Lightcone Podcast

“I remember being in a board meeting once, you had all these board investor finance people, and you had all these salespeople were like, hey guys, shouldn't you look at this AI thing? Like, isn't that getting hot? Shouldn't you guys do it? And it's like- Yeah, actually, that literally was a question from one of our execs to me. It's like, we got to get our AI strategy. What's our AI strategy, Spencer? And I'm like, this is the wrong way to think about what it is we're doing.”

“With the capabilities of AI, because they're so jagged, it's a technology first understanding of what is possible. And so if you go to your customers and tell them and ask them what they want, they're not even gonna be able to describe what's possible. Yeah, give me a faster horse. Or it'll be asking for something that's not quite possible or possible in the wrong way.”

“We came up with an AI week, and unfortunately, for a bunch of reasons, we weren't able to actually do that until June, but that was a key pivot point. What we did was we got a bunch of the existing leaders in the organization, so our VPs of product, our engineering managers, to use this technology and to see what was possible about it. And then during that AI week, what we do is we train the team. We had one of our product leaders vibe code, like a dark mode for Amplitude in front of the entire organization, which was actually very scary, but actually it happened. They ran into a bug, but they happened to sort it out.”

“We've done two reorganizations in the engineering product and design organization since the start of this year. And so there were leaders and executives and different people who were very much in the SaaS modality, but were not on the bleeding edge of AI that just unfortunately, were not quite the right fit for what we were trying to build in the future. And so ended up having to move folks out of the business. I mean, doing that level of reorganization twice in a year is very disruptive.”

“I mean, honestly, I joke about this, but anything Google is trying to do, I think that Google is the worst B2B company of all time. And there's an incredible opportunity to compete with them. So, you know, I'd look at email, I'd look at a lot of workspace stuff. I think what the Notion guys, for example, are doing as a competitor, Google Docs is very exciting. Institutionally, Google is way too slow and way too conservative to be able to do this.”

“Before Amplitude, we started this company called Sonolight, which was a voice recognition. It was like an early version of Siri. And it had this really amazing demo, actually, where you could- it listened in the background for your voice. And this was before any of the Hey Siri, Hey Alexa stuff. So we did YC with that. We did, went through the whole batch, did demo day, did this amazing demo on stage. We got tons of press written about us, but the product and the tech, it was just not good enough. And so we ended up right after demo day deciding to shut that down.”

“The number one misconception I had was like, this would be something you learn out of a book or on a website or you kind of read about. But one, you have to do it, and two, you just want to get someone who's good at coaching you. So we worked with this guy, Mitch Mirando, who coached, was a sales exec who had gone on to coach a bunch of other companies. And then he would just come in once a week and just beat me up and just being like, hey, you don't really, what's the customer pain? And I'm like, oh, they want some dashboards or charts. Like Spencer, it's not a business pain.”

“I read Founders at Work, and one of the really clear takeaways from any of these journeys is there is a point that you get to a year, maybe two years in, where from a rational standpoint, you probably should quit. But for whatever reason, those successful ones don't. And so that is the number one filtering criteria.”

“You have to be much more disciplined about your time and say no to most stuff. And then what you realize is you become the person you hate. You'd always make fun of big company executives for not doing any work for themselves and just like judging other people's work all the time. But there's a reason for that. And you have to embrace that reason. And so it's this very hard, that is a very hard thing to unlearn.”

“Our take is that we can make the price of shipping anything by ocean container shipping cheaper by between 8 and 10 percent cheaper over the next few years. And AI is a big, not the only part of that, but a big part of that. As our business model, the way we think about it is as I call it, scale economies shared, which is the bigger you get, the cheaper you get, the more automation is a form of scale.”

“This is true of all incumbents in an industry. They have some real advantages when it comes to AI and benefiting from it. And one is the scale of the data. Two is the domain experience to know, okay, which problems should we be solving? And third is distribution. Like when we build or any large company builds a great AI product, the next day it can be used by thousands of companies. Whereas a startup doing that has to go beg people for their data to train the model and earn their trust to have that data from a security compliance standpoint.”

“It's not that we just started using AI with LLMs. We've had a machine learning model for doing planning, and planning in the sense of logistics means let's say on a containerized basis, I've got a container, which ship should it go on? So our AI for that saved us 2% of our ocean freight spend while improving transit time 20%. Usually, that's a trade-off. It's either faster or cheaper, but not both.”

