I see a large IPO on the horizon.
You're surrounded by journalists. Hold your position.
That's misinformation.
>> Courtney declaring order inbound.
>> Let's just roll.
We are surrounded by general. Hold your position. Come on. Get up.
Trust the experts by we are experts.
Founder mode.
>> Five good.
>> I see multiple journalists on the horizon. Stand by.
>> UAV online.
>> Double blaze. Triple blades. Double kill.
Raw. Cook is cut.
Team Deathmatch. We are experts. Triple blades. That's just wrong. Right.
>> Clearing order inbound. >> Come on, get up.
Position. Retreat one.
Cup is win.
Team deathmatch. We are experts. Triple blade. Let's just roll. Right.
Mark clearing order inbound. Get up.
>> We are surrounded by journalists. Hold your position.
>> Strike one. >> Strike two. Activate golden retriever mode.
>> Marky clearing order inbound. >> Five good. Founder.
>> You're watching TVPN.
>> Little preview there for you. It's Wednesday, May 20th, 2026. We are live from the TV Ultram Temple of Technology, the Fortress of Finance, >> the capital >> capital of capital.
>> Uh we're both in suits today. I like it. It looks good. Some uh we can't tell us they can't tell us apart. You Is that what you're saying?
>> That Okay. Anyway, >> but people have been saying I need something like a square jar. >> Okay. >> When I don't wear a suit. >> Oh, okay. I'd like that. >> And uh so so yeah, something something to consider.
>> Yeah, >> it it should probably be a pretty big jar. 20 bucks to the OpenAI nonprofit. That's what you got to do. Um, no, we weren't unheard. Uh, we I mean we can pull up the full uh the full post later in the show, but uh there's a funny I mean it's a good good analysis of like uh you know what we've taken from ESPN, what works about live streaming. This is from Alice Key. Uh my latest for Unheard is on why tech shows like TVPN are rerunning the sports media playbook invented by ESPN and why it's working. And they said, uh, where where is it? They're hard to tell apart in their matching suits and floppy haircuts. I don't know if that's good or bad, but uh, we don't always have matching suits. Usually Jord's the casual one, but yesterday we were both casual. Today we're both in suits.
>> We do we do still get the brother thing a lot, >> but I think it's a term of endearment. I think it's positive. Anyway, uh, we have to react to Google IO, of course. There's a whole bunch of announcements. Uh, some really exciting stuff. some stuff that people are having mixed reactions to. We'll take you through it all, but first we need to watch this video about humanoid robot ramp.
>> Why is ramp? >> No, I'm kidding. >> No. Uh, uh, Sun Drip Woo says, "I'm crying. Watch this video." >> Let's get some. >> You got to go to the beginning. You're spoiling it. You got to go to the beginning.
>> Doing pretty well. Moving pretty quickly. Little bit of a Oh, catches itself. Catches itself. Not bad. Not bad. Okay. Seems like a full recovery. Seems like you're ready to go. >> Another one. >> I'm liking it. >> Yeah. >> And then >> not good. And then I I wonder what it's thinking because you would think that there would be >> Oh, they got to cut the music. They got to cut the music. Don't let the music keep playing while your your boy's down.
>> It just gets carried off like this. It's so crazy to just carry it off like this. Like this. This seems solvable. This seems solvable. Do do tele operation take over in that case or even just have like a if on the ground reset yourself by like doing a push-up like you could sort of hard code that I would think. Like that was a weird I I understand that it can't dance and walk upstairs. Like that is tricky. I understand why it fell down but I don't understand why it couldn't get back up after it fell down. Anyway, >> singularity delayed. Yeah, it's it's possible that uh the robot died from embarrassment. >> Yeah. Or maybe it was damaged. It's possible that like as it as it hit the ground, it was just actually taken out. Um but uh I'm fascinated by this. I wonder how many like this video was taken clearly in like the third row of an auditorium. Are people going to see robot dance shows? Was this an introduction to a tech conference? It's like I need so much more context here to really understand what's going on. Aaron Chen posted the full the the first video. Aaron Chen AI. They had the chance to stop this when the robot almost fell, but then it really fell. But the show must go on. So, who will dance on the floor in the round? Uh, beware. Well, uh, certainly solvable with enough time, but not quite there yet. If you're thinking of shelling out big dollars for a robot dance show, maybe uh make sure there's a refund policy in case the robot absolutely collapses. That would not be good.
Anyway, uh Google IO, bunch of different announcements. Uh Brandon Guerell on our team uh posted on the TBPN newsletter some reactions. Sort of bucketed it into four key areas. Intelligent eyewear. This is an interesting one. I want to go into this. Gemini Omni, we talked about the videos, we played a little bit of that yesterday. Uh upgrades to Gemini LLMs, those have been mixed reactions from developers. We'll go through that. And then anti-gravity, uh which is an interesting place with an interesting history. So, uh let's start with intelligent eyewear. Uh if you had to pick Warby Parker, Gentle Monster, have you heard of Gentle Monster before?
>> I've heard of Warby Parker. I know the story. I I I'm a fan of of the business story. I super familiar. I I haven't worn glasses in a very long time, so I'm not really in the market, but uh the Warby Parkers, I've always I've always enjoyed the way they thought about the brand. And I've also been impressed by the way they've built that business. They were early to the DTOC boom and then didn't some of the founders move over to Harry's. Is that the same team or is that a different I'm thinking of?
>> Maybe they were certainly when I when I think of DTOC, I think of Warby. Yeah. I think of Allirds and I think of >> Everlane. Maybe. Yes. But when you think of two, the last two out of those three, the market caps are sub 100 million. Allirds was trading at what 20 million or something and then spiked because of the AI thing. But uh Allirds Everlane uh not really sustainable businesses. Uh Warry Parker on the other hand uh current market >> has been sitting at three and a half >> billion. birds has pretty much been down only >> down only since the pump on their Neocloud. No surprises there.
>> Well, have they given us an update on how they are rolling out Kubernetes? How it's going building their NeoCloud? Did they get allocation? Are they racking Cerebras? Are they racking GB200s? What are they racking and how fast are they getting power? >> Would be it would be funny. >> Jensen ends up having to talk about allirds on the earnings call today. There's a new >> Yeah. Yeah. Didn't they also like fully rebrand the name? Was it going to be like bird AI or all AI? Like they were moving >> all birds still exists. >> Okay. >> But they basically kept the public entity. >> Yeah. >> And that's what they're building the Neocloud through. So >> fun. Well, >> lots of fun.
>> Uh Orby Parker, uh resilient. I mean, uh in 2021 it was a $6 billion company. Now it's a $3 billion company. uh not the best scenario but surprisingly resilient I think in a time when a lot of people wrote off a lot of the standalone direct to consumer it was like either get rolled in to a bigger company or uh go or or like face the fate of the public markets but uh Warie Parker has a deal with Google and Samsung uh Google says we're partnering with Samsung Gentle Monster and Warby Parker on new intelligent eyewear. Here's a sneak peek at two designs from this fall's upcoming collections. Uh, and people are This is I like Futuromics from Sam. Kind of crazy that you can wear your favorite Mag 7 on your face. Now, uh, you can. Uh, the Gentle Monster one does a really good job of hiding the camera. I I imagine that it will have a light to tell you if it's recording, but if someone wore these from a distance, uh, also met Ray-B bands have done sort of the hard work of becoming the first face computer. So, when you see Ray-B bands and they're a little thick, you start immediately thinking, "Oh, should I be looking for a uh for a camera lens? Am I being recorded?" Uh, but the gentle monster design, the silhouette, doesn't scream technology. It doesn't scream uh wearable face camera. And so, uh, these are going to be a little bit more stealthy. Um, Warby Parkers look nice, but the camera pump on this, if you zoom in on the Warby Parker, >> it see, you know, it makes a lot of sense that that the Googles and the Metas have to go and partner on different silhouettes. Yeah.
>> My expectation, my uninformed >> uh expectation is that Apple will just make Apple glasses, right? they will probably uh I I it's hard to see them taking the route at least early on of partnering and allowing another company to influence the design language, but it makes a lot of sense that uh that uh Meta would partner with Luxodica.
>> Okay, first look look at the camera bump on this. If you zoom in as far as you can, I don't know if we can zoom in any further, but uh the the camera is actually not flush with the frames. It's actually protruding a little bit. Yeah, you can see it right there. Uh interesting design choice. I wonder how that will catch the light, how that will reflect in in uh in the real world. Um is that true about Apple? I thought that they had a partnership. I thought someone I I saw this on the timeline um that Apple would be uh partnering with someone on Frames. Maybe that was just a rumor. I don't know where it went. Uh okay. So yes, I it Oh, Apple has Carl Zeiss, which I guess is like a glass manufacturer. I don't know. Is do do they do they make glasses? Glasses.
>> Carl Zeiss. AG. German manufacturer of optical systems. >> Oh, they make eyewear. They have an eye collection >> in Germany in 1846. >> Okay. So >> optician Carl Zeiss. >> The Zeiss eyewear collection. I don't know. Um, but this was all from a joke uh from Abdu says, "Okay, so Apple has Carl's Ice, Meta has Ray-B bands and Oakley. Google has Gentle Monster and Warby Parker. Boring. Which company is going to be bold enough to slap wearable technology into some 3M safety glasses?" Would you rock these, Jordy?
>> I like safety glasses. You know what I'm talking about, right? You're working with a buzz saw. Dust in your eyes. >> But these these look cool. These are sporty. I'm I'm much more likely to just commit to the bit. computer. Full full cyberpunk full. Yeah, full. You'd clank out. Full full cyberpunk. I think that's uh I think that might be the move. I don't know. For some company, a challenger company could potentially do that. Maybe friend or something. Uh anyway, um what else? Uh so, Warby Parker traded down on the news, which uh Shield Monot was surprised by. Why is Warby Parker down 14%. They announced a partnership at Google IO that's been in the works for a while. Is it because they aren't available yet? Um and uh and our friend Rat King M Isaac says, "Okay, Google AI glasses with Warby Parker are officially coming for Meta Ray-B bands." Google also said it would bring Gemini to glasses this fall with Samsung electronics and and the eyewear companies Warby Parker and Gentle Monster. The glasses which work similarly to Meta Ray-B band smart glasses come with a camera, microphone.
