In short
TBPN Podcast Episode Summary
Podcast Details
- Title: TBPN
- Description: Technology's daily show (formerly the Technology Brothers Podcast). Streaming live on X and YouTube from 11 AM to 2 PM PST Monday - Friday. Available on X, Apple, Spotify, and YouTube.
Episode Information
- Episode Title: Studio Ghibli, Andrew Reed, Rob Mohr, Tanay Jaipuria, Immad Akhund, Steven Schwartz, Jeremy Levine, Anthropic Copyright Dispute
- Date: March 26, 2025
Notable Segments
- Studio Ghibli Discussion (00:13)
- Exploration of the Studio Ghibli style and its impact on meme culture.
- Discussion on the latest image generation capabilities in AI tools like ChatGPT.
- Contemplation on the implications of AI-generated images and art.
- Anthropic Copyright Dispute (24:50)
- Overview of the legal challenges faced by AI companies regarding the use of copyrighted materials for training models.
- Insights into recent court rulings affecting AI outputs and copyright claims, particularly in relation to the music industry.
- Guest Interviews
- Immad Akhund (01:00:18)
- Discussion on his recent ventures and observations in the tech space.
- Steven Schwartz (01:30:41)
- Insights into the growth and evolution of the Huberman Lab podcast and its impact.
- Jeremy Levine (01:46:31)
- Perspectives on venture capital and the landscape of tech investing.
- Rob Mohr (02:00:55)
- Conversations about the importance of community in building successful tech platforms.
- Andrew Reed (02:19:02)
- Focused on venture capital strategies and lessons learned from major investments.
- Tanay Jaipuria (02:48:05)
- Exploration of S1 filings and the impact of AI on startup growth.
Key Themes and Takeaways
- Generative AI in Art and Business:
- Discussion around the implications of generative AI on creative industries and the emergence of AI-generated art as a cultural phenomenon.
- The potential for AI to change creative processes, with examples of companies embracing these technologies.
- Navigating Legal Challenges:
- Highlighting the evolving legal landscape for tech startups, particularly regarding copyright issues in the AI space.
- Importance of clear communication and legal strategies in navigating these waters.
- Venture Capital and Startups:
- Insights into the journey of venture capitalists and how they evaluate companies, emphasizing the importance of market understanding and product viability.
- Recurring theme of community engagement as a driver for startup success.
- Product Focus:
- Discussion on the significance of product development and maintaining a clear vision amidst the chaos of market pressures and competition.
- The importance of continuous iteration and responsiveness to consumer needs.
Concluding Thoughts The episode encapsulates a range of thought-provoking discussions about the intersection of technology, creativity, and business strategy, providing insights from various leaders in the tech space. The focus on generative AI, legal implications, and the venture capital landscape reflects the current zeitgeist in technology and innovation.
Listening Options
- Where to Listen:
- X
- Apple Podcasts
- Spotify
- YouTube
Episode Highlights
- Studio Ghibli memes as a cultural phenomenon
- Anthropic copyright dispute's impact on AI
- Perspectives from influential guests on venture capital and tech sectors
---
Feel free to add or modify any sections as needed to reflect further insights from the podcast episode!
Written by AI. May contain mistakes. Listen to the episode to check what was said.
Transcript
Automatic transcript. May contain errors.0:00You're watching TBPN. It's Wednesday, March 26, 2025. We are live from the Temple of Technology. the fortress of finance the capital of capital we got a great show for you today i'm sure you've seen all over the timeline it's studio ghibli too cheap to meter everything too cheap has been giblified uh there's been a movement by geordie to try and move us past the studio ghibli era era era i advocated for a simpsons era didn't really get much traction on that but there's people on the timeline i got uh jared callen replying to me right now he says that dr seuss is out and actually where the wild things are oh wild things are we're entering a new era but people have a lot of fun people said we're entering this era of abundance and clearly we're we're here yeah we're here this is post scarcity yeah post scarcity post scarcity yeah food eggs you know are still uh you know ten dollars an egg i hear but lutnik was saying he brought down the price of eggs or something i haven't i haven't checked the facts on that but that's what i heard in the last well we were here to spread misinformation yeah and not do the research but we did the research on open ai's new model if you haven't seen they just updated it's available in chat gpt right now you can get the app check it out it's in gpt 4.0 i think oddly it doesn't really show up if you're using the the 03 high reasoning models i would get kind of confused you would almost think that they are intentionally making it difficult to find yep because it works great in 4.0 really the enthusiasts are the only ones using it you can take a picture anything from your camera roll you can even just snap a photo right within the app and then just say studio ghibli style and boom you have a meme that you can share yeah i think the wild thing about this iteration is that there's zero prompt engineering required to deliver an exceptional yeah i mean with mid-journey you have the s refs and and all these different prompts and negative prompts remember uh no six fingers you just have to say that and tell it remind it and now you can just go in there with two words three words and you get exactly what everyone else wants and it's fascinating because i wonder how i mean we there's so much to talk about here but um i wonder how enduring this will be like you can imagine now that the like the jd vance memes where they made him look like a child and then like a very furry person and like a giga chad and all these different iterations um i'm wondering if like the studio gibbification of iconic images will happen like the same day um in well and and it seems like we could be 24 or 48 hours away from somebody figuring out how to make, you know, the matrix as Studio Ghibli, right?
2:51Like if you - Like the whole movie you think? Or like a trailer or something? It would take quite a lot to figure out how to do that today, but we're maybe six months out. Well, we have a video that actually someone we interviewed at YC Demo Day made. I don't know if we can pull it up then, but someone made with this new image prompt, they made a whole bunch of Studio Ghibli style images about San Francisco and YC and their life in Y Combinator. And then they animated it probably with runway or one of these video image to video generators. So it just adds a little bit of motion, a little bit of character to the visuals.
