Bubble talk

0:05Today on Great Chat, we ask one of our industry's favorite questions. Are we in a bubble and is this the top? Let's hope not. I'm Ashley and I'm joined by Helen and Angela. Great chat is brought to you by Mercury, radically different banking loved by over 300 ,000 entrepreneurs. We use Mercury to run this podcast and love recommending it to other founders and fund managers. It's free to get started. Try it for yourself. Visit mercury.com to learn more and apply online in minutes. hello hi hi Angela thank you for joining us from afar in the middle of the night it's like dark and rainy here in Zurich not very glamorous I'm staying right by the train station so if you hear a train rushing by that's what it is can I just say I feel like I like lived in a dark hole for a few days and like what's been happening with Disney I feel like they've been having the worst week ever between the bachelorette getting pulled and losing like 50 million dollars and then this whole sora thing like what's happening well i think they just pulled out right saying that they're not gonna have to pay whatever they promised to pay a billion dollars they were gonna invest a billion isn't it also like the disney ceo's like first week the new disney ceo trial by fire i actually have i saw the like bachelorette saying in passing i have not gone down that rabbit hole angela hell and i do not know if you have i know i'm not a reality tv person but it seems really messy and it also seems like abc knew what risk they were taking on so that's like that has also come to light and now there's like a separate conversation about like how much reality TV is selling people's sort of trauma for entertainment.

1:50Well, that's been true for a long time, but maybe we're finally calling it like it is. No, I should say also, I guess I don't know how to calibrate content costs, but I thought the whole point of reality TV was that it's like basically UGC and super cheap. That's right. $50 million for nine episodes of The Bachelorette seems kind of expensive. Like how much are they paying these contestants? maybe the booze got expensive i don't know inflation like they're i mean it's probably like the insurance and legal bills to do this like i mean i'm not even kidding and yeah yeah sponsorships too but no angela you are our resident reality tv expert so yeah we we rely on you yeah but you know people were talking about like the cost of sora and you know people were doing the math around how much you know it was burning your day but like look at humans like like how take nine nine episodes 40 minutes like that's an insane like per minute burn costume angela you you sound like sam did when he was like people talk about how expensive it is to train ai but think about growing a human like feed it and give it shelter and electricity so i feel like you've been you've been sam filled no this thesora thing is wild i guess i'm not surprised because i haven't heard anyone talk about Sora since like that brief window of like post-launch flurry of content creation and then all the people I know who are very active there and I was never one of them just stopped uh so I don't know if there was like a a little Sora subculture like still exists on Clubhouse today we talked about Sora on like Great Chat season three and we are now in season four So Sora lasted exactly one great chat season.

3:34Wow. This is how we will measure all product longevity. I mean, I think the most interesting thing about Sora getting shut down is that it is part of a broader re-evaluation of side quests. And yes, like a refocusing of OpenAI on its core business and maybe even a little bit of a pivot into the enterprise business. And people are saying this is like the Fiji era. And it's just interesting contrast to Sam, like you can just do things. And then Fiji's like, let's just do one thing. And, you know, it is really interesting. Helen, I remember you talking about like it feels like OpenAI is like doing a lot right now.

4:17And we've sort of had these really fascinating parallel case studies of OpenAI doing a lot of things. And it felt like that was working for them for a while, at least from like a momentum and narrative perspective. but Anthropic has been super focused. And now you're just seeing the compounding benefits of being really focused on Claude and Claude Code. And it feels like they're shipping new features every single day. So yeah, like fun to watch this race in real time. Yeah. I did say that it felt like Meta was fighting a battle on too many fronts, but at the same time... Meta or OpenAI? Excuse me, OpenAI.

4:54Well, that's going to get to my next point. But I felt like to the Fiji point, Like Facebook was pretty good at being like the everything app, right? And OpenAI, I mean, it's not, you can squint and kind of see how like OpenAI could be the everything app. I mean, they've got so many ex-meta executives that are familiar with how to do that pretty well as well. So, you know, one thing that I've been thinking about is like the value of chat GPT being synonymous with AI for so many consumers. Like there's value in that inherently. And I think when I think back to like the Facebook developer platform, like 2010-ish, like Zuck's vision of like, we want Facebook to be like your passport to the internet.

5:34Well, you know, OpenAI, I would say probably has the best chance of being like somebody's identity platform across like all of these new AI tools and every like sort of AI enabled app or, you know, extension of existing software. And so I don't know, like, I still think that that is a reasonable move for them to make, which is why the focus to me actually came as like a little bit of a surprise. And there's this other narrative of like, does the way that Anthropic does accounting, is it different than OpenAI? So like, are they that much far ahead in the developer and enterprise sort of like revenue than we think?