>> Yeah. At what point at what point does Google just buy Warby Parker, right? It's a $3 billion company. It's actually done quite well over the last uh over the last six months. It's up 43% the last uh 6 months although it's been um almost flat uh this year. Uh I would say I expect that uh uh smart glasses are going to have product market fit among people that need to wear glasses first. Right? If you already have to wear glasses all day long uh for your vision, uh why not throw some smart features in there? it's going to be harder to get someone uh that doesn't need glasses to add a new device to their rotation, right? Yeah.
>> Um and so uh you know, Warby Parker's, you know, done quite well. >> Um and uh has been surprisingly resilient, but they have, you know, incredible distribution. Uh and I wouldn't be surprised if they get sniped at some point. I mean deeper integration into a traditional I don't know like workflow like like a lot of the Google IO we'll get into this but was talking about uh Spark the personal AI assistant and when I think about >> Spark >> it's an AI >> oh just look at Omni >> there yeah there there's a lot of names it's Google there's a lot of products uh you're referring to of course Nathan Clark's post it's it's in Gemini just created an AI studio oh it's for your personal goo Google account for workspace, you need Gemini Business. No, not Gemini Advance. That's AI Pro now. Unless you need AI Ultra. Oh, agents, you do that in Spark, actually. No, not Gemini API management. And it's the typical uh meme. But um the uh the the interesting thing is that I I do think uh meta raybands like it was always like okay you have a deep integration with WhatsApp you have a deep integration with Instagram DMs maybe Facebook Messenger some people are still using that but uh in terms of like wiring into your life there are way more people that see Google Docs Gmail as like the central node in their personal life like pe like people think of like I the all the stuff I have saved on my the desktop of my MacBook is like my core repository. A lot of people think okay for the important stuff I'll put it in Google Docs or Google Drive and then most things flow through Gmail. Most things flow through iMessage. There are some people that just are like yeah WhatsApp is the number one screen time for me. That's where I really organize things. But Meta doesn't really have this like knock-on effect of like, oh yes, you've you you're it's not necessarily an enterprise level productivity suite, but there are people who are like, yeah, I'm using Apple Mail, iMessage. I save my files in Apple files, Apple, you know, the the cloud storage. I use my my camera roll super important. So an an AI agent running through the Apple ecosystem can be valuable and an AI agent running through the Google ecosystem can be valuable. the meta smart glasses. It's a little bit trickier to go and do anything that because you're just like sort of bumping up against the walled gardens, right? Yeah.
>> But investor Nick doesn't like them for aesthetic reasons. He says these Google X Warby Parker glasses are horrific looking compared to these Meta Ray-B bands. Someone is probably going to lose their job over this. I don't know that they look that much worse. I don't know. Ray-B bands are a very uh iconic silhouette and they do look good. So, um we'll see. We'll see how the response goes. I think uh from a product perspective, there's obviously fertile ground. On the flip side, uh the red the the the Wayfairer is just such an iconic it's more iconic than anything Warby Parker has produced. And that's just sort of the reality of brand building over a decade versus a century or something like that. However long Rayban's been around, long time, right? Uh anyway, uh Genie3, you can now simulate real places by grounding Genie 3 experiences with street view imagery. Google is sitting on a a motherload of real world data. I was always thinking YouTube was going to be so valuable for Omni and V3, V4 maybe in the future. I hadn't considered Street View as a trove of data. Demis seems very datailled. He seems a lot of the Mag 7 CEOs seem very data pelled. There's that story about Mark Zuckerberg uh screen recording or or or re logging all the computer use from all the meta employees. The these important troves of data are increasing in value and street view certainly seems like it's one of them. Uh this is cool. I wonder how interactive this will be. How how this actually instantiates into a game. It's a great demo. What does it take to >> Yeah. Do they allow people to build games on top of this? Yeah, I just think about I don't know. I I mean, Deus has a background in games and he was sort of alluding to the fact that he might go back into games at some point or or at least be able to like scratch that itch again. Famously, he uh wrote uh uh a programmatic code to generate vomit in a roller coaster simulator. Very fun story. Uh but again, when I think about uh Roller Coaster Tycoon, which was not I think I I don't think he was actually working on that game. It was a similar uh theme park simulator. uh we are moving back into the simulator world. Uh but the mechanic is what is so enticing to gamers often when I think about the games that I've spent a long time with. Some of them have incredible graphics, AAA graphics. Uh some of them have 2D graphics, but the mechanic is great. And so that is what gets me to >> the legend Bobby Chipman in the X chat says, "Can't wait for smart glasses to fully replace my monitors." >> Yeah, maybe you'll need augmented reality or something. Meta Rayban display certainly going that direction. The Orion I' I've been surprised. Uh wasn't the first episode we ever did? We were talking about Orion >> uh and they still haven't shipped it, right? >> I mean, they shipped the smaller version, the Meta Rayban displays, which have sort of the the Call of Duty HUD. It's not full augmented reality. I was expecting >> I was expecting we we we've demoed the Orion headset >> and it and in it has a bit of a narrow field of view, but it really can put a screen right in front of you. Uh, and I I assumed that, you know, everyone was saying it's really expensive. It's clunky. It's not ready for prime time. But, you know, look at how fast things are going. In a year, maybe two, we'll get it. And maybe that's coming at the next MetaConnect. Maybe this summer we'll see it. But, uh, hasn't haven't been that many rumbles on it. And then obviously the massive p pitch uh shift to AI capex uh might have taken a back seat. I don't know. I'm certainly hopeful. Uh, I'd like AR and VR. I think uh I think we're we're we're overdue for a new fun product. I'm still waiting for the next Apple Vision Pro, Apple Vision Air, something just lighter. That's all I want. Cheaper maybe, but lighter and same screen. Sc screen was great. Uh anyway, >> we know John >> Gemini Flash 3.5 looks pretty neat according to Tennibbrris and extremely fast, but still largely the sort of incremental progress we've come to expect from Google. Generally a pretty disappointing IO. Now, what's interesting is that uh Gemini 3 felt like a new base pre-train. Felt like it had some of that big model smell. Felt like it was uh you know, really delightful to talk to. And I think a lot of people were expecting Gemini 4 here. We're still waiting for the next iteration here. And also, >> yeah, we're still waiting for Pro. >> Yeah, Pro isn't out. Uh but people are speculating that that 3.5 Pro won't necessarily be a new pre-train. And so, uh, it seems like it's a there's a little bit of, uh, you know, research being at odds with like the product cadence like Google IO is scheduled probably like two years in advance and whether or not the training run finishes on time is a little bit harder to uh, package up and nail on a specific time. We see this with uh the the uh the independent labs or the open AAI anthropic the other labs XAI like they're launching models very much like when they're done and then they will like instantiates like something that looks like a conference around it or maybe a video or a blog post um a model card. Uh but if you're if you're grinding towards a specific date uh and the specific model isn't quite ready uh you come out with something that's looks a little bit more incremental. people were really really uh honing in on the fact that the uh the the cutoff date was January of 2025, right? Was that the was that the date that or was it December of 2025? Either way, yeah, I I don't know. I don't know how much cut off time cut off dates matter because you know all these models, you know, can query the web and and get up. But >> yeah, overall reactions from uh developers across the board were not good. Not good at all.
Yeah, here it underwhelming. Cursor ranked it on Cursor Bench. It is below Composer 2.
Is that a fair thing? I mean, I'd like to see you rank another live stream on TBPN Bench. It doesn't match up, you know. It's like...
No, I mean they have Bench all the other, they have all the other frontier models, and some of them are ahead of Cursor's own models on Cursor. It's just one data point.
But here's the other thing. It costs four times, it underperforms Composer 2 even though it's roughly four times more expensive.
Interesting. Interesting. Yeah, I feel like for a long time Google's positioning was, you know, frontier or near frontier but best possible pricing. And this marks sort of a shift in the strategy perhaps.
Yeah, overall starting to make more and more sense why Google has put so much capital and resources behind Anthropic.
Yeah, seems to indicate that DeepMind is constrained by data rather than compute for what they intend to do. Hence the TPU sales. Rest of Google now shipping their org chart. Ben Thompson talked about that a little bit, and there was some context on like, you know, we were asking the question like, will there be AI fatigue from stuffing AI in every product surface area? Ali K. Miller shares one of the loudest applauses in the entire Google keynote. Nisha put on the Gentle Monster plus Gemini glasses, tapped the side to summon Gemini, an all-in-one prompt said, 'Take a photo and put a cartoon blimp in the sky that says Google I/O 2026.' And within seconds, the preview of the edited photo from Nano Banana appeared on her watch. 'I want to spend less time on screens. AI is really coming everywhere.' And so much is driven by voice AI as the interaction mode. Very cool demo, impressive technology, but Greg's Gadget says these companies truly have no idea what regular people want because, yeah, that is a little bit of a niche use case. You need to be more creative with it for when you would actually use that because this is a perfect demo of the product and the functionality, but it lacks that creative spark of like, yes, I did want a picture of that on my wrist at that key moment in time. Like if you're not doing a demo, I don't think his point is that regular people would not be excited about that particular feature. Right. Anyway, you want to move on to something, where do you want to go?
IPOs. SpaceX IPO. We're getting more details by the day.
Oh, the other thing that was going pretty viral. The last thing on I/O was that the Google anti-gravity team flashed a Codex folder in their actual demo video. Gurgley says, 'I had to do a double take in the second minute of the launch video for anti-gravity. You can see people use Codex on the anti-gravity team. Did no one double check the launch video at the very least?'
Um, not a huge surprise. Obviously anti-gravity looks, a lot of people were saying looks quite like Codex. So clearly more than Windsurf, they would have just rebuilt Windsurf. I don't know. We'll have to see.
