Rick Heitzmann: The AI Boom Is Forcing a New IPO Wave

19:33Rick Heitzmann:Hard to – a complicated transition to make if you have to swap your business model, take risk, especially on the public markets. Yeah. I have to reinvent myself. I sell a new product to an existing customer that I'm going to charge less for, and they're going to want to buy less seats even if they live with that seat-based model. So my new product that I'm going to take years to develop, I might be able to sell to an existing customer for less. That's a bitter pill to swallow. Changing gears to another important topic. What do you make of the block layoff? You know, 40 % of their employees. Is that specific to block or you think more to come?

20:14I think that's going to be a trend. I mean, Shopify famously has been able to grow revenue over the last couple of years, 70 % while keeping the employee base constant. You know, there's two things that are happening there. I think people grew very inefficiently during COVID. And this might be the last piece of saying, all right, let's reset ourselves. How many people do we need, especially if we're here in person and we're able to act more efficiently? All right, that might be, how did we look at the end of 2019? And how were we able to generate efficiency then? And then on top of it, where can we add agents?

20:50So I don't know many companies who have done that analysis incredibly well. You know, obviously companies I'm on the board of that are smaller, easier, more centralized, they're doing less things. It's easier for them to do that analysis.

21:03Rick Heitzmann:But you think it's more overhiring than agents creating value today? I think it's less focused on efficiency in the near term and then more agents over time. So if Block takes out 40 % of their employee base, that employee base might not grow that much. Right. They have the confidence of sort of what they're trying to lean into in terms of new capacity. And famously, X, not only do they take out a whole bunch of people, but they didn't grow that employee base. Yeah. X will be fascinating. Unfortunately, we will never necessarily know how well he did with that whole deal or not. I can't imagine SpaceX really has to get granular about the performance of X.

21:43It'll be interesting. It'll be interesting to see. I mean, they'll probably have to report revenue numbers. Maybe that's about it. I don't, I don't, I'm not a, thank God I'm not a securities lawyer. What, in terms of sort of the AI deal making.

21:58Rick Heitzmann:Yes. That's been happening. I know you have some point of view on that. Like what, I don't know. Revenue round tripping, I guess, is the most like a charged way to put it. But like, to what extent do you think these deals, you know, including like now with Amazon wanting OpenAI to use Tanium chips are, I pay you to spend money with me. Yeah. And then there's some of it's off balance sheet with private credit that I'll give you some equity. You'll agree to buy my stuff. We're also going to put this private credit thing that's off balance sheet. We somewhat opaque that opacity has never been proven to be good for shareholders.

22:37Right. And transparency is the key thing to, if you're not really transparent and straightforward about something, what are you trying to hide and why? And it's not because it's so awesome that you want to hide it from the public. So I think that's going to be a case. I think the thirst for AI will let this go longer. So every new technology that was incredibly CapEx intensive crashed, whether that was the railroads or television or fiber and the internet. At some point, we will overextend. At some point, we will overextend because no one ever says, maybe I'm okay. Maybe I don't need dessert tonight.

23:19Everyone's like ice cream and all of some gummy bears and hot fudge. And you continue to get rewarded. As the hyperscalers have said, I'm doubling down on AI, I'm spending more, stock price has gone up. So not only is there an ego thing of like, hey, I'm looking next to me and I got to keep spending to keep up with my fellow hyperscaler CEOs. My stock price will go up. When I do that, my stock price goes up. And I remember going back to the old infrastructure, Celex, in the early days of building out software that they came up with a metric of you got valued at seven times whatever CapEx you put in the ground.

24:02And that became a metric for next generation telco companies and internet providers in 2000. So, you know, what's the incentive there? You will put as much as you can in the ground.

