90 % of the time is much better than me, but is distinct.

8:02I don't want people confused like Brian, I get like Brian Halligan. And so it's okay that it's not Sam. Like it could be a version of Sam that is better than Sam. So, okay, let's walk through these top 10 learning. So you build this AI. It has 35 ,000 conversations with people. You did that as an experiment. Almost 50 now. There are 50 ,000 conversations and learning number one, you said users will tell AI things that they won't tell humans. What do you mean? Yeah. Well, first of all, before we get there, one obvious, like, there's no way I could do 50 ,000 conversations, is there? Right. I'm not even a people person.

8:39I could do like one or two a day and I get tired. 50 ,000. Just think about the delta between two a day and 50 ,000. But yeah, this is the thing. Like, there are some things that are somewhat unique about me that are early, the amount of content, Sean, like your math, right? The amount of books we're talking about here, right? The amount of content I've written just about B2B, just about building business software. The other thing is there's enough people, I'm not like a super influencer like you guys are, but there's enough people that know the digital me that they already know what to ask me, right?

9:08They've already watched me at an event, they've read the content. So like they ask me things without thinking it's a prompt or a prompt, they just ask things they want to talk about their deepest fears. Like I've partnered with Sam on this podcast, but I'm not sure he's committed as me. I see him doing other things. He's into this accounting software and this Hamptons thing. I'm worried he'll quit my podcast on me. What do I do? And my AI is really good with talking that through you, right? There are thousands of these types of conversations. My co-founder is not as committed. I don't know if my board's going to fire me, right?

9:38I'm only growing 18%. I have 11 months of runway, right? Were you Zuckerberg-ing here? How'd you know what they were saying? Is there some privacy here? How do you know what's being said? It is interesting. And on this app, you can set how much privacy you want. You can anonymize everything. But you can. No, because there's different use cases. If you're using it as a support tool, think about it using it as a support tool. You don't want anonymity for support. How can you follow up with someone? Then there's like a therapy level where you would want true anonymity. And then there's like a slot.

10:06Everything's a slider in AI, in a sense, right? The hard part here is like, so I'm so committed to ChatGPT. So it knows so much of my information because we've had so many conversations. It does. Are we all going to have like files, you know, like our health record where it's like, you know, the background? Because it sounds like I get nervous to switch to any other model because I'm like, oh, ChatGPT, I've already asked it so many questions. It already knows everything. It's an interesting question. Here's what I've learned. First of all, memory is a big deal, right? ChatGPT could already remember questions, but now it's remembering everything, okay?

10:35So it's a moat. From the top folks I've worked with on this, there's, on the one hand, it's a moat, just like being in the Google ecosystem is a moat, just like being back in the day, being in the Yahoo ecosystem was a moat, okay? But the moat is not as deep as we think. Our ability to get trained and up to speed does not require 20 million words. So yes, it is a moat, but if you decided Claude was much better for you, and the point is, and you spent a month just including your day, each day, you're probably there. You're not 100%. It's enough. It doesn't need to go as deep. Somebody had a good startup idea.

11:08I think it was Aaron Levy or somebody was tweeting this. they were talking about Plaid for AI tools. So basically, you'd be able to just like, the way Plaid lets you connect your bank safely to any financial service, to be able to take your kind of memory and context, what ChatGPT knows about you, but port that to some new app that you're trying without handing over everything, but allow it to pull from that. And I was like, oh, that's actually a very interesting idea. If apps will even exist in the future, what will exist instead of apps? We're going through a transition phase, right? We're going through, now we're chatting once again.

11:42We've been chatting since the ICQ days, okay? Since AOL. We've been chatting at our keyboards with people. Now with AI, we're chatting again, okay? But we're starting to not even know where we're chatting with, right? Right now already, like MCP is new, but already today, OpenAI and Cloud can pull from other apps like HubSpot and others and Notion and just pull the data out. I don't have to go to Notion or even Google Calendar or HubSpot anymore. As we live in this, and then people think this Johnny Ive thing for six billions crazy, it's a great bet because we're going to go from 20 minutes a day, 22 minutes a day in chat GPT to 24 hours a day in chat GPT.

12:19And when we're in chat GPT 24 hours a day, we're not going to like pull up a CRM. We're not going to use the interfaces. Like these interfaces are all already dying. You can see it, Sam. And then the amount of time you spend in GPT, like then go to like a B2B app, you're pulling your hair out. It's so dated, isn't it? It's so dated already. It's only 2025 and it's already dated to go to any B2B app. Because everyone talks about the dot-com boom and I never got to, I mean, I never experienced it. But how's this today compared to then? Those days were the first time in Tilei where people feel like they could just do something out of nothing.