6:13The ARR and definitional battle. Yeah, yeah, yeah, exactly. Exactly. Led by their investors. Yes, of course. But it's not a bad point, right? And so like, yes, Sora shut down. Like, yes, there's probably a focus on things that are going to be, you know, more revenue generating. But like, I wouldn't count out like the opportunity that OpenAI has to like, I think be pretty broad. Oh, I don't think anyone's counting OpenAI out, like remotely, but it is an interesting moment for the company. And I thought especially the narrative around focusing on enterprise was surprising because like you, I do think they have just such a unique advantage in consumer just being synonymous with AI.

7:00So yeah, so we'll see what happens. And maybe their PR team is like feeling the pressure a little bit because one of the things we shared this week in the group chat That was the snarky comment from the AI, OpenAI. Oh, yes. About the Ramp data. Right, right. So what happened basically is Ramp puts out these reports based on its data around like what software is picking up steam across its customers, which are probably mostly like startups or medium sized businesses. And I think it showed, I think it must have showed like Anthropic pulling ahead. An OpenAI spokesperson was asked for comments. And they basically said, well, that's like measuring like the sale of lemons by my kids.

7:49Lemonade stand. And so the funny thing is I was talking about this with a comms friend. Like you actually you absolutely can push back on that. Right. Like, you know, like RAMS customers, customer base is mostly X. These are mostly like credit card transactions. Large enterprises aren't running their OpenAI subscriptions through RAMP. And so this is not the right sample set. But because the person was salty, it drew all this attention and it looked so defensive. And what my comms friend said is this feels very Facebook circa 2017 when you're kind of on your back foot and you feel unfairly maligned.

8:30And it must have felt good in the moment to land that comment. And then later you're like, I should have just texted that to my friends. And that's why we have group chats. That's why we have group chats. That's right. There is a whole category of things in my timeline that should have been in group chats. Yeah, group chats will save us all if we let them. Group chats and disappearing messages. Always turn on disappearing messages. Oh, you know, I'm not a big user of disappearing messages, but maybe I should be. Or maybe my life is like too boring. Speaking of drama, I feel like the company that's been in the spotlight over the past week is Delve.

9:10Can someone tell, like, did we know Delve before? Like what? Like I totally missed this whole saga. Like what? It's a baby Banta. Exactly. Sock 2 compliance. And they were growing so fast. And I know from a lot of startups who are evaluating them against Banta, they were just like, Like, yeah, but Delve can just produce, you know, can go just so much faster. And it's like, and now we know why. No, I don't know if it was AI enabled. I think it was just compliance software that... I mean, I'm sure there's some AI component, but like Banta kind of created this category. There are quite a few players in it now that like the Delve drama has kicked up all this dust.

9:53There's like this meta commentary around this, is this whole category kind of a scam? And, you know, of course, there has to be a broader thing. Well, I think a lot of people feel pressure to get Sock 2 compliance maybe ahead of when they need it. And then I have learned more in the past week about how Sock 2 compliance works than I ever cared to. But it's funny. I feel like Dell – so Dell's raised, I think, like$32 million. It's a YC company. And the reason I knew about them was because they've just been super aggressive on the marketing front. I had to mute their LinkedIn-sponsored post long ago, but they were very good at activating their customer base to tell stories about how they were able to get Sox2 compliance in two weeks, which actually technically shouldn't be possible, according to some of the commentary around this moment.

10:46and so Angela the most interesting thing with this is the thing that kicked it all off was an anonymous sub stack post which you know of course some people are hypothesizing was from a competitor but but someone wrote like a really in-depth takedown of Delve and I think it was Aaron Griffith at the New York Times she tweeted it out and then it just like went viral oh and of course they're Forbes 30 under 30 so it fits into that meme like it's kind of like oh once this story got out there there was no chance for them to rein it back in and then I did think their public response was like a little bit lacking and a little bit hand wavy so Helen I was thinking about you because this brought up a conversation around like the the sort of marketing like customer logo pages and slides and Delve did something really sneaky where the logos it was showing were the customers of its customers.

11:46And like the language was like, we helped close the loop. And you think this means they're like the, you know, the SOC 2 compliance partner of record for all these like open AI or whatever it was. But it's like, no, one of their customers does business with one of these companies. And so it sort of kicked off this broader discussion around pretty hand wavy marketing. um and uh and yeah so delve is our our latest uh drama story well ashley your point about like the it being an anonymous tip blog post i mean like twitter is is hard for me these days because there is a new big account that's like a secret or excuse me that's like an anonymous you know something capital like it pretends to be like a venture firm and it just does tips are you talking three more that just popped up this week.