3:28And hopefully, I think we're going to have the founder of runway on soon, Cristobal Valenzuela, who I'm a big fan of. But anyway, let's go through Ben Thompson's analysis because he broke it down really well. The Verge also covered this. And The Verge really sets the table for exactly what's going on here. So OpenAI is integrating new image generation capabilities directly into ChatGPT starting today. I think it's interesting because with the video model Sora, Sora is now a separate website that you have to go to. And there's a link to it from ChatGPT to take you over to Sora. And Sora is more of a, it's almost like a nonlinear editor, like a Premiere Pro.
4:10Like you start with a prompt, it generates a few videos, and then it can kind of cut them all together and you can make edits. and it's a little bit more of a UI-based paradigm. This is different because, and if you remember when Dali 2 launched, it was its own website separate from ChatGPT, separate from GPT-3 Playground or any of those. You'd go over to Dali and then it would have a whole bunch of UI functionality to do in-painting and touch-ups and expansion. You could do really cool things, but it was a separate thing. This is clearly that Google mindset of like there's the search bar. Go to chat.com, chatgpt.com, openai.com.
4:49You're just going to get, what do you want to do? Oh, you want to do an image? We'll do it right now. Ben, can you play that video? Yeah, one second. Okay. So I'll keep reading from The Verge. This feature is dubbed images in chat GPT. So they're not giving a crazy new name to this structure or product. Users can now use GPT-40 to generate images within chat GPT itself. This initial release focuses solely on image creation will be available across GPT +, Pro, Team, and Free subscription tiers. I'm sure the servers are absolutely melting today because everyone has been dumping prompts in there.
5:25And you can just tell from how long they take to render that there's something going on. Yeah, and it's interesting to think of this as effectively a marketing expense for OpenAI, right? Yeah, 100%. Celsius, I love it. Cheers. Crazy. And we are in this, I believe. so I think that's Brian Chesky speaking at YC and then the Golden Gate Bridge uh and then just a little bit of little bit of motion nothing crazy it's not a full movie it's more like moving paintings but it's very cool and hey that's us there we are at YC Demo Day I love it uh thank you for including us it was fun interviewing you and uh yeah this is just uh I don't know this is this is like this is lovely like it it's it is sloppy but it's cool and I don't know it's very clearly less sloppy yeah like that like the the sort of yes slop is flooding the timeline but at the same time it's better than ever they're getting the text right the fingers are right yeah i'm one-shotting every single generation that i'm doing uh while we after we had started the show look at this uh i'll post it in a second that's great yeah post that's great but uh but yeah just one-shotting prompts left and right including text which has been really hard for a long time the text is phenomenal of course if you're doing really detailed sub layer text like at the it's just a few pixels high like you know font size 10 it's going to get a little bit confused uh some people took our banner and overlay and put that through the studio ghibli filter and some of the smaller logos from brands that we know and love uh were kind of mutated a little bit but it got pretty close honestly and it did really well with the ramp logo it did really well with a lot of other icons.
7:05Oh, yeah. Oh, we have a soundboard now, by the way. We have a soundboard now. When certain things are mentioned on the show, I might hit the soundboard. Yeah, we're upgrading the show every day and we hope you enjoy it. So the free tiers usage limit is the same as Dali. The spokesperson told The Verge, but added they don't have a specific number to share. These may change over time based on demand. Per the ChatGPT FAQ, free users were previously able to to generate three images per day with Dali 3. That'd be terrible on a day like this. You don't want to be on the free tier now. As for the fate of Dali, Christensen said, fans will still have access via a custom GPT, but they're clearly moving towards aggregation around this one box, one text box.
7:49The model is a step changer. I mean, it's so crazy to think about as a dad, I know you've done this kind of thing in the past, but you could go through, take pictures of every picture of your kid's favorite book, swap your own faces onto all the characters. And you could do that in like maybe 10 minutes right now. And somebody could build software that does it all automatically. I mean, and that's like probably the biggest opportunity today. If you just wanted to figure out how to make a hundred K is like printing kids books by just doing the sort of like face swapping thing, you know, remaking these things.
8:22Yeah. I think it's, it's probably a business that you don't raise a ton of money for, but, but, but you start today, you start today, you figure out the exact cost of what it'll cost to generate all those images. What are your printing costs? And then you go on Facebook and you market to people over 65 years old that have grandkids. And you say, it's a hundred dollars, but this Christmas make your grandson elated with an amazing book. And I think you have a good business on your hands if you can figure out the acquisition funnel. So let's go to Ben Thompson. Creativity is too cheap to meet her.
8:55It really is so cool. I mean, we were debating about this earlier. Like, does this put Studio Ghibli out of business? Or will Studio Ghibli just have tools to make more content faster? Yeah, we were debating whether or not OpenAI, it's possible. We have no insight into this, but it's possible that they actually did a deal and said, hey, we're launching this new image generation tool. We already know. Yep. Let us basically like actually do this. Because I was messing around with it last night and I was running into a bunch of copywriting issues that then seemingly went away yeah it wouldn't let me make a studio ghibli image of winnie the pooh but it would let me make a studio ghibli image of donald trump or anyone or us because um i guess like the likeness of public figures is a little bit looser and the style is hard to copyright but you can copyright specific the idea of right now it's inconsistent and I'll upload a picture on my phone and do a prompt and it'll say, can't do that.
9:55And then I'll go on my computer and do the same thing and it'll work. Yeah, it's definitely, people are dancing around it trying to figure out what it can do. But it's fun and clearly the technology is advancing very, very quickly. And this functionality will probably be all over the place. And you know, Grok is gonna be getting wild with it and letting you do way more stuff. You know, the meme of the soldier sort of like tossing a grenade over, couldn't do that, frustrating. Oh, interesting. So no violins and stuff. Yeah, I mean, I'm sure there's a bunch of filters around adult material. I am very interested to hear, is this a change in scale or is this a change in algorithm?