No. So, but anyways, this isn't a huge surprise, right? Google's been using a bunch of Anthropic models. Clearly they're using a ton of different models and products internally.
What was that drama with Steve Yegge going back and forth with Demis about like what teams are using what models and stuff? There was a big back and forth, a big dust up on the timeline like a month ago about like whether or not Google's employees were deploying AI efficiently or broadly. Some of them aren't and some of them are, and Demis chimed in and said like this is just completely wrong and everyone's using AI. I don't know, goes back and forth. There's also people are benchmarking Omni Flash, which looked amazing when we saw the videos. There were a few like little quirks. Some people in the chat were saying that the firing order of the V8 was not correct. Maybe it was only a V6.
Yeah, it was missing two cylinders.
It was missing two cylinders. But it looked good to me. I don't know. But now people are actually comping it to Cance 2.0, which obviously has much looser content restrictions. Because I guess just like Hollywood can't file a lawsuit in China. I'm not exactly sure how that works. Because Cance seems to be available in America. It seems like maybe...
Oh, I think Chinese businesses have been relatively immune to US copyright law for a very, very, very long time.
And it also might just take like years to file a lawsuit, do discovery, actually go through and litigate.
Oh yeah. You can just go like, there's malls in China where you can go to a Nike store. Yeah. And Nike has nothing to do with it and yet all the products... They've been selling Swatch APs over there for decades.
Yeah. Just ask Rolex and Patek.
Yeah. I've heard fake cars too. Like you can get a full replica of like a G-Wagon that's just made in a factory and then you could buy it, bring it over here and you take it to a Mercedes dealership and they're just like, this is not a Mercedes but it looks like one, like to the millimeter from the outside, but internally it's just frauding. Anyway, Cance 2.0 looks great. Omni Flash looks great as well. These are both like super useful. We'll see how they actually play out and how they get implemented, how they get used. The interesting thing will be like at what point, it still takes a long time to generate videos. Very hard to get them right. The last 90, like we're at 99% fidelity, but when you click in, you start noticing little details. When will we be in a paradigm where you ask a question and you actually get an explainer video 6 minutes, 10 minutes, like you would on YouTube? Very computationally expensive, very difficult to maintain the logic. Like what is the deep research report of Omni Flash? These 8-second, 10-second, 20-second videos are impressive but not perfectly substitutable for a 20-minute YouTube video because of the time and the level of detail that you can go into. Some people that are looking for information about a V8 engine, they want a breakdown that lasts 20 minutes. And so, that's the next benchmark. We got to move the goalposts. Anyway, SpaceX IPO, the prospectus is incoming, according to Zero Hedge, as soon as May 20th. That's today. We will see. Goldman lead left. This was a surprise. Michael Grimes has worked with Elon for a long time at Morgan Stanley. There was some back and forth. He went back to Morgan Stanley. There's a question about whether or not there would even be a lead left because it was such a big IPO. Maybe they all share equally. Obviously, they're all going to make a ton of money off of this. So, good news from start to finish, but it is interesting that Goldman was selected. Do you have a soundboard cue you want to play?
Okay. Katie Roof has a scoop.
The scoop athlete of the century, Katie Roof, has a scoop on the biggest venture returns ever. Founders Fund and Valor are set to make more than 60 billion in gains on the SpaceX IPO. Sequoia have more than 20 billion.
Yeah. Was this somewhat of a reaction to D1 getting a lot of credit earlier in the week, right? They're set to generate roughly 20 billion in returns, and maybe some of these other funds thought to put their hand up and say...
I don't think, I think that at this scale like there are so many LPs in these funds that are getting updates and they've known the numbers for a long time. They've known the ownership, the holdings, and you do some back of the envelope and you get to some pretty huge numbers. Will be very interesting. Huge for Sean Maguire, huge for Luke Nosek and a lot of other folks over at Founders Fund and Antonio Gracias at Valor and all the other Founders Fund folks. Really like Sequoia, Founders Fund...
They needed a win. I mean, you go back like, you know, these investments were made like 2004, 2010. Like it was not obvious. Certainly there was no Starlink narrative when these were made. There was no space data center narrative. This was a rocket company that was blowing up rockets left and right and not quite getting to, you know, massive business. So you really had to be a believer and they were.
They benefit. PY was having some fun on the timeline. He said megafunds are too big to generate returns. They're basically just be collectors, and of course they're printing.
Well, we have a very special guest joining us today, the CEO, the founder of Bigma, Dylan Field in the waiting room. Let's bring him into the TV. Dylan, great to...
Hey, great to see you. Thanks for having me back.
Too long. Great to see you.
First question, are you alive or are you dead?
It is such a white pill. Why? What? I mean, how have you processed all this? I guess we can go into the business and the product and the financials, but I'm more interested in just like the mindset, the emotional journey, like how do you stay sane? Is it an advantage to have been building this business for so long that you can draw on other experiences? Like what has the process of going through all of this been like for you?
I mean, what's the quote when in the age of AI, those who stay sane win?
Oh, I like that. I think that that's pretty accurate and yeah, you should own that.
Yeah, I think I'm going to attribute that to you.
We're going to attribute that, not mine. But really I think it's super important. You have to have an even mind about it. And you know, I know that you guys know this, but social media is not reality.
It is not reality. We see this all the time.
Yeah. All the time. But yeah. And the thing that I've been thinking about this week is how AI-generated design is seemingly becoming as easy to clock as AI text.
It's not this, it's that, right? I feel like people kind of... like right now, I mean the entire market is just going through like rolling stages of AI psychosis and everyone's having it in different ways, right? So like if you do like equity research, you're obviously going to go through a period of AI psychosis because you can just like type in a few words and get like what looks like, you know, days, weeks of really, really great work. And sometimes it is great work. And if you're a designer, web designer, and you can type in a few words and get what looks like a great website instantly, you're also going to go through that period. And you can just see these sort of like pockets of AI psychosis all around the industry broadly. But it does sort of like fade at some point and then if you're able to stay sane, you know, through those periods and not let yourself succumb to it, you can get excited about the potential and benefit from it but not lose your mind in the process. And I think like we were seeing that with, you know, early on with ChatGPT and like writing, right, before it was super obvious to clock. And then we were seeing that with design specifically where, yeah, when you can just go and just generate stuff, design that looks like good design. I'm not saying it is good design. And I think we're kind of coming out of that now where you can just instantly clock when something was generated. And sometimes for a specific type of asset or use case, it's totally the right move to just generate something quickly. But I think we're already in a period right now where if I go to a startup's website and it's clear that it was just like a one-shot, you know, prompt, like that says something about the company. Could be good, could be bad, we don't know.
Jordi, I think you're absolutely right, but seriously, I mean it's just the balance... Oh, thank you. I think it's the balance we had to strike with Figma's design assistant, design agent launch today. And what we tried to do was make it so that we can really help you bring the context from your file into the agent's context window and with that do things that maybe are what will be most helpful for the designer in the moment. Spawn agents to do explorations, variations, do the sort of rote tasks that are more boring. So things like design system maintenance, for example, how do we make it so that you're able to do that way faster with Figma's design agent than you would if you're just manually clicking on all these variable names or manually changing a bunch of components or text translation or any number of things. And so, you know, it's been really fun to kind of explore how does the design agent work and where can we make it better because there's so much more we can do to handle more of these tasks that, you know, designers are able to elevate themselves and work at a greater level to actually push aesthetic past AI slop, to actually push past, you know, the cliches and get to real innovation, solve real UX problems for users rather than making it so that you just have to go do these boring tasks to just get through and unblock people or unblock your team. And I think that the design agent will really help there and bring more of it into the canvas where then you can share and collaborate with your team or with other agents. So, I'm excited for where it goes and I think today's really big and kudos to the team because they've been working really hard making it so that we're able to take all the advances with models and LLMs and apply them to design. You know, getting these models to speak design well has been non-trivial and very cool.
That's awesome. I feel like a lot of software engineers who are interacting with coding agents are experiencing like incredible amounts of like bloat in the code and there's questions about like, are you laying really shoddy foundation that will come back to bite you later? And a lot of people say the answer to slop is just more slop. In the design context, are you feeling pain from Figma users where design systems... or are you at least designing the agentic design tool in a way that is aware of that potential pitfall in advance? Because there is a world where, you know, oh the drudgery is taking my website and, you know, doing design iterations for every single possible viewport or device, and if that gets out of sync that can be very complex. At the same time, if you wind up, you know, instantiating all that and then it doesn't tie together, you could wind up with some sort of, you know, misalignment and like it could be very difficult to manage in the long term and you could wind up with like bloat.
Yeah. I mean, look, I think that it's different context for sure when you're in a sort of codebase and you're using Claude and it's over-claiming and making stuff up. Or you're using, you know, OpenAI's ChatGPT models. 5.5 is amazing, but like definitely over-complexifies things. And, you know, these teams will deal with the sort of problems that are being identified and they'll fix them and figure them out. But overall I think it's a different sort of class of problem than like when you're in a design file and really part of your mode is exploration. If we can help you with that exploration, that can be useful for the user. And in terms of semantics, I mean the way that you represent a component or, you know, the way you apply auto layout, even, these are all examples of things that we've had to get more rigorous around. How do we build evals for this to make sure that we're doing things in a way that is both clear and opinionated but also doesn't get in the way of the user. And I think the work here for us has been evaluation side that, you know, we have to always be improving on. And part of the feedback that we'll get from this beta will be towards where have we missed the mark and what do we need to do better, and I think there's lots of room to grow there. I'd also say that I think that when you look at products like Weave, there's so much potential to not have necessarily this pylon effect but rather to explore more with AI. The outputs of the models are like clay, you can mold. And, you know, it's like how do you take stuff that might be a one-shot slop that's generated but then actually apply it across multiple models in a pipeline in order to actually figure out a way to make it into what you want? And sometimes the best way to do that is to get on a canvas and use your hands, sometimes the best way to do that is through a workflow you define, other times it's by going back and forth with an agent. And I think you have to be very smart about what you do when, and it's our responsibility as a tool creator to give you all the options.