“If you look now at the last two hackathons we've done, it would have been like 90% LLM based projects. I haven't studied it, but it's just my feeling and my gut. Whereas probably 18 months ago, there were like four or five. There's probably 50, 60 teams that do a hackathon project each time. I remember thinking afterwards, I'm like, you know what, we could just only do that stuff and we'll also win.”

“One of the other things that we've done is create a program for non-engineers to learn AI skills and kind of formalize programs. So your manager has to agree, but you get one day a week for 90 days. It's a 90 day program, one day a week, where we teach you kind of a AI bootcamp, vibe coding and different ways to apply. The promise of the leader who created this and convinced the managers to give up someone for 20% of their time to go into it was, I will return them to you as 10 times more productive than their peers.”

“Everyone's so worried about automating away the jobs. And I just think that misunderstands the role of companies in society. Like the role of companies is not to employ people. It's to deliver goods and services. And in fact, whoever employs the least number of people will have the lowest cost and win. And that's how they benefit society, is lowering costs and making things more available for us to buy and sell.”

“There's a period in history called the Axial Age. It's about 500 years BC. And that's when coins really started to spread. What you had with... You think about it with coins is taking transactions between two people and really making them very impersonal. And simultaneously, across the world, you had four major profits that emerged. Well, profits of sort, you had Buddha, you had Laozi, Confucius, and Socrates. They all lived at the exact same moment in time, right, as coins were taking hold.”

“I tell founders, friends of mine who raise a large round, sure, go raise a big round. As long as you're up round, you're doing good, great. Raise a large round, then do a hiring freeze for 90 days. The next day, to tell your team culturally, no, the money's not going to solve our problems. We're going to solve our problems and keep that. And then, sure, go higher. But it's because it's super, it happened to us over and over again, where you just like headcount, got out of control.”

“The things that Flexport did really well compared to all the other tech companies who have tried and failed in our space, both before we came along and in parallel, is we didn't look at ourselves as a pure technology company. We're willing to pick up the phone and solve problems with humans, drive down to the port, still to this day. And I think that's the mistake that a lot of tech people in traditional markets will fail at, because they're like, oh, if there's no API, I can't do it. If my agent is unable to do this task, I guess the task can't be done.”

“Engineering teams at these orgs are filled with people that themselves don't actually really believe in AI, don't use cogen tools, think it's all super overhyped, are really excited when an MIT study comes out saying that it's all like hype and retweeted, and really want because it's a narrative they want to believe. But the consequence of that for the companies is that they can't build the product. So if your engineers don't believe in this, then how are you going to build a product that actually works? The knock on effect for start-up centers, if you can actually build something that works, the enterprises will talk to you because they have no other options. You can't build it internally, you can't go to an established company, so the start-ups are actually getting the shot that they never had before.”

“To be fair to IT consultants, Apple is very bad at software. You know, my favorite example is Apple of the company that can have infinite access to capital and infinite access to the smartest people in the world. All of us use iPhones, and I use the calendar app. I think you guys do too. We use it many times per day due to our schedules. And even the calendar app is a piece of trash. You know, you probably run into some sort of weird bug in that, like, almost every single day. So Apple, a company with infinite resources and infinite access to the smartest people in the world, cannot make a good calendar app. So, you know, if that's true for Apple, how could any normal company, let alone an internal IT system, let alone, like, Deloitte or Ernst & Young, like, very well-meaning people, but, like, you know, most of the time, the output of something like that is bad.”

“And there's a company that I worked with called Greenlight that also sells AI systems to banks, and they were telling me a story that is exactly along the lines of what Garry was talking about, where there is a bank that they were trying to sell to, and the deal fell through, because the bank had an existing relationship with Ernst & Young, who apparently builds all the software for the bank, which is apparently not that uncommon. And they're like, well, you know, we trust our vendor, Ernst & Young, we've been working with them for years. They say that they're going to build this AI system.”

“This is a very impressive case study. The company is called Reducto. They just announced their Series B recently, and they actually closed a fan company 154 days after the batch, which is, I haven't seen that happen. And this big fan company reached out to them because they did a YC launch. That's how they found them. So our launches get people watching them. And they reached out and they go, this is interesting, we'll love to try it. And we've been working on a solution. Turns out this particular company has been trying to, what Reducto does is document processing for AI. And this company has been having a lot of internal systems and build internal solutions for years to run a lot of the operations. And a lot of the solution, they tried open source, they tried AWS, Tesserac, all sorts of OCR solutions. And they were not cutting the mark. And this is where product excellence really got Reducto to win the deal and be a pretty big one.”