10:30And that will probably be open sourced at some point or discussed on Dorkesh's podcast, probably. Someone will explain what's going on here. I don't understand it at this point, but it's really cool. Let's go to Ben Thompson. He says, the big change here is that Dali and most other image generators use diffusion the model would start with an image of nothing but noise and iteratively denoise it to create an image. The unit being operated on was in fact the image itself. Changes in ChatGPT on the other hand is auto-regressive, images in ChatGPT, the new product, on the other hand is auto-regressive, which is to say it works the same way ChatGPT works.
11:11First one token is predicted, then that token is fed back into the context, and then the next token is predicted, etc. In this case, however, the next token isn't text, but rather pixels. This pixel by pixel generation makes images in ChatGPT particularly good at iterating on images or stylizing them, for example. Here's a picture of his cat next to a picture of his new water fountain. I love that he's sharing pictures of his cat. And then, boom, Studio Ghibli style. The Studio Ghibli stylizing was inspired by this viral tweet, which almost certainly horrifies Hayao Miyazaki, the creator of Spirited Away.
11:48and this one went super viral 29 ,000 likes when Ben Thompson screenshotted it I'm sure it has more now Grant Slatton said tremendous alpha right now in sending your wife photos of y 'all converted to Studio Ghibli anime I did that last night before I went to bed with my kids I took my son out to ice cream and I Studio Ghibli The kid thing is an issue I tried to put my I tried to make my I didn't have a problem with it kid and they said we can't generate images of kids. Just stick to the boys. Stick to Jeremy Giffon. Stick to Will Manitis and David Senra. Yeah, your Jeremy generations were fantastic.
12:26Yeah. This is a cute one. I'll show you this. Wow. Yeah, it's funny. Anyway, so it is particularly good at handling text now. You can also iterate on image with high fidelity, and he gives some examples of editing a minimalist the sticker of a raccoon. Let's have the raccoon eating a strawberry. Let's give it a thick white border. Now let's make the raccoon gray. Can you make the raccoon actually take a bite out of the strawberry? It does that. It's all very progressive. And it's a very different style of prompt engineering because you're not starting from scratch because the image is being fed in every time you prompt.
13:09And so that leads to character consistency and a lot stronger. and it's just a more natural feedback loop. Like a lot of people would get frustrated when they were prompting previous image models because they would say, oh, okay, it's really, really close. Like this is amazing. It looks great. But I just need that tree. The small iteration. The tree on the left mountain needs to be on the right mountain. Yeah. And it'd be like, well, here's a whole new image. And you're like, no, no, no. You really were close. We just wanted to move this over here. And it feels like working with, an actual creative partner where you can say, okay, this is great, but you need to change this small part or change the color of this.
13:52Yeah. And so he gives some, he gives another example here. You can, as noted above, use an existing image as inspiration or draw on the model's world knowledge. And so draw a design for a vehicle with triangular wheels, use these images as reference, and then now put this, now make this a photo that's taken in New York city and it just makes photo reel, which is very cool. So anything that you have, and I saw a bunch of, I think the underrated discussion right now, obviously the Ghibli memes are really fun, but a lot of people were showing how they could use this for e-commerce advertising. Take a picture of your product.
14:25Okay, put it in a hand, put it on a table, put it in an everyday carry, get my Ridge wallet, take a photo of that. I actually saw some people generate Ridge wallet ads that looked extremely convincing and were well past the uncanny valley at this point. What is so striking about these examples? is the kind of thing I imagine this massively improves icons product. Totally. Totally. Yeah. What is so striking about these examples is the clear day-to-day utility. Diffusion is probably still better for original image generation given its global nature. Autoregression can get stuck in local maximas.
14:57The clear barrier to actually using diffusion for day-to-day image workflows, however, was the relative lack of control that images and chat GPT excels at. And he goes on to to talk about graphic design and Google. I wrote earlier this year, Ben Thompson says, in the introduction to deep research and knowledge value, when did you feel the AGI? This is a question that's been floating around AI circles for a while, and it's a hard one to answer for two reasons. First, what is AGI? And second, feel is a bit like obscenity. As Supreme Court Justice Potter Stewart famously said, I'll know it when I see it.
15:31And Ben Thompson has given his definition of AGI. what O3 and inference time scaling point to is something different. AIs that can actually be given tasks and trusted to complete them. I'm certainly feeling this when I use deep research or fire off some huge query. It feels like it's going to take a while, but it's going to get it right. And I don't do nearly as much fact checking after as I did with the original ChatGPT product, which was built on 3.5, which had a fair amount of hallucination. And so he goes on to write, this by extension looks a lot more like an independent worker than an assistant.
16:08Ammunition rather than a rifle sight. That may seem like an odd analogy, but it comes from a Keith Raboy talk that he gave at Stanford. My definition of AGI is that it can be ammunition. It can be given a task and trusted to complete it at a good enough rate. My definition of artificial super intelligence is the ability to come up with tasks in the first place. The feel part of that question is more recent discovery. deep research from open AI feels like AGI. I just got a new employee for the shockingly low price of$200 a month. Pretty soon we're going to get ASI and ramp it up to artificial giga intelligence.
16:42And that will just be the new sort of what we're all sort of racing towards, right? Yes. Yes. Talking about definitions of giga intelligence. Yeah. We're going to need a new term pretty soon. for sure. Well, good news for me. I just got another new employee. I already gave it a job for the section below. This employee was like deep research, one that I wouldn't have hired otherwise. Making a cartoon helped make my point, but it was hardly necessary. At the same time, however, you can certainly see the level of control afforded by images in ChatGPT, making real a lot of the fears that accompanied the first wave of image generators.