Dylan, what are you excited about in terms of using AI to help augment the creative process? Because when I think of part of what I think is exciting about this moment is like, I don't... right now, right now it's hard, like a human still has to have like the fundamental kind of idea, right? And we know that because you have to like type it into a box, right? You need to do the typing still, so we're still necessary there. But when I think about like iconic startup brands, I think of Linear. I think of some of what like Cursor has done over the last couple years. I think of like Figma's brand where there's so much work that went into just like coming up with that, coming up with like the concept and the look and feel of the brand and then like way, way, way more work in terms of like instantiating that brand across every possible surface area. And part of what I think is exciting about this moment is like we're potentially entering into a...
World where you can spend more time just doing that, like heavy creative work, coming up with like a concept that can stand out. And then once you create a system, then you can like much more rapidly scale it across again all these different surface areas, whether it's like digital products, web, you know, email, ads, etc. Um, how are you thinking about like augmenting and giving creative superpowers on that, on that like brand building process when you're kind of coming up with, coming up with like the idea?
Yeah, I think there's so much we can explore and do there, and using design systems effectively is hard. We certainly have not perfectly nailed that. I think we've done much more on the maintenance side, and we have a lot more coming on the execution side of how do you put the design system you've created together so you can fully productionize that. And it's one of the things that, you know, we are looking at from so many angles, and it is critical context to be able to figure out across code, across design, how do you use this effectively. But brand, I think, is something where on the Weave side we see even more exploration with these workflows and people being able to define workflows that are more canonical and then branch out from them. And I think we'll see the same with Assistant and with Figma's design agent as well. And I also think that this idea bank, content bank, can be established and added to over time and by more people. And what we're seeing with the last quarter results that we announced is that more people are starting to use Pigma in the organization. Thank you.
There are fantastic results.
Thank you. But the breadth, I think, of the usage expanding is a huge part of that. And the more that you can get it to the point where people can come in with their ideas, contribute to the conversation on a canvas, collaborate with others to refine, doesn't mean that, you know, these like sort of non-designers who don't know all the conversations are being had about the UX or the brand are going to be the ones that are bringing the idea, that's the idea they're going to ship to the table right away. But I do think that bringing more voices into that conversation, more viewpoints only helps.
How do you think about the concept of Gell-Mann amnesia in like the, in the context of like staying sane in the age of AI? Like, because there's probably non-designers who use a GenAI tool and are like, oh wow, like designers are cooked. And then there's real designers who use a GenAI tool and say, oh, there's, you know, yes, it's maybe 99% of the way there, but the battle is to get to 99.999, maybe 100% of what you can do. And it feels like staying sane all, all outputs, right? If you ask, if you ask AI to generate you something around a topic you know really, really well, it's a different feeling than a topic that you're just learning about.
Yeah.
Yeah. I mean, I think that basically folks are in a place where they learn by comparative cases, and the more that they see great examples next to sort of output that's not as good, and they can learn why that's ultimately where you see folks then push further, and, you know, they can then take it on to do more in an area and evolve their skills and their judgment and their sort of taste or not. But I do think it's really important to have those conversational loops going and also get your arms around as a design leader what is happening in your organization, because what we're seeing is more people prototype things, and then, you know, there's like this latent fear of, oh no, what are we going to ship? And is it the thing that is sort of like a random exploration that someone thought was real, or is it something that is like gone through our cycle and is ready to go and is like very well thought out. And in general, the other thing that I'd mention is it feels like there's a lot of tunnel vision right now. You know, I think that there's certain models that kind of went, oh, let's learn from 40, and, you know, are sort of synoptic and doing that, especially for engineers, and get very latched on to an idea, and then it's kind of like the AI psychosis sets in, and you're like, this is the best idea ever, and you're fully tunnel vision towards creating it, and you can show massive progress, but like, are you going the right direction? And I think it's really important to steer and to actually be building what you need to build and to think, not just wear a thinking cap. So that is something that I think is critical right now.
Yeah. And I completely agree. Yeah, there's been a bunch of great takes about that. Help me, if I meet someone and they have the next great app idea, and help me pitch them starting with Figma as the front door to the experience if they're a non-designer but they're also non-technical and they're hearing about Codex and Claude Code and those like codegen tools versus starting with a design, starting and having the design be the place where the idea person sort of instantiates their app or their product, and then yes, there are agents and they might use different tools for, you know, spinning up backends and doing other things. Why should Figma be the front door to building?
Yeah, I think that as you're exploring an idea, part of it is thinking it through and actually understanding what is it that I want to build. And you should not just start in Figma, you should go talk to users. You should go think about like, what is it that I'm trying to build? What are my goals and the problems I'm trying to solve? So that might look like starting a sketchbook or a doc or in live conversation.
I think there's a lot of people that have not asked that question to the LLMs and jump straight to, I spend seven nights staying up all night. I'm cutting this thing and forward. It's fun. It's box effect of like, will it write this time or not? And that's variable reward. Yeah, very overreward Skinner box. And so, but then I think going from there to being intentional about, okay, I know what problems I'm trying to solve. Let me think through the solution. Figma provides an excellent spot to do that with in conjunction with, you know, brainstorming tools like FigJam or, you know, document editing, spec editing. And from there, I think when you've got your direction, you know what you want to build, it's a great time to use our MCP, go bring into code or go bring into Make.
Yeah. And Make itself, well, you'll see a ton of evolution on surface ahead. We've already had a lot that we've improved, and you'll see even more in the weeks and months coming. And then I think that's a great spot where you can go back and forth because sometimes as you build it out, you realize that actually there's more to explore and define on the design side. So use the best tool for the job. But I think don't be tunnel vision about there's one direction and that's the only thing that can work. That's where you'll fall into traps.
Yeah. Last question for me. Huge revenue growth, massive success on the financial side. I'm interested in going a click deeper and trying to understand what is working the most because it seems like there's great retention among big clients and customers that are using Figma at scale across a huge organization. There's also new entrepreneurs that are a team of one and they might be using Figma. There's almost consumer cases. The heaviest users of coding agents are heavily using Figma because they're making more software and they need to add stuff to it. I mean, I look at some of the stuff we've vibe coded and I'm like, okay, like it's time to level up and have Figma in the flow if it's not already. Even because you don't want some, even like the dashboards, like these things need to be designed at some point as they grow and scale. The first like most basic thing, it could just be a CLI, but as soon as we start building UIs and multiple users, like it gets more complicated. But what is driving the growth? Is it both? Is one more of an opportunity for you over the next few years? Like what are you most excited about?
I mean, I think we're seeing growth kind of across the board right now, and we're very thankful for that. But our eyes are always in the future, and, you know, the same things that I've said when I've been here in the past, and, you know, I think have become their own memes around design being the differentiator and design as the layer above code. As code commoditizes more and more, design is, I think, increasingly the battleground. This is where everyone is going to really duke it out to figure out what the direction is that they should explore and what they should go build. And we need to build for that world, a world where design representations, code representations can live together, and a world where, you know, you don't have to make these false tradeoffs between direct manipulation and AI or, you know, being able to explore broadly or make fast progress. Like both is the answer, and so that's what we're really building for right now.
Exciting. Well, thank you so much for taking the time to come on the show. Have a great rest of your week. Great to see you. Good to see you all. Thanks for talk soon. Goodbye. See you. Up next, we have Feross from Socket back on the show to discuss a Series C by none other than Thrive Capital. We'll bring in Feross from Socket from the waiting room to the TBPN Ultra Dome. Get that gun ready. Welcome to the show. How you doing? What's up? Good to see you. How's it going, guys? Thanks for having me back. I think last time you were on, oh, we'll see you soon. And here you are. Give us the news first. What happened?
Yeah. So Socket announced its $60 million Series C at a billion dollar valuation. Did it again. Did it again. Yeah. Led by Thrive Capital. Congrats. And it's a big day for us. Absolutely massive.
Is what you do important right now?
Yeah. You know, you couldn't have really asked for a better, you know, perfect storm. You know, you have just the confluence of, you know, AI, cybersecurity, and all the different attacks we've been seeing. And we kind of built the perfect product for this moment. And it's been something that we've been building towards for, you know, it seems like it's okay. You know, we just, we had the right thing right now, but the reality is, you know, we've been building towards this for, you know, four years plus. And so, you know, it's an overnight success that took four years, right?
There we go. I love it. Yeah. What was the key unlock for the new round? Growth needed the capital for specific expansion, just the broad tailwinds of the industry, people looking forward to more demand. What's changed in the business?
So the business is just inflecting. You know, every year I feel like, you know, okay, this is great, you know, certainly we're going to get like a regression to the mean or something, but it just keeps accelerating. So over the last while, you know, we've had 500% plus ARR growth over the last 12-month period. And, you know, it's just, it keeps going. So I think it's kind of, we're seeing kind of three forces converging all at once here. We have this kind of like perfect storm between, you know, AI generating more code than ever before. You got developers and increasingly non-developers too that are, you know, pulling in open source dependencies and third-party code at like an unprecedented velocity. And you have more code being written than ever before, both by humans and AI agents. And the kind of maybe counterintuitive thing is it actually brings in a lot of third-party code, and that code is really vetted less than it's ever been before. And that's kind of like the first thing. The other thing is you got the frontier AI models like Mythos that are finding, you know, thousands of high-severity vulnerabilities across, you know, every major operating system and open source library. And so the total volume of vulnerabilities in code is exploding right now, and it's going to keep going as, you know, we start to throw these models against, you know, this backlog of code that's been out there for a long time and has a lot of vulnerabilities. And then kind of the third piece is that the attackers have really started kind of realizing that they can exploit the software supply chain to get into companies. So they're not really like coming in through, you know, finding vulnerabilities, but they're actually kind of realizing if we go into an open source component, that's actually an easy way to get into an organization and not just one, but usually thousands of organizations that all use that same component. So it's this like perfect storm of all these factors right now.