24:15Rick Heitzmann:It's insane in retrospect to think internet providers were once obscenely valuable. Well, and then every dollar, I mean, once that metric came out that every dollar you spent, he was worth$7. Your only incentive is just to spend as much as you can as quickly as you can by getting dollars however you can get them. And I'm not sure we're at that point now. I think you're still seeing GPUs be somewhat. But there's no dark GPUs. There's no dark data centers. What do you mean by that? It does not be used. When people were laying fire, they called it 1 % was dark. It wasn't being used. We haven't lit that up yet.

24:57And therefore, we're building things for tomorrow. And don't worry, the customers will come. And eventually, you just build too much. Because right now, NVIDIA has been great about projecting. There's more demand. Yeah. We could seemingly set whatever price we want. Exactly. So they said, you know, in that analogy, the dark fiber, people were overbuilding, saying eventually the customers would come. And they did. It was just five years too late and all the companies went bankrupt. In this case, they are saying as many chips as I could produce, someone's buying. And actually, I could charge whatever.

25:32And there's a line outside the door. You know, they're also investing in companies to buy their chips and round tripping it. So you don't know what real demand is, but until you see a little bit of stopping of real demand and some of that demand being driven by, you know, consumers, some of them driven by enterprises, maybe a lot of that demand being driven by fear in both enterprises and hyperscalers, it's not going to stop anytime soon.

25:59Rick Heitzmann:But you think it has another year or two years? Yes. I just think that given how we're seeing some of these companies scale, given the fact that the capital markets seem wildly open for them, even as OpenAI has hit some choppy waters, they still continue to raise a lot of money at very high prices. OpenAI in particular is insane because it is still a nonprofit. Obviously, they resolve some of the structure. But they're still going to court. That'll be interesting. Another interesting disclosure when they go to court, I think, later this year, right? Yeah, they're in a big legal battle with Elon.

26:34Rick Heitzmann:Yes. Yeah, you can see how Anthropic is a less complicated company to invest in. Yes. Where they're selling to businesses, they're more eager to be profitable. Whereas OpenAI, it's like weird structure, all the legacy problems of being a first mover. Yeah, they have some consumer-facing stuff. They have some agentic stuff. They have some infrastructure stuff. They're going to get into specific agents. You don't have a bet, right? You're disinterested in this? I don't. I'm not in either of those. And then how does that stack of, hey, there's going to be consumer agents. Some of those consumer agents are going to be shopping.

27:07Well, they have a commerce strategy, but they don't have commerce revenue. They have an ad strategy. They're just getting into ad revenue. They're going to do specific roles for you. They're going to be involved in health care. No one's quite sure exactly what that means. Are they going to be your payment agent? Are they going to be your health care coach? Are they going to be your doctor in your pocket? and then you're like wow these guys are kind of trying to do everything which was you know why sam did the code red a couple months ago it's like hey we're doing everything and we can't win

27:40Rick Heitzmann:we can't fight a battle on every front i mean i i wrote a piece being bearish at 157 billion which seems clearly wrong now yes um i'm sure you're gonna edit that out yes no i'm still bringing it up i can't help myself you know i wear my uh my mistakes on my sleeve i mean it's just very hard to call you know pricing tops in new markets i mean i remember when silicon valley was super divided about uber at three billion and thinking that was you know so far ahead so it's really hard to know you know we will overextend given the nature of how these things go but then you you know the markets can the i forget who's drunken miller or sand or julian robertson the markets can remain irrational longer than you can remain liquid so betting against a market especially a runaway train market like this is really, really hard, which is as in the venture markets, how do you kind of, you know, don't get irrationally excited about it.

28:38Stay close to your knitting, continue to invest in the strategy that you've committed to yourself, your LPs, your partners about, and continue to do good deals in that space and good things happen.

28:50Rick Heitzmann:How are you playing out in the private markets right now or where do you see investable opportunities in AI? So we see really three different things. We've begun to see consumer AI experiences. And actually, we do a lot on the consumer side, even as people have abandoned consumer. You did Riot Games. We did Riot Games, yes. We're a seed investor in Riot. And they've gone on to be the biggest video game in the world over the last 20 years. So gaming is really interesting, that non-player characters can be agents and therefore it makes the game more fun. Assets and games can be created much more cheaply.