17:20What's also notable about images in ChatGPT is that it wasn't the first autoregressive image model. Google released Gemini 2.0 Flash Native Image Generation two weeks ago. And I remember seeing a lot of screenshots on X about this Gemini 2.0 Flash Native Image Generation. Another great name in AI, of course. And people were talking about how good this product from Google was for specifically editing images. and saying, okay, I have a Celsius here. Let's remove that from the image. He was really, really good at that. But Ben Thompson's pointing to two limitations compared to what OpenAI just launched.
18:02So Gemini 2.0 Flash Image Generation Experimental, that's the model name, is only available in the Gemini API and Google AI Studio. And so there's memes about the Virgin Gemini that you get from Google and then the Chad Google AI Studio because AI Studio was vending in the latest and greatest model. So when you saw those AI influencers on X posting, they were using the best model from Google AI Studio or the API. But Gemini 2.0 Flash Image Generation Experimental doesn't support converting images to new styles. It's designed to start with images it generates from scratch, which then you can iterate on.
18:41And so when he went to Google's product, Gemini 2.0 flash image generation experimental. What a mouthful. G2 FIGE. I don't know. It says make an exact version of this with Studio Ghibli style. It creates an entirely new image from scratch. It's not doing the proper style transfer. And style transfer has existed for a long time, even before Dali. And it's been really cool, but it's always been just not quite right. There were all these examples of like make this like a Van Gogh painting. I don't know if you've ever seen those But they were always like close but not super reliable so but now you can literally go and just you know take your kid to the park Document it with a couple photos and then say make this a comic book and boom It will look exactly like you expected it to all the framing will be the same all the characters will be the same It'll capture the essence it really captures the soul of the person that you're taking a picture of it's kind of crazy Yeah, and you were saying every time somebody makes a Ghibli version of you, they take a piece of your soul.
19:48I think the Native American... They were onto something. I don't know if it's even a true story, but we've certainly been taught that in grade school. That idea of the image of the camera capturing a part of your soul, it feels more real than ever today. In short, Google released some very cool technology, but they didn't release a very cool product, something we've talked about before. OpenAI did, and that's why everyone's talking about it. And oh, by the way, Google also just released Gemini 2.5. Ben has this uncanny ability to say something very plainly that just is... Devastating. Devastating.
20:25Yeah. Cool tech. Now build a product. Now build a product. Ouch. Well, yeah. I mean, it's a good point. More on security. One day I will learn not to wade in with orthogonal takes on current political topics, but apparently not this week. This is OpenAI's depiction of me reading all of your mostly thoughtful emails to yesterday's update. What about, what about, what about, and Ben's sitting there thinking, I've already written about all of that. Specifically, I received a lot of emails noting that secure communications, and he goes on to talk about Signal. I don't know if we want to get into this.
20:58I think we should stay on Ghibli. Very fun. But you can go read the full update and of course go subscribe to Sir Techery uh one of the greatest writers to ever do it um but let's move on to the next slide because uh lots of lots of people are having fun with Ghibli on the timeline and I wanted to get into it because none other than uh show sponsor Ramp switch your business to ramp.com put up a beautiful studio Ghibli image of Saquon Barkley there it is Super Bowl ad there it is and I there are a couple cool things about this image so So, I mean, you just see like the level of detail, like the hands are perfect.
21:36It's really like past the uncanny valley, but it does a great job of, in the actual Super Bowl video, they're using a camera that has shallow depth of the field. So you can't really see what's behind. And you can imagine it just maintains that depth of field. It's just like, oh, blurry stuff in the background. Let's just make blurry comic book in the background. But that's not how the Studio Ghibli style is. is like like these cartoon anime studio ghibli style it doesn't have depth of field and so it has to figure out what's back there and then render it properly and so you can see in the background of this image there's like very sharp lines on the paintings on the wall and there's things that actually make sense like there's books stacked on top of boxes stacked on top of other boxes and all of it kind of follows like a very clear logic you don't have this situation where There's just something floating or something that doesn't really feel right.
22:29It feels great. The other thing that's interesting is that in the Super Bowl ad, I didn't realize that the snow globe that's sitting on the table says world's best boss. But in the Studio Ghibli version, it read that, understood that, clarified that, and created it even clearer to read text. Did you notice the world's best boss? I never saw it. I never saw it. But it was actually there because I saw this and I was like, did it just think to put that in there and i looked back at the original image and it did say world's best boss it was just very subtle in the actual super bowl app so i thought this was very cool it even it the logo is not a hundred percent right but it's pretty close uh they certainly got the font right on the ramp it's just wild this kind of image was possible to generate through extremely intentional prompt iteration and massive iteration totally but we were going back and forth with the ramp team this morning about a bunch of stuff.
23:27And I know that they generated this like basically in a one shot. And that's, and that's just what's so powerful about this, you know, most recent iteration. Yeah. And it's, it's interesting because it, it, it, it clearly cuts corners and takes like creative liberties, but it doesn't take creative liberties that wind, that bring it to a negative place or a place that degrades the image. And so like even the expression on Saquon's face, it's not exactly what was happening in the Super Bowl ad, but it tells this emotional story that makes it just feel like, oh, okay, this is actually thought out.
24:05It's not just a screen cap that's been colored over line for line, which I thought was really cool. And so Ramp on X says, we're assembling a team, and they put both of their star athletes up in Studio Ghibli form. And this is a great time to tell you. We are talking about... Using RAMP. RAMP. RAMP. RAMP. RAMP. RAMP. RAMP. Yes, if you haven't seen it, we have talked about RAMP so much that we made a whole song about it and our editors edited together every time we've said RAMP and made a little jingle. Not every time. I think it was like 1 % of the time. Oh yeah, 1 % of the time. Anyway, time is money.