So walk me through how your business model interacts with the open source community, because when we see things like Mythos finding a bunch of bugs in open source repos, you know, we hope that everyone will just submit bug reports early for the good of the world. But I'm interested in like the economics of patching open source. If there's a really important package out there that maybe doesn't have a Linux Foundation behind it, like how at some point, like at the very least, there's inference costs. Who's paying to patch all of the open source software that so many different companies rely on? If you go to a big company, they pay you. They might patch like their dependencies, but how does this all flow together into an actually safer internet?
Yeah. Well, I've been in open source for I think over 15 years now, and I maintain a bunch of open source packages and, you know, have a real soft spot for the open source community. And so let me tell you, they were suffering under an enormous burden even before the current, you know, AI trends have kind of made it even worse. You know, they don't get much support, whether financial or just in terms of, you know, the number of people working on this really critical infrastructure is super low for how critical it is. And so, you know, we started seeing a lot of AI tools being used to create these slop PRs, these slop issues, and that was already kind of causing a burden. And then now you have the various frontier labs that are, you know, finding with their models are finding a lot of vulnerabilities that, you know, were there, were always there, but the community didn't know about. And I will say they're doing a good job of providing a patch along with the bug that they've identified. So they're doing a lot of the work for the maintainer, but it still takes effort to review the PR. And the thing that people don't realize is when you accept a pull request as a maintainer, you're not just kind of, it's not this one-time cost. You're actually accepting the burden kind of indefinitely into the future to maintain that code and make sure that that code is secure and, you know, is kept up to date with all the other changes. And so I think what we're going to see, and we're already starting to see, is despite the kind of best efforts of the frontier labs, and they are making good effort, you know, maintainers are going to fail to actually accept these patches. And so we're going to see a lot of libraries that may have a vulnerability in them that like literally have, you know, effectively, you know, code sitting on GitHub that anyone can look at and use to generate an exploit against that vulnerability. And that's, there's not going to be an easy path for companies and for, you know, developer teams to go and actually patch their libraries because there's no version they can upgrade to because the maintainers just sort of sitting there on that patch and hasn't accepted it. So it's just creating this real risk where, you know, like there's more volumes than ever before and there's not really like easy paths. And so this is actually something we try to solve. So we built a solution to this called certified patches. And what this is is basically kind of a deterministic way to in kind of one click make your vulnerabilities and your open source code go away. So we use a whole bunch of AI and kind of produce these patches that make the vuln go away without any work on the part of the developer. There's no kind of burden of upgrading packages. You just sort of in a click the vuln is gone from your dependencies. And you can kind of keep the rest of your code in your application the same. And so it's really a kind of a quick fix to the problem that we hope, you know, is going to be part of the solution. And by the way, we're giving away all the critical patches to the community for free so we can try to disseminate this widely to as many people as possible.
Are there, are you able, are you finding any companies that are completely asleep at the wheel or is everyone like giving this? Who should I hack today? No, I just want like, I imagine like part of why the business is growing so quickly is this is like a hair on fire problem concern every single day, right? You're seeing these like vulnerabilities or issues pop up. And so I would have to imagine that every software company online where the technical leadership there uses the internet, which I would imagine is all software companies, are like pretty like leaning into this problem. But do you ever, do you ever like, I'm curious even from like an outbound sales marketing standpoint, are there people that are like, ah, this is just like not that important for us right now? Or would like a no be like, we have a solution. It's with someone else.
We, okay, so we've met one company that has literally no software. They produce toilet paper and they didn't have to of business. No. I think, yeah, I think that it's become a universal problem. When we started, it was kind of this like maybe more niche problem where people said, okay, you know, I got a lot of other problems that are more important than this. And we initially had a really good uptake of the product in the cryptocurrency community because they have a lot of software supply chain risk. And when these hacks happen, it's kind of an irreversible thing for them where they like lose all their funds. And then I would say kind of the AI labs were the next to pick it up and sort of, you know, SF tech companies realized that the kind of people that are ahead of the curve. And now it's become, I mean, it's like a top, I'd say top one or top two concern at like nearly every company we talk to. It's a board level concern a lot of the times now. And CISOs and heads of engineering are being asked to figure out like what is our solution to make sure the next time one of these things happens we're protected. And the next time is probably going to be like, you know, tomorrow. Like it's so, it's become so, it's literally like, you know, every day. Yeah, literally our end of quarter was, you know, the end of last month. We, you know, we already were super busy and then there were three major software supply chain attacks that happened that same day. It's wild. It's like more than we've ever seen and it's totally unprecedented.
Was the GitHub issue that I saw a supply chain attack or was that something else? I saw like a couple posts, but I didn't get to dig into it though.
I don't think they've released that information yet. Yeah, there's definitely been some speculation. I mean, the timing right now, every time a company gets popped or their source code leaks, it's the first question that people think of now. And the group I believe that claimed responsibility is this team PCP group that has been responsible for a lot of the attacks. And so, you know, I wouldn't be surprised if that's what we learn, but I don't think, you know, that's something that we know right now and I wouldn't want to speculate.
Yeah. I wonder if there will be movements to find like the compute infrastructure for team PCP at some point or like the hackers because at a certain point you would imagine that they're ready.
Well, they don't, I mean, some of this, a lot of this is still social engineering, right? So it doesn't, they don't necessarily need, you know, a country of geniuses data center, a couple dudes with phones who knows.
Well, congratulations on the progress for coming on the show. Thanks for keeping us all safe and we'll talk to you soon. Cool. Thanks guys. Have a good one. Goodbye. Up next we have Tae Kim from Key Context, the hottest Substack on the internet. We have Tae Kim in the waiting room. Let's bring him in to the TV room. Tae, how you doing? Hey, guys. Here we go. Kicking it off strong. Good. Good fit. You know what we need to do? We need to figure out how to do a like real time, real time recognition of the face, face tracking, and then land the glasses on the correct face. Your next appearance for next video. Yes. And maybe before we get into AI and different companies, I'd love to just know how Key Context is going, how like what it's been like since we last checked in. How is life?
It's much better. It's going great. I think the last time we talked was March 30th and the market was right at the bottom. Yeah, that's right. Weary. Everything was going terrible. Yeah. You never worried though. I did not. I tripled, quadrupled down on here and that's CPU shortage by CPU by memory stocks. Yeah, all those ideas are up 50 to 150%. So I said I think 13 out of 14 ideas are solidly in the green. I joke that like show Ohtani numbers with like three, it's not going to last, but it worked out. I mean, we trusted in the actual show would love memory stocks. Yeah, right. I think he would. I'm sure his financial advisors long HBM stocks. I'm sure. Yeah, he has. Yeah, I read some report that in South Korea there's folks who are liquidating insurance plans to go long SK Hynix. It's a huge, huge moment over there in South Korea and everyone's making fun of these Koreans for being long memory stocks, but I mean, we're talking single-digit P multiples probably going to last two, three, two to three more years, triple-digit growth. I mean, I quoted this Michael Dell thing where he spoke to a Wall Street conference. He says 25, 25, that's the thing you need to know about memory stocks. AI accelerators going to have 25 times more, like AI GPUs from Nvidia two years from now going to have 25 times more memory per GPU and there's going to be a need for 25 times more GPUs. Wow. So he said multiply 25 by 25, you get 625 more revenue, etc. And you layer on to the fact that four years ago all these memory companies saw their revenue, you know, get cut in half. So they didn't expand capacity. It takes three to four years to expand capacity. So we're going to see like mega pricing power. Like we haven't seen anything yet. These stocks are going to keep going higher and their revenue rates are going to be astronomical. So everyone makes fun of these Koreans for piling into single-digit PE stocks growing at triple digits. And this cycle is different because there are only three companies that can make the HBM memory. Yeah. So don't make fun of the Koreans. Don't, don't.
Is the pullback that you pushed to, that you mentioned, like when they were beaten up, was that the post-crypto slump when semiconductor equipment and GPUs were being used to mine crypto and then there was a pullback there? Is that what you're referring to? Like what is the prehistory? Why were these stocks beaten up?
So in 2022, like we had a kind of like a tech recession, like everyone's, all these tech companies were laying off people. Even Intel and Nvidia had these 50% drawdowns. Yeah. The whole, you know, everything was, it's almost like the post-COVID kind of overhang where people bought too many digital computers, electronics and there was too much overcapacity and so that really is different.
What was Nvidia doing during that 50% drawdown? Were they buying the stock back?
Not really. Nvidia did, they bought a little bit of stock back but not really. And back then this is a different Nvidia where like data center is 10% of the revenue and gaming is 80, 90%. Now it's split right where data center is nearly everything. Yeah. What happened with Nvidia back then is everyone's mining their GPUs to mine Ethereum and then Ethereum did the whole proof of stake, proof of work thing. So demand for GPUs collapsed and Nvidia had like a terrible quarter where they missed by like two billion which was a big number back then for that. Now that's a rounding error. Back then it was 20% of their revenue and the stock got hit hard.
Yeah. So what are you looking forward to from Nvidia this quarter, this year? What are you watching?
So we had like two unbelievable quarters the last two quarters and I expect another great amazing quarter. I mean, Jensen is out there saying GPU consumption is through the roof. On Monday he said AI demand is far exceeding supply and capacity. So I think the numbers are going to be great and this is without even China. So I mean, the absolute scale of these numbers are mind-blowing. We're talking like 80% growth on an $80 billion number, 79 or 80 billion. Just think about in three months, right? Yeah. So the absolute scale and the stock is like almost as cheap as it's ever been. Like it's trading at 19 times forward below the S&P 500 which is growing at like 10%. Nvidia is growing at 80%. So this dichotomy where Nvidia is becoming more and more undervalued and we could talk about why.
Yeah, yeah, I'd love to know. Is that just because they're the largest company, 5 trillion, it's hard to wrap your mind around a 10 trillion company.
So every year we go through this cycle. The last three years during this whole upswing, everyone says peak is here, right? Nvidia can't grow anymore and Nvidia keeps growing at some ridiculous growth rate. So the skepticism, the only reason why it's trading at below market multiple is that the AI skeptics believe that the peak year is near, right? There's going to be a no.