12:38And they're all just giving tips on who's raising and at what valuation. It's just sort of this, this is such an old school thing for me to say, but it's like, that was Facebook's value proposition back in the day to advertisers that people generally behave on the platform because they use their real name and likeness. And that's a rule. And Twitter doesn't have that, right? And now I'm kind of seeing it where it's like, you know, this platform's become unusable to me, even if the content is entertaining, It's just sort of like these, these anonymous hotlines. I mean, this is just like, like you said, like, it's like secret, right?

13:12But like, yeah, that, that crashes out pretty fast. It is really interesting because like, um, and we've, I know we've talked about secret on the pod before, maybe cause I have so much trauma from it. But the interesting thing about secret was it was still in an era where the risk was you were going to get an email from a reporter asking for comment. And so like, it wasn't like, oh, there was something on secret. And then suddenly everyone was talking about it. And the story was out there. It was like, oh, there was something on secret. I remember when Box was on file to go public and it had been delayed.

13:46Someone posted on secret, Helen, maybe from Dropbox. I hope it wasn't you. It was me. This is when the pod ends. No, the truth comes out. Yeah, exactly. I've been holding this for 12 years. Yeah. But, you know, it was like Box is getting acquired by HP. And suddenly I have like three reporters like calling me or in my inbox asking for comment. And like I was able to handle that, right, in the way that you do where you don't comment on the record, but you're like, this has no. First I have to ask, are we getting acquired by HP? And then, you know, you're like, you know, we don't comment on rumors and speculation, but also no, we're not getting acquired by HP.

14:25Yeah, yeah. And it was so much harder in the internet age, right? Because, like, I mean, it's funny that Erin Griffith from the New York Times was someone who helped, you know, give this story. It was like her tweeting it out gave the anonymous sub-staff credibility that it wouldn't have otherwise had. But, like, the story was baked well before, you know, TechCrunch eventually wrote it up. But that was, like, two days later. And, like, you know, on the Internet, we were already in, like, a different phase of, like, you know, analyzing the situation. And I think it's really hard on social media to deal with information and especially misinformation.

15:02And so, I mean, I have no horse in the sock to fight and no direct interaction with Delve and no idea how scammy or not scammy they are. But, you know, watching this unfold, you're like, oh, shoot, like, how are they going to you don't get to put this like genie back in the bottle? Yeah. And I just feel like it won't end well. and you know like the thing with these these anonymous accounts they they gain a ton of followers really fast so they look legitimate and now you know we're wasting like actual company resources like what you just mentioned that box like actually like you know responding to these sort of you know acquisite accusations like they're not when they're anonymous right it brings up a whole host of questions that wouldn't otherwise and so i don't know like if that's what the future of X is, it's like, I'm just like, not that interested.

15:55Yeah. You also don't want to set a precedent that you weigh in on every anonymous, like, I mean, and that was challenging in the like media era where you didn't want to set an expectation with reporters. Anytime they brought you any crazy rumor, you were going to guide them. It's a hundred X harder in an era where those conversations are happening publicly. So it's interesting. Maybe, maybe the best path is still the Huberman path. Just post through. Maybe Delve should have just been tweeting out its logo slide and moving on with life. But yeah, it's a whole new crazy environment. And I wonder also if there are companies paying attention to this as a potential competitive tactic.

16:38Because companies do do shady things like this when it comes to their competitors. That was sort of the point I'm getting to, which is I feel like seeing how this went down in a very short amount of time, I'm sure we're going to see more of it unless there is some action to like, you know, disincentivize this from happening. So you know what I saw that was so interesting speaking of disincentivizing? Did you guys see Nikita's tweet this week? So Nikita runs product at what we now are supposed to call X. I guess on the revenue side, they are deprioritizing engagement from audiences that aren't in the country that you are in.

17:19And it is specifically designed to prevent people in other countries trying to mess with like the US political conversation or something. If you are in like Sri Lanka, and you are getting people all riled up, you're like making money from that. So I thought that was interesting. There were definitely some Twitter influencers who are not happy about this, because I think a lot of their audience is like not in the U.S. and they want to keep making money. But it was interesting to see Nikita and X like actually make some some kind of move. Yeah. Yeah. Well, I think Elon then commented and said that this decision has been put on pause until.

17:58Oh, really? Oh, my God. Oh, my God. The chaos will continue because I think, you know, like this talk of incentives, like I just feel Like there's, there are too many incentives for, for chaos and we need to like figure it out. But hey, in this sort of like laissez-faire, you know, free markets approach, you know, I actually, I'm a little, maybe a little cynical that we're going to just going to keep seeing more of this. That is truly funny that it is already, I guess if your whales are upset, you don't want to, you know, want to rock about too much. I'm very impressed that it stayed like only focused on companies though.