24:39Save both. Go to RAMP.com. Easy to use. Corporate cards, bill payment. Thank you to RAMP for not just sponsoring the show but presenting the show. Presenting the show. We love you. Presenting sponsor. Anthropic. Should we go to Anthropic? This is kind of interesting because it relates. This happened, like the timing is unrelated, but it's an interesting story because it obviously relates to what's going on with image generation stuff. So, this is from the Wall Street Journal. Anthropic scores a win in AI copyright dispute with record labels. Music companies claimed the company was harming them by using copyrighted material to train its AI chatbot Claude.
25:17And so I'm sure with all these Studio Ghibli images, there will be a big question. Where'd you get the data? And was that, did you have the right to train on that data? Where'd you get the data, buddy? Where'd you get the data? That's always the question with these AI training runs. But I've always said that we have a process for sorting this out. It's done in the courts and we have fair use. Is this transformative? is drawing Saquon Barkley like an anime. Is that fair use of that anime style or do you have to pay some sort of royalty? And that economic equation can be solved and it will be solved by the courts.
26:01And the mere fact that there are lawsuits is not indicative of true wrongdoing. It's more just like, hey, there's a new thing going on and we need to decide where the dollars flow. And this was true when music streaming started and whatnot and eventually got to a place where I think everyone feels like they're getting their fair share. So Anthropic scored a win this week after a US court denied an injunction that Universal Music Group and other record labels had sought to prevent the artificial intelligence company from using copyrighted lyrics to train future AI models. So of course, if you hold back all of the lyrics that are associated with universal music, AI is going to be a lot worse at writing lyrics.
26:46And also just less aware of the world. You ask it, you know, a question about Taylor Swift and it just is blind to everything that Taylor Swift has ever sung about. Like that could be a pretty big hole in world knowledge, right? Even if you're just asking for, hey, write me a biography of her or write me, or just answer a question about what she's up to today. Like it's important to have that extra context. but at the same time, music is copyrighted. So how does this sort out? So Concord, ABKCO Music and Records, Universal Music, and several subsidiaries sued Anthropic in October of 2023. This has been going on for a year and a half, saying the company was harming them by using copyrighted material to train its AI chatbot, Claude.
27:30The record labels alleged that Claude's responses to user queries contained verbatim or near verbatim copies of the work, saying their reproduction violated copyright. The music companies, which represent a large cohort of artists ranging from Taylor Swift to an Ariana Grande to the Rolling Stones, said Anthropic infringed copyright in lyrics from at least 500 songs and sought a preliminary injunction that would prohibit the company from using the works to train future models. However, a judge in California on Tuesday denied a motion for that injunction saying the record labels hadn't demonstrated how using the works to train Claude caused reputational or market-related harm.
28:09So this is a good, this is, I always go back to just like the most normie analysis of fair use is just like, could you imagine a Taylor Swift fan saying, nah, I'm not gonna listen to her new album. I'm just going to go to Claude and ask Claude to reproduce the lyrics and then I'll read the lyrics and that will be satisfactory, right? No, no one's doing that. It just doesn't make sense. but there are plenty of scenarios where fair use gets a little wonky because someone is like, there are examples of like streamers who are streaming like the, like, yeah, I'm streaming the whole boxing match, the whole UFC match.
28:51And I'm just kind of being like, cool. Yeah, that was a good one. And they're not really adding anything. They're not breaking it down. They're not really reacting to it. The bigger streamers from what I'm aware of are not actually showing any of the imagery or the sound. They're just providing this sort of additive commentary that you can position as a watch along or watch companion. So you would still, to get the full experience, you still need to buy the pay-per-view or whatever. And it depends a lot on who owns the IP. Like some video game companies love, they actually pay streamers. I think Call of Duty right now is promoting their new product, Black Ops 6, and they're paying streamers to stream it.
29:29and effectively like share that intellectual property, which Nintendo for a minute, I don't know if it held up, was saying, hey, maybe we don't want streamers streaming Nintendo products because it could be seen as a substitute. It's kind of silly to think about, but because playing a video game and watching a video game is a very different experience. But for some people, they might say, hey, yeah, this particular game is very story driven. I'm just gonna watch some pro play it and they'll be better at it than me. and I'm too lazy to press buttons, which is kind of sad, but I understand people are busy and maybe they want to fold laundry instead of actually hold the controller.
Read the full transcript
30:05Yeah, I remember being probably 10 or 11 and I would watch Call of Duty sometimes. I talked to this guy yesterday, Sam. Vitesh? Yeah. Yeah, yeah, yeah. He was X-Jaws. X-Jaws. Yeah, he was like, oh yeah, Jordan, you're a big fan. Yeah, when I moved to, I was a fan when I was probably 10 or 11. Sam was only a few years older than me, but was one of the biggest gaming youtubers yeah i think he hit a million subscribers on youtube for his call of duty content so he was massive and i would watch it because my parents like were kind of against video games and let me get a a console but it only could be in the living room we had a small house and like it just i didn't want to play video games like with my family as a as a you know preteen or whatever so i just be like cool i'm gonna watch it on youtube i'm gonna watch x jaws or whatever um so shout out shout out to sam um it's extremely zoomer coded yeah exactly in selling yourself watching video games yeah but that's where back in my day if you wanted to watch a video game you had to sit on dial-up internet for 12 hours downloading a two-minute highlight reel but in many ways 360 no scopes are a form of art it is it's performance art yes right it's high stakes uh well uh let's move on to more of this wall street journal article despite the court court's narrow ruling we remain very confident in our case against anthropic more broadly a representative for the music company said this case remains vital to protecting creators from the wholesale theft of their copyrighted works by anthropic and other ai companies we expect that as we proceed with this case and fully develop and develop a full discovery record our claims will be validated there's going to be uh you know emails saying yeah we got to get taylor swift lyrics in here the ai just won't work without them it's impossible we'll never we'll never it is it is uh interesting because many people are paying open ai as like a writing a writer as a service right it's like writing as a service or image creation as a service and what do writers do when they are being paid to write they sort of look at various sources it informs their um you know approach and it informs the content even though it's new content and you don't see record labels you know suing uh some some author and saying like we know you read our stuff uh you know yeah stop yeah right it just doesn't there's no case there it's It's not causing consumer harm.