Is it AI? Is it AI skeptics or is it TPU and Tranium enjoyers?
That too. They believe competition is coming and they're going to gain market share. I kind of laugh at that. And the numbers that people put out there for the competition, it's like a rounding error. It's like a small single, low single-digit number compared to what Nvidia is going to grow in the next few years. And what people keep forgetting is like if you actually look at the numbers, I mean, Nvidia has a trillion dollars in orders, right? They're the ones that have locked down all the memory. Like I met a guy at GTC. He's the optical startup and Nvidia locked up all the capacity for lasers and optical, right? So Nvidia has the supply components with memory, wafers, optical. So they're pretty much the only game in town if you want to actually buy AI GPU capacity. Like TPU gets headlines, but I mean, $5 billion is nothing when you're comparing it to a trillion dollars, right? So they have the volume, they have the great power per watt numbers and metrics that basically for if you're doing inference, Nvidia GPUs, even though they cost a lot more upfront, when you actually do the inference, performance per watt is excellent when these are four.
Yeah. How did you grapple with the question of whether or not Nvidia is a car? There are competitors. You said low single-digit percentage of the market, but AMD is getting its act together. Intel's maybe back in the game in a few years. You have Tranium and TPU. Competition is rarely a good thing. Is Nvidia a car or not?
It, okay, obviously it's not a car talking about how, you know, I would have loved to, I would have loved to be in the room with you while you saw that segment. You just like taking your computer smashing. I mean, I think I was tweeting about it. I think I wrote a piece on it. I wrote like, you know, four or 500 words about that podcast. I don't know if we want to go into that on, but they, you know, they're going to be 80, 90% of the market. Maybe they'll lose 10% of the market share in two, three years, but 80, 90% of a market that is growing, you know, 50, 70% a year and the hyperscaler capex numbers went up huge. There's like 780 billion going to a trillion dollars next year. Yep. Like the market believes with Nvidia's valuation that Nvidia is going to stop growing within the next year or two. But the overall market I think is going to keep growing at 50%. At least if it grows 50% in the next two or three years and is going to grow at, even if they like maintain market, and I don't believe that's true. I think they're actually going to maintain or even increase market share. But even if market share goes down 10%, I mean, the numbers are just insane. And I don't think Nvidia is going to trade at like 10, 12 times earnings when they're growing at 50%. I mean, that doesn't make sense to me. So I think it's going to re-rate higher. And then the key thing here is the stock buybacks because that's like, if you guys remember during the whole iPhone, like when the iPhone 5 and 6 came out, when they actually had the large screen iPhone, Apple was trading at PE 7, like Xcache, like people thought Android was going to almost like the same thing, like Android's come in and destroy the iPhone market share when iPhone went on this generational run with the large screen iPhones, right? But a big thing that Apple did was they started buying back stock and that's when their multiple started going from single digits to 15 to 20 to 30 times. And I think what, and they already kind of hinted at it, but I think Nvidia is going to start buying stock buyback in size. They said 50% of their free cash flow in the next 12 months. They're a little bit cryptic on if that was after, before prepaid, the prepayments they have for the inventory and suppliers, but I think we're going to get some more clarity tonight. And if they actually put numbers, maybe it's this quarter or next quarter, if they actually put the actual, we're going to buy hundreds of billions of stock, you know, in the next 12 months, I think the PE is going to re-rate much higher. So, you know, as soon as we get visibility that next year is going to keep growing at 40, 50% or higher, I think it's going to be higher. And the year after that, you know, that stock should go on top of the re-rating from the capital returns.
Gavin Baker was on our friend Patrick's show this morning. He is making the point that it's possible that TSMC could prevent a bubble just given that they are not investing in capex as aggressively as let's say Jensen might want them to. He said, you know, basically said what you're saying, which is Nvidia could sell a trillion dollars of GPUs, you know, pretty much immediately if they had the supply. What's your read on that? Do you think that TSMC is...
I think, well, so I think I've been, I mean, I've been following the chip industry for, you know, 30, 40 years. So, you know, I'm very good at reading, I'm very good at reading the tea leaves. And I've noticed that Jensen and what TSMC, what the CEOs have been kind of hinting at, that their tone has changed in the last two, three months. Yeah. So TSMC just reported their Q1 earnings and, you know, if you read the transcript, it's pretty obvious that they're just going to raise capex on a different order of magnitude for the next year and the year after. They said their capex is going to be much higher over the next three years than it was the last three years. And basically, and all the stuff they're saying, you know, they're saying, and Jensen's been going out there during these speeches and Q&As, he's confidently saying the supply issues in two, three years are going to be on a different level. Like even with memory, he's saying all these component issues that we have now in two, three years, I think he even said that on the Dors podcast interview, two, three years, this all will be like less of an issue. So I'm confident that TSMC, memory might be a little problematic, but TSMC wafer-wise, it's going to be dramatically higher over the next two, three years.
But even if they raise capex, it feels like there might be like a, I don't know, a speed bump essentially because if they triple capacity and that takes two years, but everyone's sort of pricing in or expecting 10x a year or half an OOM a year, there's a mismatch there that could be fine.
Right now, right now the Nvidia valuation is trading like it's growing at 10. It's going to grow at 10% the next year or two, right? If they grow at 50% the next two years, I mean, that's definitely, you know, like TSMC is going to raise our capex the next. So it's like the numbers are way off.
Yeah. Yeah. I'm more just thinking from a market perspective.
The 10x is not going to happen. Yeah. Yeah. But, you know, 50 to 75% the next three years, I think that can happen. Yeah. And stocks will go much higher if that happens. Yeah. Yeah. No, that makes sense. 10x is, you know, it's not software. I mean, this, you know, you have to actually move physical and adds.
Yeah, that makes sense. How do you think Nvidia will be positioning China exposure, China opportunity? That was an interesting point of debate on the Doc Cash episode.
Yeah, I mean, I've given up on China. I mean, at GTC, Jensen said, you know, he got approvals on both sides and then weeks later, like that, you know, fell through again. Like, it's forget China. It probably isn't going to happen. There's also another report that gaming GPUs were incrementally banned in China. Like, I just think like you can't rely, temper your expectations. I don't think China's going to happen.
Not relevant to the business case or the valuation, but potentially important to like the geopolitical story.
I think it's just going to be the status quo where Nvidia is not going to be able to sell to China. And anything that happens is incremental upside, but it's not going to be a huge number anymore. Even the stuff that is allowed is a limited number. The way the US has, you know, approved the licenses, it's China is the one that's saying, no, we're not going to let you sell right now. Like, it's been going back and forth between China and the US, you know, who bans what and whatever. Yeah. But even the incremental numbers, even if China starts approving, it's not a huge number the way it was in the past. So I would temper expectations on China. It's just, it's been so, you know, so back and forth and unreliable the last couple years.
Yeah. What was your reaction to Google I/O?
I'm actually pretty negative on Google right now. And they're, you know, they're basically non-existent in the whole coding agent stuff, which is like the one that's going to, that's growing exponentially right now in terms of coding agents. And everyone's paying tens of billions of dollars to Anthropic and now Codex for OpenAI, you know, they're adding a million users every few weeks. Google anti-gravity is nowhere to be found. You know, no one's talking about Twitter. Anyone that talks about it says they don't like it. So the hottest thing, which is coding agents, that's going to, you know, that's going to automate things and do actual work for an enterprise is the biggest market that's taking off like a rocket. Google isn't there. So I mean, Google Cloud is doing great, you know. Yeah. Growth, etc.
Did seeing yesterday, did it start to make more sense to you why Google had been committing so much, so many resources to Anthropic? Like, do you feel like...
Yeah, so I think they at the time, at a lot of, at, you know, at the time, everyone was like, wow, I wonder how, you know, Demis feels about this. Of course, Demis is an angel in Anthropic and obviously, you know, thinks they're a great company and they have a good relationship. But at the time, it was like, wow, what, like they have this new model incoming, they have this massive compute advantage, they have this massive distribution advantage, they have massive data advantage. Like they have every single advantage. They have this huge head start, vend it in everywhere on all these platforms from Google Workspace, which is like a massively, you know, underrated distribution point to the Google, you know, Docs and Sheets and basically everywhere, like every single possible advantage and an incredibly talented team and exponential demand for tokens. And it was still, you know, somewhat surprising to see them, you know, invest that much in a competitive lab. And they're benefiting, you know, with Google Cloud and getting that business. But I think long-term strategically, I think Google has to be very careful. And I think Microsoft experienced this with OpenAI is once the value accrues to Anthropic and OpenAI, they're going to be the power players, you know, four or five years from now and be able to push people around. I mean, we saw that when Yahoo literally used Google as their search engine, right? Google was nobody and Yahoo put it on their homepage and powered their search engine. And then Google became the power player in a few years. Same thing with Netscape and Yahoo. I mean, this is like history repeating itself. Netscape was, everyone used Netscape. I don't know if you guys were around then, but Netscape was the browser, right? And the reason Yahoo became powerful is because you clicked on that little thing on the right on the homepage and it drove people to Yahoo, the portal. Then Google, you know, took off because they started using the Google search engine inside yahoo.com. Yeah. So I think all these companies have to be careful. Like if you help OpenAI, Anthropic become super powerful and they have the best models and, you know, this thing is all a flywheel, right? The more people that use it, more people that use Claude Code, you know, Anthropic is going to get smarter and make the better version of Claude Code. Same thing with chat. And if you let Gemini kind of like lose the market right now. Yeah. Like and Anthropic gets all the business and all the revenue, like Google starts losing power in the whole tech ecosystem. So yeah, it's almost like the whole industry is on a merry-go-round, but there, you know, it's like a Spider-Man meme with like, you know, guns. Everyone's kind of just like going around. Yeah. Very, very.