18:34Like, I guess people aren't gossipy or vengeful enough to turn it into like, you know, filmers or VCs or, you know, like it's been like at the company level. It hasn't gone down to like the personal level yet. So, yeah. I think the personal level is more uncomfortable for other people to pile on to. Like you need a real, a real accusation. And so I also just wonder if like a random startup that people aren't on the cap table of is like much easier fodder, especially if you get to sound like noble and good. But yeah, like going after an individual is probably, I don't know. I mean, speaking of individuals going after each other, there was the Emile Michael headline in TechCrunch about how he still had a grudge against Bill Gurley.

19:22And I'm just like, oh my goodness. Yeah. What was the particularly artful line he said? Listen, that is a man on a vengeance mission. Forgiving would be like letting the Wuhan Institute of Virology slide back into a good reputation because the new senior manager of pandemic causation has made more friends than his predecessor. Was the number actually confirmed, first of all? That's a great question. He's undermining his own argument here. by i think he's he's playing to an audience that considers it confirmed so it works for those purposes did bill girly ever respond or did benchmark respond no i think bill is like staying out of this he's on his book tour like he doesn't benefit from rolling in the mud on this at all especially in this particular like narrative environment and let emil have his headlines and maybe work through some of this grief.

20:22Okay, let's talk about something that's been on our minds. And I doubt this is going to be the last time that we bring up this topic. But this past week, there has been more scrutiny around like the sort of math on post NVIDIA conference excitement last week, right? Where, you know, you're seeing just how big of a business that they're building and that they are like a, they're a great toll booth collector. And, you know, they're one of the great beneficiaries of this AI boom. And so people started to, more people started to dig in on the math of like how these hyperscalers, like how they're spending insane amounts of money on AI data centers and just generally like the CapEx calculations looking like something that we just haven't seen before.

21:11And so this question of like, are we in a bubble? What does this mean? What does it mean for my day to day? What does this mean for the industry? has just come up. And I think we've seen more think pieces and analyses on this. So wanted to bring it back to this group on sort of like how we're feeling about like where we're at in this moment. And, you know, yesterday was YC Demo Day, which is a great sort of, you know, it's a good marker of where we are in time because it happens a certain number of times a year, right? It's a good cadence and it's a barometer. Absolutely. And, you know, we've been meeting companies over the last couple of weeks, pre-demo day, and, you know, their expectations on raising were like, most of them were raising like four on 40.

21:51And then some of them got term sheets for 50 or much higher. And last, you know, batch, the number was more around like 35. And so you got to, you know, I was just curious, like, well, what happened between last batch and this batch that the, that the, maybe the median valuations went up like 10 to 15 million. It feels like, you know, If you live in San Francisco and you're meeting early stage founders, right, and their expectations on their valuation is just compared to all of their friends who are raising at astronomical prices, like the argument of, well, everyone else is raising at this amount.

22:24So I don't know why I wouldn't either is true. And then VCs are paying up for it. So, yeah, I think we are in a valuation bubble. I don't know enough to say that we're in a financial bubble. I definitely don't think we're in a technology bubble. I think that tech is actually meeting the moment and then some and exceeding expectations on how quickly it's moving and it's improving. But the valuation part makes me feel like something's up. I've heard from a lot of different investor venture capitalists over the past couple of months who are either seed investors or their Series A investors. And they are kind of like priced out of where rounds are ending up because they have their ownership targets.

23:06They have their check sizes and those no longer align. And so I do think it's interesting. I wonder how much of this is the like multi-stage firms that are, you know, like 30 posts, 50 posts actually like doesn't matter that much because those seed investments are not going to drive fund returns. And so it's all about having the relationship with the founder, having the right to like double down at the series A and put like lots more money in, you know, later on. And so it is really interesting to hear people sort of grapple with the fun math. I will say I'm still seeing lots of pre-seed deals in our wheelhouse.

23:47And so I think YC is a very specific part of the ecosystem. And there's still enough belief that those companies have a better chance of bigger outcomes that you can justify that math. But yeah, it just keeps creeping up each time. Do you see that in later stage, Angela? I feel like valuation has in some ways lost all meaning. I mean, I'm sure you guys, I mean, it's a public company, but the news about Meta's new exec comp pay package and the sort of$9 trillion market cap where it starts paying out value and like, I don't even know how to think about, I mean, it's one and a half trillion, which is already hard enough to think about.