“I've heard it's actually a particular archetype of big company employee. It's someone that really wants to do a startup or has always had dreams of a startup, but they're not actually ever going to do it. They're too risk averse. And so they can kind of live vicariously through an exciting startup with founders that they get along with. And if you find someone like that to be your champion, it's like they want you to succeed because they're going to feel like they're on the startup journey as well.”

“Another good tactic is to find founders whose companies were acquired by big companies and get them to be a champion. With Triple Byte, we were able to work with Apple and almost no recruiting companies working with Apple, and that was all because of a YC company, Q, started by Robbie Walker and Danny Gross, actually, that had been acquired by Apple, and then they helped us get in there. And then, actually, I remember we got a pilot with Oracle through a founder who had sold his company to Oracle and was just pushing for them to hire better engineers and helped us through procurement and gave us all the internal politics and step-by-step playbook to get the pilot going.”

“Here's one other point from the study that I also thought was really interesting in terms of why enterprise is such a big opportunity for startups. I'll actually read this quote. This is from some enterprise buyer person. We're currently evaluating five different gen AI solutions, but once we've invested time in trading a system, the switching costs will become prohibitive.”

“Yeah, so I read the tweet. The tweets are essentially, oh, Kapathi says agents are overhyped and can't do the work. So then I listened to the interview, and it's like the point he's making is, you can't just give an agent a prompt and expect it to do everything perfectly the first time. You still actually have to do lots of work to provide the right data and do all the correct context and actually do the evals and all the actual tooling. And my interpretation of that was, that's a fantastic opportunity for startups and anyone who can build software. Fantastic. There's tons of stuff that's still yet to build. And I just found it an interesting raw shark test almost. If you fundamentally want to believe that everything is overhyped, you're going to read into that, that, oh yeah, look, AI expert confirms it's overhyped. But if you listen to actually what he's saying, there's tons of opportunity to build really great tooling.”

“I mean, I think it's deeper than merely like, is it obvious or non-obvious? Non-obvious sounds like in your body might feel like, you know, neutral. But actually non-obvious feels dangerous and scary. Like I could devote my life to this, waste 10 years on this and have no outcome.”

“And they weren't wrong. I remember talking to the Lyft founders the week before they launched Lyft. And they were extremely worried that they would go to jail. And they decided to like roll the dice and launch this thing anyway. I think a big reason why other people didn't launch Lyft and Uber before was utterly that. It's like it was basically illegal to do that and they were worried that they would go to jail.”

“Coinbase to me feels a little bit less. I didn't feel like crypto was well understood enough for it to be clearly illegal. If you hung out with a lot of people who were really, really into Bitcoin in 2010, 2011, 2012, the majority of people you ran across were cipher punks who said, F the state, F the laws, and we're going to have this radical freedom through Bitcoin. That's what he's contrarian bet was. It was that it's worth doing all of this extra work for a time where it wasn't clear the market even wanted it.”

“DoorDash entered a very crowded space. DoorDash, I think, was a prime example of this, because when DoorDash started, there was actually another YC company, Spoon Rocket, doing food delivery, where they would actually cook the food in these kitchens that were spread around the city. But there was this period where we're seen as just building software is not ambitious enough, and the big opportunities are going to be in going full stack. And so in a way that like DoorDash is contrary in bet was actually say, we're just going to do delivery, we're just actually going to have like an app and a marketplace, and we're not going to try to be a full stack startup, which was obviously the right bet in hindsight.”

“So the other thing that's cool about Campfire that we're seeing in a lot of enterprise startups is with CodeGen, increasingly, you can actually bring the switching cost closer to zero. Even with the idea of a forward deployed engineer, before CodeGen, it'd be like six weeks of writing custom scripts to convert one data schema to another. If the demos are very good, that six months could be two weeks. And then if you can code gen and get very, very good at a suite of tools to convert data from the schema to yours, you could have time to value in less than a month when it used to take a year.”

“So the story of Flock Safety, I remember I was still at Initialized. We were looking at companies at Demo Day. And then, funny enough, that morning, all the cars on my street in Noe Valley were broken into. And, you know, it was a professional crew. They came in and broke into every car on our street. And the police basically said, sorry, you know, unless you have a license plate, we can't do anything. Flock Safety today solves 10% of all reported crime in the United States.”