29:26So we've looked at a lot of things there. We've looked at things of what are next generation social networks and social constructs. A company called People that does, you can have a digital twin who exists out in the world as an agent for you. You can have a digital friend. You can have a digital pet. And what does that mean as these other things that are out there in the world, whether your pet or your digital twin, are now interacting with yours? so like hey digital rick go go digital eric and go get an ice cream cone at the yankees game

29:58Rick Heitzmann:and then send us pictures and that's i can't claim to understand what's the appeal or what's the it's so different than roblox of you you're you're basically able to have something else that extends your social social experience and everybody we were hanging out i guess we're friends now you know it's like we've done that we outsource we were in school we were working We're doing a podcast. Send our digital twins over to Yankee Stadium and send us pictures. So you can be social even when you're working. But don't I – what I want from being social is primarily like the experience, right? You're getting old, man.

30:38Rick Heitzmann:I want the networking of it or like the – Or you want the experience or do you – there's obviously a whole generation that's grown up that want the likes and the friends. and you want the sort of the adoration like hey look at little Rick if we're going to the aquarium performatively on Instagram like why go at all just like have the if you could get likes thumbs up points for having a very cool picture of little Rick and little Eric at the Yankees game that's just another way you could do things when you don't have to do anything accelerationist view of getting rid of Instagram Well, I think there's going to be agents.

31:19It's not just a flat photo that you're hoping because you put a good filter on that people will like. You have people doing things for you. So whether it's the dating market, the social media market, the gaming market, the commerce market. We've done a lot in digital health on the consumer side, which we think is going to be transformed because health care is so broken. AI could definitely make that better, faster, and cheaper. And then on the enterprise side, we have two sides. We call it application layer enterprise AI that we talked a little bit about having data and proprietary data, proprietary access to data and the ability to use that data to create new applications, new agentic applications.

32:03We're also seeing companies like CRISP who create enterprise applications. So think about supply chain. If I'm able to have an agent, if I'm able to have my data and your data, you know, I'm Mondelez, you're Walmart, you know, somebody else is Unify, which is, you know, the logistics company. How do I understand how that works and then be able to put intelligence on top of a multi-hop data system? And you think about the lock-in of that if you're across enterprise and the data network effects which would exist and why that's really important. And then even the third bucket might be on the infrastructure side.

32:43A lot of Fortune 500 companies are still trying to pull data out. A lot of these companies have grown through acquisition. They have a spaghetti of all kinds of different databases, software languages, all that stuff. And they're just trying to unlock that data. And there's a whole generation of companies that are doing that. And then what happens when you unlock the data? Holy shit. you know, is that secure? And who gets access to that? You know, so now I have this data. Who's getting access? Are the agents getting access to that? How long does the agents have? So if the agents has access to your credit card and financial information all the time, well, what is that agent doing?

33:26What are they going to do and why?

33:28Rick Heitzmann:I'm very interested in what company is actually willing to play gatekeeper for agents spending your money, right? It's like, is it the actual application? You know, you can see existing, you know, like a plaid or something, but it doesn't seem like a lot of companies, I think, are wary about signing up to be the one that says, yes, this is safe to let your agent go spend your money on your behalf. And where does that data go, even if it's just your bank account number? Like, oh, you could go buy me that sweater. But, you know, so what does that agent give that bank account to another agent who gives it to another agent.

34:02And what does that mean? Or if it's even worse, social security number, which you can't just cancel and get another one tomorrow. How does that data exist, whether it's payroll data? So we think about that. Do you have a bet in that space? Yes, we have a couple of different things around that. We have Daytona, which does sandboxes around agents to keep them more secure and running specific things. We have LM Studios, which manages data around this problem. We're looking at a lot of things because one of the best and worst things about agents is they're always working. So they're always working.