What are your expectations for the SpaceX IPO? What does that do to the market? They sort of jump straight into the Mag 7 in terms of valuation. Are they going to suck up liquidity? Well,
It's really, it's really tough to tell without the S-1, and we'll see. Versus, you know, all we have is like these spot leaks, and yeah, you know, Starlink, and now with the, you know, selling the AI compute colossus to Anthropic. I mean, the valuation is extremely, extremely high. Even the D1 Capital guy who's up, you know, up. Yeah, he was interviewed on O'Shaughnessy's show, and you could tell if you just read his body language, he's like, 'Wow, this is a really high multiple.' I mean, go back and listen to it. Even he, even though he's going to benefit from it, like he's like, 'Wow, this is really high.' Yeah. So, I mean, a lot of SpaceX investors were happy with, you know, investing at 60 billion, but it was doing 10 billion in revenue and had great margins, and it's a completely different business both on the valuation goes to them. They saw stuff that, no, I mean, like, you know, the launch costs going down, and Starlink coming out of nowhere and adding billions of, you know, unbelievable telecom business, total market, great. But like 1.5, 1.75 trillion in off those Starlink revenue. He's earned the, what was this from Gavin Baker? He was saying like, Elon has just always delivered for shareholders, and that means that he can marshal unlimited capital at all times. And I'm just, you know, I'm just hesitant, like, you know, at those extreme valuations, that's less upside for a retail investor. Oh, sure. Yeah, yeah, yeah. Hard to imagine like the Tesla scenario. Yeah, the passive indexes if they go in at that price, that valuation. Yeah. Like on the next downturn, it's just, I just wish Anthropic went out first. Sure. Right. Because the numbers are dramatically real. I mean, OpenAI probably is accelerating now with Codex 2. Yeah. And, you know, that will help things a lot better than, you know, SpaceX. Oh, sure, sure, sure. Might be an interesting thing right now looking at the valuations of the Mag 7, and then you add SpaceX in there, and then you add, you know, Anthropic and OpenAI, and SpaceX will potentially make Anthropic and OpenAI look cheap on just like a multiple basis, but then it still looks very expensive versus, let's say, like a Meta, right? Um, sitting at one and a half trillion with one of the best businesses of all time. I did want to ask you about Meta, your kind of updated mental model. They obviously had a big, painful round of layoffs today to be able to afford their inference bill. Um, I'm joking there, but it does seem like, you know, they're effectively, whatever they're saving on comp, they're just sort of reinvesting into AI, both capex, but also their inference bills. Um, how are you?
I think you asked me the same question, you know, six, seven weeks ago, and, you know, they reported an unbelievable quarter. I mean, they're growing at 33%. Like crazy amounts of profits. Like, um, there's market skepticism that they could do well in AI models, but, you know, there's also the plan B. Like, you could say the same thing about what happened with Elon and xAI. Like, they were able to, there's so much demand in this mega trend hyperscale cycle that even if they falter a bit, they could sell compute capacity to the highest bidder that's out there. So I wouldn't count Meta out. I actually think, like I said, um, their core advertising business is more durable than the Google search engine right now, which, if you go, it's filled with ads everywhere. It's not a great experience, and a lot of that traffic is going to shift over to AI chatbots. So if you actually look at the numbers, Meta's core business is very durable, has competitive advantage, it's growing, still growing like a weed. Uh, I just wouldn't count them out.
Yeah. Basically, as soon as they either have an AI breakthrough or they pull back on AI, the stock has to re-rate, right?
Yeah. Yeah, I mean, that's the same thing that happened with the metaverse stuff. As they start cutting back spending, you re-rate higher because all those losses don't hit the income statement. They look tiny now.
Yeah. It'd be so wild to think about what Meta's stock would be doing right now if the metaverse debacle had never happened, because the metaverse was like this high-conviction bet on a category that was just way too early. And yet AI doesn't feel, and yet they feel like late on AI. I think Mark felt late, or at least clearly felt behind enough to sort of, you know, 10x the seriousness. So, um, I think, yeah, I think they're a little more skeptical because of what happened with the metaverse, where they spent tens of billions of dollars that has pretty much gone. Yeah. Yeah. Becoming a neocloud is a weird outcome for a hyperscaler that could totally have a cloud platform but has not historically. And, uh, you would imagine that there's a whole bunch of hard-won lessons that Azure, GCP, AWS went through, organizational decisions, structures, just key people in place. It feels hard to spin up immediately. At the same time, there's neoclouds that are scaling revenue extremely quickly every day. So, uh, I agree with you that it is a possibility, but a whole set of new challenges if that's the direction. That's the worst case scenario, right? Like, they want to use the AI compute to make their internal businesses better with all the ad personalization and monetization and create this stuff. And actually, I've been playing around with OpenClaw, you know, for the first time a few weeks ago, and, uh, yeah, I've been using WhatsApp, and that's the future. Everyone's going to have a digital assistant. It's like so amazing. I ask it to give me the top 20 stories. It's powered by the latest version of Codex. Uh, it goes out there, searches every day for me, and emails me the top 20 stories I should read, and it's really smart. Um, you guys should try it. It's actually amazing. And the thing I'm thinking is like, why can't Meta do this? I mean, they have this Meta AI that's not powerful, but, you know, having this little channel for my little T-bot, I call it T-bot. Um, and all I do is WhatsApp it and say, 'Do this. Do this. Find this for me. Find this for me.' Everyone's going to have this. This is a basic thing. And I'm using... I mean, the big headline token number from Google was that they 7x'd token production across all of their services. That actually seemed a little low to me given how broad the surfaces are and how crazy the models have gotten in the last year. Um, but I would expect to start seeing that out of Meta because, uh, now when you go to Instagram and you search, it gives you a little AI overview, and obviously they have the Meta AI app, and I would imagine that there's more AI hydration happening across the ecosystem anyway. And this digital assistant thing is going to be like OpenClaw. Like, I was skeptical until I tried it. I'm like, oh my gosh. So Meta should be all over this because it's on their... and probably with the Ray-Bans they'll do it. Uh, well, thank you so much for coming on the show. Always great hanging out.
Great to see you, Tae. What do you do for Nvidia earnings? Do you just like turn off all the lights in the room and, uh, like just like sit there? Case of beer, nachos, big screen TV? Super Bowl. Is it more like a Super Bowl or like a... No, I wish, like, we should bring back those bar crawls like they had it. They did it once and then it stopped. Like, we should bring that back. Too much of a top signal. It's going to be the number one most valuable company for a long time. Yeah. Like, I'm sorry, Alphabet. It's not going to happen. Yeah. So, we should bring those bar earnings parties back. That sounds great. Well, we will talk to you soon. Have a great day. See you. All right. Thanks, guys. Good to see you again.
Uh, before we bring in our next guest, I want to actually pull up this Wall Street Journal article. See how SpaceX is about to eclipse every other blockbuster IPO. Scroll down on this because they have a beautiful visualization of IPOs over time. You can see the first graph. These are IPOs over the last seven years, ranging from 0 billion raised, probably 10 million, 100 million, all the way up to 2 billion. Continue to scroll down and you'll see what happens as the blockbuster IPO starts stacking up. You get Airbnb, who we're talking to the founder of just in a few minutes, ARM, Cerebras, Uber, and Porsche. And then as you keep going, Saudi Aramco comes in at 25 billion if you keep scrolling down. And then eventually SpaceX comes in to dwarf them all at 80 billion. It makes everything else look like a flat line on the chart. Absolutely blockbuster IPO and should make for a ton of great takes and analysis, and it'll be a fun, fun day on the timeline when SpaceX goes public soon. We're getting the S-1 any day now. Well, our next guest, Immad from Mercury, is here with a big funding announcement. So, we'll bring him in from the waiting room into the TBPN. Immad, how are you doing? There he is.
Yeah, great to be here. Am I not going to get the gong?
Oh, you'll get the gong, but you got to give us the news. Tell us what happened. You got to give us what happened. All right. There we go. How much did you raise?
Yeah, we just raised 200 million. Uh, 550 million. There we go. Sequoia is leading. Uh, so excited about that.
That's fantastic. What was the traction that unlocked this new fundraising round? What was the catalyst?
Uh, you know, we've just been kind of grinding away for the last, uh, what is it, 7 years since we launched. Um, we've seen a lot of growth in Q1 especially, about 2.5x our Q1 last year in terms of applications. Uh, I don't know if you've seen the stats, but there's been a lot of new business formation, a lot of, you know, companies being built using AI. So obviously we hear about like these big startups in AI, but there's also just like normal businesses being built with AI being a big catalyst. And we saw, yeah, big increase, 2.5x over the last year, which, and last year was a big year for us.
How are you reaching those customers? Like, what are the key funnels that draw in these new businesses? They don't exist. It's hard to reach out to them before you have to be very aware.
The other crazy thing for us is most of our growth is organic. Uh, but even the percentage organic has been growing for us. So even though we're growing, we're getting more and more organic, which, you know, is unexpected. Uh, and the other big channel for us is LLM recommendation. So, you know, if you go into your favorite LLM and you say, you know, 'What bank should I use for my startup?' Uh, Mercury is often the recommendation. So that's, you know, we try to deliver the best customer experience possible, and, you know, because people really love Mercury, they tell their friends about it, and that's for us.
What, is like 90% of the reason that you guys are recommended so highly by LLMs just because people recommend Mercury so highly? Like, I can't be, I'm just trying to get at, is it like your team was like, 'Hey, we should really optimize this,' or it was just...
Yeah, I think it's really hard. Like, there's a whole thing called GEO, which is like all about optimizing for LLMs. But at the end of the day, you know, if it's on Reddit, it's on X. Like, if all the content is recommending Mercury, then that feeds into the algorithm.
Yeah, it helps to have been doing press for years. There's a huge advantage to being in business seven years. Uh, not so long that the articles are about the downfall or like how you're a dinosaur. All the coverage is that it's new and interesting and valuable and worth checking out. And so that surfaces. How are you thinking about going deeper in those LLM recommendations and sort of like the agentic commerce of actually opening an account? Is that something that's on your horizon to integrate?