24:29How do you wrap your mind around what business is worth$9 trillion? So it feels like there is a barbelling out in terms of the big companies have just reached levels and have ambitions to reach levels that we haven't seen before. And then there's a bunch of other companies at the other end of the spectrum that are all sort of competing to exist at this smaller status and have new ambitions of becoming$9 trillion companies. So it's almost you have to make your bet. Either you're going to plow capital into companies that are just going to aggregate all the capital or you're going to make your return on a different set of companies.

25:08But I don't know if maybe that's not how earlier stage investors see the market. We want the nine trillion dollar companies to Angela. I'm like happy to take one of those. Maybe even the one and a half trillion dollar company. One and a half trillion. I mean, the math on that for a small fund is beautiful. Beautiful. Okay, but even like, okay, let's, the thing I just said before about measuring time across great chat seasons, I feel like I was like in this room when I said, when we were running some of the numbers on some of these new venture fund sizes and what the return profile needed to look like in order for them to return the fund or to, you know, whatever, 5X the fund.

25:50And it's like, you're betting that there is a trillion dollar outcome inside the portfolio. we were like, ha ha ha ha, how many trillion dollar outcomes are there? And now we're talking about how like, yes, there need to be multiple trillion dollar outcomes or Facebook now has a nine trillion dollar target. I mean, that's how quickly things are moving. And yes, I actually think, you know, reiterating the point of I do think technology is moving that fast. So it is possible. But I just watching the valuations and expectations of like market sizing, like moving this quickly. Like it feels unsustainable.

26:24And I don't know. On the other hand, I don't know what will make it stop other than like a correction. Like we're not just going to be like, you know what guys have been going too fast. Like let's slow down. Like I think there needs to be like a, you know, that's not the game. Like you just, you just, you go, you go faster and you continue in one direction unless it comes to like a screeching halt. Isn't it wild that it wasn't like that long ago that we had our first trillion dollar market cap company. And like, that was such a big moment. And then suddenly there were two and suddenly there were three.

26:53And now we have these IPOs lining up for later this year where, you know, it is possible that one of those could hit a trillion. And so it's just like things are changing so quickly. I do think that like the reality is tech is just a much, much bigger part of the economy. And so we are going to see, you know, some of these crazier outcomes. The question is how many. And then I think we're in this era where a lot of people believe that the value capture is going to be concentrated in just a few businesses. That's why people are still scrambling for Anthropic SPVs. It feels like compared to 2025, 2026 is only more concentrated, right?

27:34We've had the massive Anthropic and even bigger, like$110 billion OpenAI raise. Feels like every day there's a new fundraise. I think Kleiner announced another$3.5 billion today, which is actually modest compared to what A16Z did recently. And so, yeah, these are big numbers. And I also think these big numbers are more responsible for the valuations we're seeing than, Angela, to your point, any real evaluation of value, right? These mega funds have to plow meaningful amounts of money into companies that can absorb it and where they believe that they can have that significant outcome that moves the needle for these funds.

28:22And so I don't think this has any sign of slowing down unless, yeah, there is some bubble. We hit some bubble or some moment that makes everyone recalibrate. I mean, I think we're in the bubble. I mean, most people agree. We are in the bubble. It's really the question of like, when's it going to pop? Oh yeah, that's what I meant. Yeah, yeah. Like, are we at the peak or when was the peak? Was it behind us? Is it still ahead of us? I think it's probably still ahead of us, but like, I'm a little bit of a doomer thinking that it's probably closer than a lot of people think it is. But to your point about like, what the valuations being meaningless, like you also have to like understand that the mega funds, like they just, you know, they have a business too and they need to win competitive deals.

29:07And if it means that they offer a better price that wins the deal, then they'll do it. Right. And so like you just see how these things sort of compound. And it's not any one chaos agent's fault as to why we're here. But I think to me, the definition of a bubble is like when the rules of the game are such that like it creates this sort of situation, you're in a bubble. And that's what I think we're at. And your question about what do early stage investors do and what do early stage founders do when they look at this is a good one. And I just continue to go back to I think the best thing about being a really early stage investor and being early in your founding journey as a founder is that you can choose paths, right?

29:50The farther you go in the journey, there are fewer paths available to you. And the best thing about starting now is that you can decide that you're going to have a cash flowing business that maybe needs a little bit of startup capital to get off the ground. But you want to be in charge of your own destiny. And so getting to profitability and having your choice about what the road ahead looks like is the advice I give broadly. And I think what's great about being a pre-seed investor is that it's not crazy that a founder takes your check and then they never take more money. It's interesting being here in Europe when I'm walking around the streets in London, in Zurich, people are not talking about open claw.