34:40Rick Heitzmann:I know. You hope they're doing good stuff. If they're doing bad stuff, they could really drive you in the ditch. A new anxiety we're inventing right now is this sort of when you wake up, you have to see what your agent did. It's like, have I been productive overnight? Did I set everything up so that the nighttime agent work was super useful and I have stuff to do today? Or did they take my credit card on a shopping spree with Toto Eric at the Yankees game? I mean, games and consumers are obviously the most fun. Yeah. What in gaming in particular are you optimistic about or have you made any bets yet?

35:15Rick Heitzmann:Not recently. We looked at some stuff in terms of content creation. We looked at some stuff in terms of pure content creation. So if I wanted to play you in a game, we could create a much better game than we ever could have, right? So there's kind of next generation Unities, which is game creation, game engines. Game engines are able to leverage existing IP, and they work with companies like Disney or Epic. They leverage existing IP. So that game creation is going to be easier, no different than it's just easier to create applications or anything else like that today. So we've looked at those things.

35:49it's hard to create a lot of long-term competitive differentiation so there's a lot of these companies are doing it they're all creating a lot of value but it's hard to say where they're creating a long-term competitive differentiation so we haven't done anything there uh we have companies that are creating um really good agentic non-player characters right so npcs is you know the term for them they used to be you had an npc it was like it was like i know well mpc we would

36:17Rick Heitzmann:trivialize humans by saying you're just an npc exactly now we're gonna have agentic npcs so you know those were like the stormtroopers you're gonna go straight ahead and you're just gonna shoot them and now these npcs are gonna be smarter and there's people who are making smarter npcs and that's kind of interesting but again there's lack of differentiation you know we're really probably most excited about on the game side because i think that more people are playing more games and maybe even some AI will let people work less and therefore play even more games, and we both love playing games, is the opportunity to create independent studios, right?

36:55So Riot, despite being enormous today, was an independent studio who said, you know, we were able to bring League of Legends to market for about$6.7 million, which is -

37:05Rick Heitzmann:You would just bet on a game studio? So I think there's going to be small teams of game studios who have different play patterns. We've looked at things around audio being a really interesting thing because the ability to use AI to ingest audio and come back with audio. So are there audio formats which are really interesting and different? Are there interfaces that are different? You know, we looked at a company that does kind of audio formats on TV. So can you play us? Yes. They're some of my best friends. I host events with them, so I know them very well. There's a couple. Yeah. Really good.

37:43So are there different ways that hopefully I'll meet somebody that'll say, you know what, we thought about all these things, and we believe with a handful of people, we can come up with a game. It'll be completely different. It'll be audio or voice driven, and it'll be super cool and fun. And we'll pull into all these things. It'll be social. They'll have Enchantic NPCs and they'll do all these things. But oftentimes we spend a long time looking at megatrends and figuring out what's going to be it. And then you have to find that special entrepreneur the way Brandon Beck and Mark Merrill were at Riot who were like, ah, you're actually saying what I've been trying to articulate for years.

38:30Rick Heitzmann:This is an old one and we don't need to spend a lot of time on it. But the whole like major league gaming, like you guys made a play in that world? We didn't. Because I wrote about it and I got excited. Founders Fund, didn't they even make a play? Like the sort of, you know, Activision went big on having more broader eSports. So there was a whole bunch of things in eSports, a whole bunch. So there was teams, there was leagues, there was some other things. It died, right? Like it's not happened. If not, the pandemic went, right? Well, it died a little bit. There's two reasons I think it died. That different than like major league baseball or the NFL, a lot of the rights and the IP was so controlled.

39:11By the games. By the games and the developer publishers that you really were just playing someone else's game and you had less control.

39:20Rick Heitzmann:Because then they got built sort of as marketing for the games. Exactly. But then because it wasn't a real business, they didn't have to set it up in a way that weren't. And you can only play in certain places. and then therefore the rules were very much dictated by that. And there were some very big companies. 100 Thieves is still a really good company growing in that space, 9 Liquid. There was a lot of companies that were very valuable at a time, but their ability to gain audience didn't let them capture as much value as I think people would have thought. But they're still really active. You know, Riot sells out any kind of regional event.