It's kind of interesting thinking about like the LLMs using Mercury. So we actually launched Mercury CLI. We launched that a month or so ago, and we launched an MCP actually in December, and our growth on like those is like through the roof. So tons of people are using Claude Code or Codex or ChatGPT to connect to Mercury. And then, yeah, I think, well, not for everyone, but for a lot of people, they're kind of running their whole business and their life in AI now. And, you know, if you can pull in creating an invoice or like approving a payment, and it's all done within your kind of existing workflows, uh, that's really powerful. And I think if you, there's a flywheel between like, if you build great tools that AI can use, it actually recommends those tools even more. I mean, that's partly why like people like Supabase and like those kind of tools.
Yeah. Then how do you deal with the tension of like, you could build a dashboard visualizing payments that are made into a bank account over a period of time. You could also expose an API that allows me to suck out all the data and vibe code my own dashboard the way I like it. Uh, there's probably demand for both. Do you just have to do both? Do you pick a lane?
I think for us it's like do both. So we're building both kind of this CLI, and we have a whole team that's improving all the features available via the API and those services. We also have another team that's working on, and we just launched Mercury Insights, which, you know, summarizes your transactions, gives you like kind of AI insights on it, and lets you talk to your transactions. And then later this year we're launching Mercury Command, where you can do kind of complex workflows like, 'Hey, I just hired someone, set them up in payroll and create a card,' and, you know, do all of that without me having to kind of click around all of the kind of Mercury surfaces. Uh, and I think number one, I don't know exactly where the world's going to go, right? Like maybe everyone's going to be using AI assistants to do everything, or it's going to be a combination. So we want to service both of it. Uh, and B, you know, if you think about like the gap between us and kind of these incumbent banks, you know, even whether it's like a mobile app or like searching for all your transactions, whatever it is, you know, Mercury's been far ahead. But if AI keeps changing the world and like people's expectation of interfaces and things like that changes, we want to be at the forefront of that, and, you know, that's going to increase the spread between like what people get from Mercury versus like what they get from the incumbents.
Yeah. I was reading an article in The Economist a couple weeks ago about stablecoins going through a boom last year and then going through a bit of a winter, or like maybe just, you know, stable volumes through the earlier part of this year. What are you seeing in terms of demand or tickets or usage of stablecoin-related features in fintech broadly?
Yeah, I think it's a lot more important on like a global basis. So I think if you're in the US and you're sending money to a US business or a US consumer, uh, you know, wires, all of the cards, all of that works pretty smoothly. Uh, if you're paying money internationally or, you know, you're not in the US and you want to have like a stable currency, uh, that's where USDC and these other stablecoins are doing really well. So, uh, you know, we don't currently support it natively, but lots of our customers use, uh, you know, some of these kind of stablecoin services, and we'll probably have, uh, kind of at least payment features that are based on stablecoin, especially for those kind of international payments where it is particularly useful. So I don't necessarily see it as a replacement for the US banking system in the US. I think it's more of like this kind of global connectivity layer.
Oh, sorry, Jordi. Uh, why was this year the right time to get into payroll?
Uh, you know, we've been expanding our services. So obviously we're known best for banking, but we've had a corporate credit card for three years, and that's growing really fast. We have a bill pay invoicing solution, and, you know, over time, I really want Mercury to be the place that you do most of your business finances, right? It's kind of annoying, like if the money's here and your users are here and your finance team is here, you know, you want to have all of that kind of in one place. Um, so yeah, we made an acquisition in payroll. We acquired a company called Central. They're running, uh, you know, we're continuing to run them, and we're going to over time make it more like a fully integrated Mercury payroll solution. Uh, but yeah, payroll is like a big category. Uh, you know, it's kind of annoying right now when you're using a payroll provider. You have to send money, it kind of sits around for four days before it shows up on your employees. Uh, and, you know, Central's done a really good job with kind of AI-first, like they have this whole interface where you can do a ton of things in Slack, and, you know, we're kind of taking some of that learning and putting it in our own kind of command launch.
Yeah. Is the AI boom helping you sort of rethink product roadmap expansion differently than a few years ago? Like when I think about neobanks, fintech, modern technology-enabled banks, I think like you could eventually do auto loans, mortgages, consumer, you could do all sorts of the insurance. There's so many different products in the financial technology world. I know that there's incredible value to focus. At the same time, there's this foundational unlock in being able to move very quickly into new markets. How are you wrestling with that tension?
Um, you know, I think number one, we want to talk to our customers and see what they really want. I don't want to like invent things that they might want kind of thing. Uh, it's definitely accelerating. Uh, you know, I think the two vectors are like, A, it accelerates software development. It doesn't necessarily make like kind of the more financial, like it doesn't make lending that much better faster or whatever. Uh, maybe you could underwrite a little better, but, uh, there's still like kind of limitations. But you can build software, especially new software, much quicker. So definitely thinking about that, and like we're already seeing an acceleration in kind of new software launches we're doing. Uh, I think the other thing that's kind of interesting is, you know, interfaces tend to get cluttered over time, but AI has a decluttering effect. Like you can kind of personalize it to someone's need, and like, you know, I really think the future is people are not trying to do a specific thing, but they kind of just talk about a problem. They're like, 'Hey, you know, I just need to make payroll, like, you know, help me think it through' kind of thing. Uh, so you can go like a level above, like solve the problem rather than like be like a feature, uh, which I think like has a decluttering effect to interfaces.
That's good. Yeah. It's so amazing. I just think back to, you know, 10 years ago at this point, like starting to build companies and just seeing how many things I expected to be intuitive. Like I expected my bank to be able to help me pay my employees because like the function of a bank is to like hold money and let you do things with it. And yet obviously, you know, it took some time for you guys to enter payroll, and then all these other things like just the concept of getting an invoice, wanting to pay it. Like it really should be as simple as like dragging it somewhere and then that's it, right? And so there's all these things that like right in this moment there's going to be an entire generation of entrepreneurs that like only knows this like completely frictionless business finance. And I'm very happy for them. Uh, yeah, and I'm happy for all of us that had to go through and do it the hard way for so long.
Yeah. When I started Mercury, that's actually what got me excited, that like after you build a bank account, there's so much more to do after that. You know, I really didn't want to, this is my fourth company, so I really didn't want to do like this thing where you build the thing and then all you do is like incrementally improve it a little bit over like a decade. Like I really wanted to do something where like the start is just the start and it's kind of a platform to build a lot. So I'm excited that we're like finally like getting to build that kind of multi-product strategy.
Last question. Deployment pace for angel investing. Are you accelerating it? Are you decelerating? What are you doing on the investing side?
Maybe I'm a little old school, but like when I think when the valuations are the highest and the hype is the highest, you should like not accelerate. You should like, you know, we saw 2021, like I'm trying to keep like a continuous pace. Like I think it's total.
Don't you know this time is different?
Yeah, I, you know, I think that there's a reason for the hype and there's really exciting companies being built. Uh, but I think it's going to be even more fun when the valuations come down after the hype dies down and you can like then deploy. You know, it was crazy, no one was deploying in 2022 and 2023 when like it's the same companies at like a massive discount.
Yeah. Yeah. No, I mean the smart funds were, and, uh, a lot of great companies made it through, grew even bigger, and are bigger than they've ever been, as you are. So congratulations on all the progress. Thank you for coming on the show. We will talk to you soon. Have a good rest of your day. Goodbye.
Uh, there have been a number of AI images hitting the timeline. One was of Peter Thiel eating ice cream in Buenos Aires. He has been spotted at a chess match. I think that image was not AI. This one is AI-generated by Cat McGee. Uh, but it still sparked a bunch of funny reactions. Here's one of them. The question you need to be asking is, 'What delicious ice cream is almost nobody eating?' The question is, uh, is mint chip properly rated or underrated? Uh, mint chip might be underrated. I think it gets thrown in with the chocolate category. It's sort of a second fiddle to the vanilla, strawberry, chocolate. Always a mint chip kid. I like mint chip. I was always strawberry. I like strawberry ice cream, but chip's good. But I think as you zoom in on this AI image, you can tell that it's AI. It's just too crystal clear, too well lit for, you know, someone to snap a picture unnoticed in Buenos Aires at an ice cream parlor. Uh, this would require professional lighting to get something like this out of a camera on the fly if you saw someone. Uh, zero to yum is what they say. Anyway, um, uh, question, John, that I was asking you this morning. Yeah. Uh, do you think it's better to rest on your laurels or sleep at the wheel? I think I have my answer, but I want to hear yours.
Uh, I think in the modern era, sleeping at the wheel is increasingly safe because of self-driving cars. You can be sleeping in a Waymo. You could be sitting in the front seat. I don't know. Do they let you sit in the front seat of a Waymo? I don't think so. Maybe if there's a fourth.
Yeah. Can you not... Not behind the wheel.
Not behind the wheel. But what if I'm in the back and I wiggle my body through the front two seats? Sit in the front. It'll stop the car.
They're going to come on the phone. They will stop. Okay. Well, uh, then it seems like it's increasingly difficult to sleep at the wheel. And so perhaps resting on your laurels would be the better option.
I think so because it allows you to, one, it's obviously comfortable, quite relaxing.
Yes. Uh, but just because you're resting doesn't mean you can't get back in the game.
Yeah. Also resting on your laurels implies that you have laurels to begin with. Whereas sleeping at the wheel, you just have a car. Yeah. It could be... It might not even be your car. It's also very Lindy, right? People have been doing this for thousands of years. You can imagine a Roman emperor resting on their laurels. But they weren't sleeping at the wheel. They hadn't invented the wheel. They barely invented the wheel.
I think they had the wheel.
Well, they would ride on top of wheels. They had the wheel. They were planning to reinvent the wheel. But no one had a wagon that had a wheel that would steer the wagon. You had reins. They would take the horse by the reins. Isn't that the phrase? Take the bull by the horns. What's the reins idiom? There's some sort of reins idiom anyway. Um, uh, yeah. No, I think personally, resting on your laurels implies that you have laurels to rest on. Sleeping at the wheel means that you just got behind the wheel. You could be in a sudden death situation.