32:43Yeah, because it is transformative. It's not threatening. This, and from the record company's point of views, they're fine if they just sue the model companies until they get a big settlement, and then it's like, cool. We're still in the music business. People still want to buy music. I mean, a lot of this is just figuring out where the value accrues and what's fair. The bigger threat is that you get this AI-generated music that becomes so good that people stop streaming Gunna on Spotify because they can just generate unlimited Gunna songs, which I've done. At one point, I included a Gunna song in a video.
33:24It was fully AI. It sounded exactly like Gunna, and it was a cool vibe and whatever. So that will be more contentious, and I think they'll have probably more of a case potentially because, hey, you're using our content to generate a substitute for that content that is harming our business. Yep, yep. So the judge noted that the company doesn't meaningfully dispute that Claude's training data included copyrighted lyrics. Anthropix said Claude's intended purpose isn't to reproduce existing works in our user queries, but to generate original output. So they said, yep, it's in the training data, but we think that's fine.
34:08It's fair use because we are transforming it. Earlier this year, Anthropic struck an agreement with the record labels to apply guardrails that would prevent current and future AI models from generating responses that might infringe copyright. The judge said the deal effectively resolved one aspect of the case, but didn't prevent Anthropic from using songs to train the models. What's interesting is I feel like the real company that's getting maybe wrecked more than Universal Music Group would be like genius.com, like Rap Genius or like a lyrics website. Because a lyrics website, if you're just looking like, oh, what's that lyric in that Rolling Stone song?
34:47I don't really know the name, but it goes like something like this. And that's why the LMs as answer engines. And then it just gives you the answer. Yeah, answer. Typically you end up on the lyric website because you're like such and such song that by hip hop artists that says these three words and then it'll pull it up. But now you can just go into the, directly to the LLMs for that kind of thing. Yeah. The case is the latest in a string of disputes between AI companies and publishers on whether and how easily accessible content like music or news can be used to train AI models. Publishers are moving to shield themselves from what they see as violations of their work from AI startups.
35:27Last year, the Wall Street Journal parent Dow Jones and the New York Post sued generative AI search engine perplexity for alleged copyright infringement, saying the company used copyrighted news to generate responses to user queries, siphoning away traffic that would otherwise go to publishers' websites. Meanwhile, the New York Times is suing Microsoft and chat GPT maker OpenAI for alleged copyright infringement, kind of the same idea. Very rough. Well, I mean, it's good that this is working its way through the courts because I think it is, I mean, we have a lot of free speech and fair use doctrines to build off of, but there clearly needs to be foundational settlements and new case law created, essentially.
36:09Anyway, should we go to something? And we're always talking about supporting private equity, making sure private equity sponsors, fund managers are thriving and doing well. Lawyers got to get paid, too, right? These guys are levered up on fantastic real estate all over the country, and they've got to keep driving fees. And one way to do that is suing big companies and getting big settlements. And so, yeah, they've got to get in on the AI boom somehow. I mean, there's another lawsuit going on between one of the legal AI companies and one of the legal data providers saying that there's like these case notes that kind of define the case law and summarize it.
36:52And they say, hey, you scraped all those and you're reproducing those. And so this is gonna keep happening in every single industry until there's a lot of, like a very clear path to how AI companies can partner with different troves of data and different intellectual property. But I certainly hope that in the long term, these products are monetized in a way that there's enough money to go around to kind of keep everyone happy. Like one of the cool things about TikTok was that they started by bootstrapping the lyrics and the music off of the iTunes preview API. Have you heard about this? We talked about this, I think, during TikTok.
37:30So iTunes, when you go into the iTunes music library, you can pull up a Taylor Swift song and it costs 99 cents to buy the song,$9.99 to buy the album. Now everyone just streams everything and it's an all you can eat plan. but you can still, and many years ago routinely did, buy single songs for 99 cents. And that was the ITS promise, right? But you could play a preview of that song and they would give you 15 seconds of the song for free. And so I believe apps had access to that API and could pull in song previews basically for free. and so those little snippets became very easy to create vertical video lip sync videos around and it was all that was hugely innovative because before that everybody's spending hours a day on social media but you had no none of the the music that people really loved yep it was just a huge it was just a huge missing element of content well yeah and then on youtube like you could just download a song, put it in Premiere Pro, export and upload it with, and you know, have your, you know, driving video or vlog with like a popular song, but you would get flagged and actually demonetized.
38:47So you wouldn't be able to make money off that content for a lot of really professional creators. That was not an option because they're making thousands of dollars in ads on every video they upload. And for, for other folks, you could get content strike or content warning. You could actually get your channel deleted if you did it enough. It was a risk. And so now there's services like Licked that allow you to pay for rights to licenses to really popular songs. But also there's a lot of libraries out there that are licensed by the social media platform. So when you go and you use a popular song on TikTok, that artist is actually getting a cut of that ad revenue and it's all happening seamlessly on the back end.