39:59I think they sold out. They had a three-day champion worlds in Paris this past summer, and it sold out in eight minutes. So people still love it. People still enjoy watching it. I think that just the developer publishers are capturing a lion's share of those economics.

40:16Rick Heitzmann:A big portfolio company of yours is Rowe. Yes. I mean, it's a fascinating space given the rise of GLP-1s. What's your read on sort of their business today? Yeah, I think, again, we're talking about megatrends, thinking about the rise of GLP-1s, the rise of direct-to-consumer healthcare, the ability to put the patient first. So one of the things you look at as venture capitalists, incredibly large markets where people are paying a lot of money and are very unhappy with the service. And you think about healthcare being very clear in that, that people pay a lot of money for healthcare, either as an individual or even an employer, and government being the other very large payer who doesn't feel like they're getting their money's worth.

41:02The patients are unhappy. The payers tend to be unhappy about everything. And so how do you fix that? I think it's around six weeks to get an appointment with a doctor today in America. So if you think about all the things you could use help on, And when you get an appointment, you get about, you know, 10, 11 minutes. So it's not like, you know, people really, really have the time to really provide personalized health care. You know, can you do something in a better way with remote access, with AI to help that provider know what's important to you and why? And then to solve very large problems.

41:45And, you know, GLP-1s and, you know, obesity have been a huge problem for a long time. GLP-1s have now been proven to be very helpful in a lot of other conditions. Are they still able to produce them? Because for a while they had the opportunity. So there was... There was like a scuffle there. Yeah, GLP-1s became so popular that there was a global shortage. If there's a global shortage of a necessary medication, the FDA says, okay, we're basically going to suspend the patent and we're going to let people compound their own drugs. And then when that shortage is over, it flips back and says, hey, you know, you got to respect the patent again.

42:27And that's been a lot of chop in the market with different players. You know, we're always been incredibly good about respecting those patents and saying, okay, when we need to compound to be able to fill that consumer demand because there's excess demand, we'll do that. And now when that's not there, how do we partner with the Nova Nordics or Eli Lilly's of the world to provide the best possible medication to the right person at the right time at the right price? And that's moved on. And they've continued to provide a great service. They're kind of, Morgan Stanley put out a great note saying they're kind of the Spotify of direct-to-consumer healthcare that you're getting a premium service in a white-glove way and you feel like you're being treated like this is a really good service and you're a patient and you're being treated even as a patient like a customer finally.

43:21And I think the growth of Roe, kind of riding that tide of GLP-1s and that being maybe the most important drug in the last 20 years is going to be a great story for some of the stories they tell and how they've changed people's lives, the ability to access health care, and then from a digital health side, how it's really changing the way that pharma companies go to market and the impact they're able to have.

43:48Rick Heitzmann:They can get the consumer excited about a drug that's good for them and not wait for doctors to serve. Yeah, well, they hear it and they go see it, and then if they want to have access to it, they have access to it. There's now a bunch of different formulations. You could get a pill. You could get a shot. You could go from Lilly or Nordisk. There's different providers, and therefore you have the ability to say, hey, this one's right for me. It's differentiated in some way. Yeah, well, there's enough providers out there that this one's right for me. Hey, I'm afraid of needles. I'd rather take a pill.