12:53Like it was miraculous. It was miraculous. And that is what AI is like today. People are just doing miraculous things and it's exciting. But man, it's super scary for incumbents. It's super scary for incumbents. Do you think that everybody who is maybe a content creator at least is going to have this sort of like the stunt double, this intellectual stunt double the way that you created here? Do you see once you've did this experiment, are you like, this is the future or this was novel, but no, I don't think it's going to last. Like, where did you land after the experiment? I don't think anyone is going to create content without AI going forward.

13:26And the tools are going to change. Like, for example, just for our last big Sassler event, I did all the speaker promos in Higgsfield. Have you ever used Higgsfield? No. It's incredible. How do you spell that? Higgsfield.ai. It's the ex-head of AI from Snap. It's one of the coolest apps on planet Earth. Everyone's talking about this new Google video thing, right? But they were already doing it. Higgsfield.ai. All you have to do to Higgsfield.ai is I can take a screenshot of us right now. I'll put it in Higgs field and it will create a movie out of it. Okay. And it will, and you can, and there's a lot of different ways, but there's a really simple way you can do it.

13:58You can take an initial photo and an end photo and it will just connect them with the narrative. So I would take like a picture of Yamini just off her headshot off HubSpot and then take a picture of like a stage at Sastra. And it would just make a video of Yamini waving to the crowd and running on stage on Sastra and being excited. And so a year ago, all the, all the digital promos were a static, you know, a classic static headshot with some Chrome around it. Today, everyone got a video, a really cool video, right? Next year, it will be me and Yamini talking for 10 minutes about AI, and neither of us will have had the conversation.

14:31It doesn't mean I won't exist, but of course, that's already, we could already almost do that. But I did all these things. Now, I did, instead of waiting for our designer two months to do a static thing and upload it to Dropbox and me forget get where it was and lose the URL and have to do it. Now I just did it myself in Higgsfield in like 10 seconds. So... I'm looking at Higgsfield. This is insane. It's beyond cool. It's so good. So good. This is why AI is exciting because the really badass people can do crazy stuff now. It used to take a badass person like six months to do a prototype and then two years to get it good.

15:05Now someone epic like the ex-head of AI and Snap can like what you can do at Higgsfield will blow your mind. Dude, I can even tell just why the way you're talking. and I feel like if we did the same podcast, you know, five years ago, your rate of talking would have been like 10 beats per minute lower. Even 12 months ago. There's like a baseline level of excitement that has just risen. Like our new blood pressure is here. Well, as long as I've known Jason, you were borderline grumpy. Like you were pissed. You were pissed off all the time. Like, yeah, yeah, yeah, yeah. You were like, look, like I remember you, you weren't tired, but you had this energy of you where you were like talking to someone and you're like, you're making a mistake.

15:41stop this you're making a mistake you had like this like almost anger and you now it's the exact opposite now you're you're optimistic i am you're right i i'd like to think differently but you're probably right the way you perceive me is is important um it is i am very optimistic about it but the difference is i used to think all these products were durable that they were durable now i don't think any of them are durable so what's that mean when i first started doing this saster content i made up this expression and people have used it ever since zero to one million in revenue impossible. We don't need another product.

16:14Okay. One to 10 million in revenue. Unlikely. There's so much competition, 10 to a hundred million inevitable from impossible. Inevitable. We had a book that sold a couple hundred thousand copies. That was his theme. So you go this stage. You're saying getting from zero to a million is really, really hard. Getting from one to 10 feels impossible, but once you get to 10, it's inevitable to get to a hundred. If you had good founders, it was a punch card. Now it's disrupted. Now zero to one is very likely in AI. Like you build an incredible tool that no one else can do. You can do a million in a week or two, but like now everyone beyond a hundred million is getting disrupted.

16:51There's so many software companies at a hundred to 200 to 300 million in revenue that aren't growing anymore. It never used to happen four or five years ago. Like you hit your punch card, you had 120 % revenue retention and you had a brand. And all you need was a brand and revenue retention and products changing about every five years. Right. If the product changed every five years and you had a brand and tens of millions of revenue, that meant you had product market fit, right? High revenue, like high NRR meant it kept going as an engine. And this 120, 130 % revenue retention, it just kept going.