“I mean, speaking as a VC, like maybe the best exhortation I can give to the people listening is like, do not use that. Like, you know, it's merely an indicator. The more rules you have about investing, the more ways you can basically talk yourself out of making a lot of money in venture. It would be the stupidest thing in the world to use that as your only criteria for whether or not you should work on a company.”

“And also something that people forget is that when OpenAI launched, it got mostly negative press. But by and large, most people, especially the AI researcher establishment in academia and in other companies, mostly just like were extremely negative on the idea that a bunch of like twenty somethings could create AGI. They're like, you know, we're the experts in this. We've been doing this for like 50 years. If there was a way to do it, we would have already done it. These kids don't know anything.”

“I mean, nine out of 10 people might tell you you're stupid or crazy, but then one out of 10 people might be exactly the person who believes what you believe. And then you're contrary and you become right, because it's necessary to actually attract and be a magnet in the world for all the people who agree with you. The only people that matter are the people who you care about, who have certain problems and your ability to solve it and your ability to attract all the other people who want to solve those problems too.”

“And I think Varun from Windsurf, who we hosted some time ago, he said it himself, the early stages at the beginning, the only moat that startups have is really just speed. Once you pass that and build something people want, then you figure out and go deeper into these types of moats that we're going to discuss.”

“The incredible story about Cursor, when we hosted Michael Truell to come talk to the batch, he was sharing how his product development cycle for shipping features and sprint cycles were one day. At the beginning, during 2023, 2024, around era, they would start to, every day would restart the clock and try to ship things every day. I mean, that's like insane speed. There's no big company that could ship something at that speed.”

“I think these are interesting examples because for all these AI agents, you could build a version of Greenlight or Casca or Casetext, like a demo version, in like a weekend hackathon. And I think when college students are thinking about these AI agents, I think what they have in their mind is like the weekend hackathon version of the product. And they're like, I could build that in a week. Like, how could that be defensible? And like the reason is like the version you build in a hackathon isn't useful to anyone. It's like if Casca or Greenlight fail, like the banks will lose millions of dollars. This is like mission critical infrastructure.”

“One way where this is playing out in the counter-positioning is that almost all of these house incumbents, their pricing model is they charge per seat, i.e. per employee. And this is, I think, a very big Achilles heel that they have strategically, which is that if their AI agents do a good job and actually work, those companies will need fewer employees doing this work because the work will be automated by AI agents. And in a simplistic way, that will just actually reduce, the more successful they are, the more they will reduce their revenue.”

“I mean, something sort of emerging that's very interesting in a bunch of YC startups like Avoka, for instance, they're doing customer support software, kind of like Service Titan, but for HVAC. And I think Service Titan has something like 1% wallet share, 1% of the gross transaction value of a given HVAC company, which is very small, right? But the wild thing that Avoka discovered is that they can come in as software, but then over time, they're actually getting a bigger and bigger chunk of the wallet share because they can get the HVAC people to pay them actually for the customer support piece, which is not 1% of their spend, but 4% to 10% of their spend.”

“This space is moved so quickly that in every vertical, many verticals, there's early on emerged one company that's seen as the early winner in the space. Often it's actually like the second movers, at least in the YC context, we have seen over and over again that there's advantage to being the second we were in a space. Stripe came after Braintree and Authorize.Nen, a bunch of things and was able to like actually win by just building a better product. DoorDash came after Grubhub, Postmates, various other delivery services and eventually went on to win.”

“Like the thing that still stuns me is OpenAI, ChatGBT has more consumers using it per day than Google's Gemini. I think anyone who understands the models and uses them daily would say that Gemini Pro 2.5 and Gemini Flash 2.5 are like equivalent models. Google was already one of the biggest consumer brands on the planet. It was almost certainly the biggest consumer brand on the internet. And yet somebody else came along and built the brand as the consumer AI app, and Google is like playing catch up.”

“Or try to use it to pick between two different startup ideas because they're trying to forecast five years in the future, which one will have a greater moat. Which just isn't how it works. I mean, literally, you shouldn't do that. Like a moat is inherently a defensive thing, and you have to have something to defend. Otherwise, like... If you have nothing to defend, don't worry about your moat.”