44:17I'd rather pay less or pay more for maybe a different brand, different than anything else. But here's what I definitely know I want. I definitely want to be able to take this drug, which has proven to help people lose weight, push off dementia, have a lot of these other knockdown effects. And I want it delivered to my house. And I want to be able to do my regular checkups. I want to be able to check on, make sure there's no side effects. I want to be able to do that right through my phone and get what I believe is a very high quality service. and that's you know this next generation of companies are able to provide a very high level of service delight their customers and then have you know a huge viral effect first mark was an investor in draft kings yes have the prediction market companies have you made a play there we have not and not because we were in draft investors draft kings we looked at some of them maybe because we were concerned maybe about the regulatory environment you're like we were deep

45:17Rick Heitzmann:in the regulatory environment and these companies don't seem worried about it. We lived through it. Right. And got and felt some of the pain of that. And these companies didn't seem concerned about it, which gave us some concern. Do you think they were always, the prediction markets were always sports betting companies in sheep's clothing, basically? Well, I think now you could look at the, and whether it's the banks or everyone else doing research um i think it's 85 plus percent of the prediction market stuff definitely now it is i mean for a little while they had a success about you know elections and everything and then all of a sudden and then i mean i was even over in europe for a board meeting um during the nfl playoffs and uh i wanted to look up i think it was you know how long the pm how long people thought the PM was going to stay intact in Great Britain.

46:10And I go to pull up my prediction market and I couldn't even get to how long that PM was going to be there because it was eight pages of NFL playoff ads. How do you think there's eventually a crackdown? I think it's going to be interesting. I think the current administration is not going to crack down. I think that might change under different administrations. You know, the key thing is California, Texas, and Florida, a huge part of the U.S. that were, you know, at least California, you know, actively voted to not allow sports gambling. And then now sports gambling exists. So what does that mean?

46:57Maybe under a different administration or someone's different way to view it, or has this completely just, you know, horses are so far out of the paddock that it's like, we can't turn this back anymore, you know, which I think would just grow that market tremendously. I think, you know, DraftKings just had their investor day and they said, you know, maybe we were not on our front foot on the prediction markets because we were trying to be so good on the regulatory side, but they were able to buy a company and prediction markets. And they believe this has grown their TAM tremendously. So I think all those companies will benefit from it.

47:37You know, Flutter and DraftKings on the current public side with their brands they have today, but then obviously Calci and Polymarket. Are you still a DraftKings investor? As an individual. I still own a lot of the stock I got as an individual as part of FirstMark, but we've distributed all our stock.

47:56Rick Heitzmann:This sort of political swing is interesting, sort of closing out with, you know, the IPO conversation again. I mean, do you think some of these companies, yeah, they sort of need to go public in a Trump era, like the SpaceX and Elons where he's so close to the administration? Yeah, I think it helps them. You know, I don't know if you have to rush out before the midterms. I don't know if the market cares that much about, like, rushing out before the midterms. You know, what do you were, you know, if something happened in the midterms and then that emboldened, the Democrats to say, hey, these prediction markets are, you know, we're after them.

48:35We're after them. And that's going to be a huge piece of our 28 platform. I think, you know, if you're a poly market investor, you'd rather be out than in.

48:44Rick Heitzmann:Right. You'd rather have just already gone public. Already just gone public. And at least I always think, and this might be a bias I have, that if you can get out and you think there's going to be something rough ahead, you're better being out where you have more access to the financial markets, even though it might be harder or more expensive than you might have thought. And you at least have some liquidity for some of your early employees and investors. They might not be happy if you hit a speed bump with the stock price, but they'll be more happy than if you were private. So you just have greater degrees of freedom to operate.

49:24You have to provide more transparency. But if you're uncertain about what's ahead and you have the opportunity to be a public company during that time of uncertainty, and obviously if you have a great CEO, you want to be out.

49:36Rick Heitzmann:Rick, thanks for coming on the podcast. Awesome. Thanks. Great seeing you, man. That's our episode. Enjoyed having Rick on the podcast. Thanks for sticking around to the end. Please, if you've made it this far, you've got to like, comment, subscribe. Excited to grow the channel. And of course, you can find our writing and reporting on startups and venture capital at newcomer.co. We also host events. You can check out what we're doing at newcomer.events. Share the podcast. Help us get distribution. Support the channel. Comment. Support. Like, comment, subscribe. You know the deal. Thanks. Thanks for being on the journey with us.

50:11Rick Heitzmann:All right. See you next week.