17:24It took you from 20 to 30 to 50 to 100. And it was like, and that's why, you know, in this crazy, in 2021, the average public SaaS company traded at 70X ARR. Public in 2021. It's almost unimaginable today when it's at five. But it's because it worked then. Like, everything just worked. And now, you know, Okta growing 10%, Salesforce growing 7%. Like, they're good companies, but it's not durable the way it once was. we've had a bunch of people on this podcast that have built apps that were really fast money grabs where they got really big really quickly with tiny teams and what i would tell them which what the hell do i know but i would tell them i'm like this doesn't seem durable but so keep your team small and ride the wave and take all the money and like do do something else eventually but it's a little bit different than it used to be because you don't need to build these huge infrastructures you can have one hit wonders that crush it you know what i'm saying uh yeah i think one hit wonders is the wrong phrasing but i know what you mean which is that you can yeah a base hit now can provide you enough cash flow to a change your life and b set you up with like five more doors whether it's investing or it's reinvesting into another company you start and you don't have to build one 20-year durable company in order to be a winner anymore so i think i think i understand what you're saying Like back when we were in San Francisco early on, like the small lifestyle business would get you the equivalent of a job.

18:50Now it gets you equivalent, you know, like you would make a couple hundred thousand dollars if you had a lifestyle business, you know, that was like, it was working, but it was small and it was solo and it was bootstrapped. Now those same solo bootstrappers have like 8 million ARR. So it's just a different scale. I remember being a - But you know what the difference is though? That is, I think that term is you have to be thoughtful about it. in the age of AI, when there's so much competition, you can build a two or three person company that's small, but the lifestyle businesses are being slaughtered.

19:20Because when these three kids come to SF working in the mission and they're working eight days a week and they've taken your little idea, but made it much better, your lifestyle is going to be unemployment. So this term I think needs to die. It made some sense, I think a few years back. It's okay. I just worry people are going to take the wrong lessons from it. You have to start in this conversation. You can do more with fewer people, but you better work harder if there's, or find a space with no competition, right? But the weird thing about AI is, is there's competition in spaces that two years ago had very little competition.

19:52Like what's an example that comes to mind? I'll give you an example. When I started investing, I did a couple investments in legal. I knew a little bit about legal, okay? And no one wanted to invest in legal. They're like, oh, it takes forever. Lawyers don't buy anything. It takes forever. It's boring. It's a slow market. Now there's a 500 AI legal startups. And there's reasons for that, but you could not be at that sleepy pace. Same thing, some of the best investment I've done has been in support in post-sales, okay? No one wanted to do this type of startup, boring, resolve tickets, pick up the phone, right?

20:23Now there's thousands of voice startups. And so if you think you can run a lifestyle customer support startup, good luck. Right. Good luck, because Delphi out of the box is probably better than you. You own a media company. You own basically a trade show that's incredibly successful. So let's just categorize that as a boring company that is not a technology company. For sure. Which is what I do as well. And Sean has done something like that in the past as well, I believe. For the people like that, how are you using AI in a not tech company? Are your employees now becoming significantly more efficient?

21:00We have five people and we do 5 million per person now. Did you have more people before that? We had much more people before. We had people in 2020 running. I didn't even, we have four designers. Now we have a little bit of a designer in AI tools. We used to have five people on this content team reviewing sessions. Now we have zero. Now we just have AI. So we have none of the designers, none of the content people. All the ghostwriters are gone. Well, I write my own content. Let me drop you off at the airport. No, but what would happen? Let me go in for sort of like, okay, nine months ago, Yamini would come to a Sastra event.

21:32She's done it three times, okay? And she would speak. and three months later, a ghostwriter would write up a summary of her thing, just terrible. And I would cry and we'd ask them to fix it because I can't do everything, right? And then about nine months ago, they got better. They actually, they got, they weren't great, but they got better and I hadn't actually used Claude. And then I went into Claude and I'm like, she just put it in Claude and they start charging us$5 ,000 a month for this. So like we can do this for 20 bucks in an hour instead of$5 ,000 and waiting two weeks with their clients, right?

22:00So we didn't need the ghostwriters we got rid of first. Then we had five people Then we had an agency we worked with for years. And for a while, it was great. And we just said, well, let's have our AI do it. And it reviewed 300 sessions, 300 slide decks, 300 presentations. Do you have to chase those 300 people down to give you the... Yeah, we still have to. It doesn't get rid of everything, but it does 90 % of the work. And it better. And it does 90 % of the work three times better. Was that you, the business owner, who had to architect and come up with all these solutions? Between Amelia, who runs Saster, this media business and events, and me, yeah, we came up.

22:32no one else was like motivated to do it. Of course. That makes sense. And then like we have a time, we have a little bit of a sales team, but then RAI just started to do all the screening and initial sales conversations for us. That helped a lot. Even for, this is a very niche business, a trade show business. Exactly. To say you do 25 million in revenue with five employees is pretty breathtaking because I would have thought that would be on the lower end. Your business would be on the lower end of like, well, I still need people, but you have just proven that not to be true. And so - It could do better with more people.