39:24Your creator payout might be a little bit lower, but you don't even notice. And so the real good ending here would be a situation where hey chat gp is making 200 a month from people generating studio ghibli photos uh and studio ghibli is making a couple bucks from every person that's doing that and and there's enough to go around and everyone's getting paid um but we'll see yeah the real the real company uh and this was going around i think justin uh texted us something about his is fiverr being just absolutely like the the i'm interested to go in public and see what uh this was something that you would normally go to fiverr for find an illustrator to do a little one-off thing for you their logo is it down i mean they're down to 2.7 today hard to say if that's day but i i want to know more about like over the last two years historically the best way if you wanted to get a nicely photoshopped image hey swap this person's face onto this person's face totally and you didn't couldn't do it yourself really good at photoshop somebody do it for five dollars that was the best way to do it yep and so they're just gonna have to go uh more upmarket um but uh anyways well uh should we go through some of these let's go through some of these studio ghibli reactions because they're all golden the timeline's been on fire this morning we had a lot of fun try to add to it in creative ways some hit some flopped but that's the name of the game uh you had a hit on your hand hitter with uh i mean this is this is just so perfect because it feels like it it just genuine when you have the text there it feels like it's actually a real scene totally it is so cool that it kept the text and the text looks fantastic it's not wobbly or rough around on the edges, like that is text.
41:15But at the same time, it's text that you would expect to see in a cartoon or in an anime, and it's not actually perfect perfect. Like there's slight changes in the kerning, like I'm looking between the C and the A there. And it's like, they're slightly together. But it's so good that it looks like it was done by hand. It looks way better, way better. If you think about the font generation in these previous models it was always the worst yeah it was always the fastest thing to expose that it was ai slop oh yeah yeah yeah i remember i remember see a picture of a car and immediately go to the license plate and say oh what does the license plate look like garbled junk if it does then it's ai yeah when we had lesson and seth on to talk about just ai broadly yep uh uh seth came on with this background yeah and there was a bunch of text in it and it was all just it was like it It was like the vibe was correct.
42:11It was like generally like, okay, there's some San Francisco text back there generally, but you'd get repeated characters, you'd get fonts out of alignment, you'd get extra rows or stuff would just kind of trail off, but this nailed it. So Sam Altman saying he's doing this because he loves it. I mean, at this point, I believe him, honestly. He's got to be loving it today, John. This is like, it's the best technology to be working on. It's so much fun. And like you get to drive these like massive moments online. And like, yeah, obviously he's getting paid now, but that's good. I want the for profit.
42:43This is going to be, have a bigger impact on open AI's consumer awareness than any Super Bowl ad. Totally. So even if they lose$20 million today from everybody and their mom just like generating images over and over and over, it still doesn't matter because it's the best marketing that they could possibly have. And it really shows this cycle of like, it is a really, really intense race right now to level up the models, launch new products. There's clearly different forks in the tree that people are going down, whether it's image generation, deep research, just humor and like chat, just natural feel.
43:24And OpenAI seems to be the first one to hit these breakthroughs, the first one to issue, to do chat GPT that felt like you passed the Turing test. This feels like a crossing the chasm, getting out of the uncanny valley moment for these AI generated images. And you can imagine that Elon is meeting with the XAI team today. We need this yesterday. Why is it OpenAI dominating the timeline and not us? Yep. It's like we actually, and this will be the cool thing. You can imagine a very close future in which you see an iconic image on X. you can immediately generate oh yeah like some variation of it and repost it the grok button right there should just take you and be like hey what style do you want to restyle this image like yeah get a prompt or use one of our pre pre pre-populated templates you should just be able to click click a button and say yeah i want this to be cartoon i want this to be cgi or 3d render or something or whatever else um it's kind of crazy how much the ghibli style has dominated to the timeline today because they launched a whole product that you can do a ton of different stuff in but everyone seems obsessed with this.
44:30It really became a meme and it really shows like our propensity for like aggregation. It's a shelling point, I guess. It's like everyone kind of gets this one and so even the Dr. Seuss posts which are funny, you haven't shifted the entire timeline to that yet. Maybe it'll happen. So we got some more iconic tech pictures, Studio Ghibli edition. We got Jensen Wong. signing a female's body uh very accurate picture and it looks i mean it's funny this this treatment just it's so funny in this style yeah uh but arun absolutely went off i think he had one of the better oh yeah he was this was last night oh yeah and he also posted one of us oh really in that same thread that's so nice of him that was sweet um but yeah yeah he has trump in front of the the mcdonald's got that got a thousand likes on that yeah it was this one you remember we were in the same thread oh he posted that that's great yeah i love that one yeah um delian posted a picture of his uh wife nadia who's actually coming on the show on friday she has a new book out um you should go pre-order it um it's called anti-mimetics why ideas don't travel which is very interesting tell you're always snapping into the borat yeah he loves the you can tell he's genuinely like thrilled oh yeah because he's like my rockets
46:00my wife and yeah i mean it's a fantastic photo uh and and again just like even the painting in the background is is like just super high fidelity but also uh fits in the fits in the world and the vibe of the Studio Ghibli stuff. Sam Altman has been enjoying the moment. He says, be me, grind for a decade trying to help make super intelligence cure cancer or whatever. Mostly no one cares for the first seven and a half years. Then for 2.5 years, everyone hates you for everything. And then finally wake up one day to hundreds of messages. Look, I made you into a twink Ghibli style. Ha ha. Wild. I love it.
46:42Banger. When you're having a good day on the timeline, when you got the hot hand, you got to just keep posting, have fun. Yeah. I like this image of Alec Baldwin in Glengarry Glen Ross. He's wearing a gold Rolex presidential. He says, look at this watch. This watch costs more than your car. He's giving a motivational speech to the car salesman at this dealership. And it reminded me of bezel where you can shop over 23 ,500 luxury watches fully authenticated in house by bezel's team of experts. Go to get bezel.com and every brand. I mean, this is like just like a freebie today. Just take your brand images, throw them through the studio Ghibli filter, put them up on the timeline.
47:21You know, your job, you can, you can go home early as a social media manager today. You can have some fun. Yeah. And I saw on X, jeremy jeremiah lancaster was working on taking my dr seuss image of mark andreessen and swapping a patek minute repeater onto his hand there we go so it's still in the works but i'm excited to check this out you had another post uh the uh the good morning with the saratoga sparkling water the uh the ice bowl the bowl of ice water and the bananas feels very calming and there was something else going on here. You said that, not to go back to the Ashton Hall viral morning routine that broke Monday, but you said that there were some brands that were partnered that did not go viral.
48:07And there were some brands that were not partnered with him that did go viral. And so - Painful moment. So I posted this and a company called Hostage Tape responded and said, I think you forgot something. And if you actually think about the video at the very beginning of the video he's taking off his mouth tape i don't know if you've ever tried sleeping with mouth tape i haven't patrick o'shaughnessy recommended it to me yeah didn't take his advice but i've gone back i've gone back and forth on it i i i think it's worth doing it's certainly worth trying it's kind of annoying to to tape up your mouth but it just helps you breathe through your nose basically so that you're not mouth breathing and does that mean so hey if you're a mouth breather, try some mouth tape.
48:48But super painful for hostage tape to be in that video, be so front and center. And it seems like they're sponsoring him, right? Maybe. Okay. But nobody is - It's unclear because I would imagine - Nobody's memeing them, basically. I know that you can just use athletic tape. You can just go buy cheap athletic tape. Hostage tape is like a premium version of that. Yeah. And I think hostage tape might even have some branding on the tape. But when I saw it in the video, I didn't notice Hostage's brand on the tape, which is kind of a weird problem. Saratoga Water, of course, has their logo front and center the entire time because the blue bottle, not only is the blue bottle iconic and it doesn't just look like any bottle of water, it looks like a blue bottle.
49:34And so you're wondering what that is. But the Saratoga logo is very clear throughout that entire video. You see it constantly. And I didn't see the hostage tape logo. And so a little bit of a miss if they are partnered with him. But his videos seem to continually go viral. So I imagine he will drive a lot of sales there. The next routine video, he's got to go and say, hey, I average a billion views per video. And I imagine that if you Google mouth tape right now you will get an ad on google for hostage tape and maybe wind up in that funnel and so hostage tape is probably going to have a good month just because the overall awareness of mouth taping went up this month and so they should be doing well and already i think it's a fantastic business i think they're doing like over 100 million yeah selling tape it's pretty good 95 gross margin pretty pretty good good business anyway uh zero interest rates rahul sunwalker the founder of Julius AI was spotted on the timeline studio giblified with his buddy growing Daniel and their boxes outside of X headquarters after Elon Musk took hold.
50:48Dwarkech Patel and Leopold Ashenbrenner It's still under discussed that he's holding up I think Michelle Obama's book. He really leaned into the bit. It's great. Leaned into the bit. I love it. Leopold Ashenbrenner and Darkesh having their chat iconic moment in tech podcasting and AI after situational awareness dropped. We were talking about essays with Lulu that had major impacts and we had identified going direct and founder mode. And I think we were digging into this more with David Sendra this morning. And I think that it's not so much that those essays were groundbreaking. It's that they were compressed into memes that had even more impact than the essay themselves.
51:35And so I think people use the term founder mode more than they actually reference the ideas in the actual essay. And Leopold's essay, Situational Awareness is kind of the opposite. That essay was really, really widely read in AI circles, but people don't use the term situational awareness as a meme in the same way that they use founder mode. but it's interesting. A little bit of a Coogan's Law moment there. Leopold should have come up with something a little bit spicier maybe but I don't know. He's usually does his brand for his fund and I'm sure it's doing well. Luke Metro posted this. Really works on everything, huh?
52:14And do you know who this guy is? I'm blanking on his name. Mike Vining. Mike Vining was a special operator for the US military and is the basis for the special operator in Sicario who wears the glasses. and is in that car scene on that bridge. And so if you look up pictures of Mike Vining, you will see, this is his military photo, but you'll see a photo of him in very thick glasses in a white button-down shirt and a tie looking very buttoned up. He looks kind of like Ben Affleck in The Accountant. He looks very kind of just dorky almost, but he's known as the deadliest operator in history. is that like being like serious serious like operations and is just like a fantastic shot and a fantastic yeah operator and and yeah there's something i think what luke's hitting into here it's extremely easy everybody's one-shotting these yep and the style is just very fun and iconic and it makes people have seen so much of studio ghibli yep online and memes and stuff like that even if if you don't know the show or you're not a fan of the show, the style is just very cool.
53:28Even just like repurposing the memes from even just a few weeks ago, there's this White Lotus meme that Brandon Gurrell over at PirateWire shared. He took that, this was all over the timeline after that episode dropped and now it's been Studio Giblified and it's very fun. I was just thinking of one other image that you generated that won't make the cut. That was from a screenshot of a post from a long time ago, over a decade ago, that was a good one. Bad content for the show though. But yeah, that's another thing. You don't have to post them on X. You can generate these images and share them in your group chats.
54:03Just share them with your absolute boy. I was telling Jordy that, you know, sometimes it feels really good to get a thousand likes on a post on X, but sometimes just getting five heart emojis in a group chat of six people, that'll do the trick. That's the dopamine you need for the day. Did you see this Studio Ghibli image of a billboard on the 101 in San Francisco? I did. A lot of people have been talking about this, John. This was a really cool one, and it reminded me of AdQuick, because if you wanted to do a real-life billboard like this one pictured in the Studio Ghibli image, The only place to go is AdQuick.com.
54:35Adquick.com. Adquick advertising made easy and measurable. Say goodbye to the headaches of out-of-home advertising. Only AdQuick combines technology, out-of-home expertise, and data to enable efficient, seamless ad buying across the globe. That's right. And the Studio Ghiblification continued. Stage four, the sign bears no reality. no relation to any reality whatsoever it is its own pure simulacrum and of course those bizarre jd vance is a child memes have now been giblified and so it's just a snake eating its tail and or a boros of slop at this point but uh there's still a little bit of jd in there and i'm sure he's having a laugh looking at those one on the left is too good it's really i want i want versions of of us with those we'll go to the next slide we're we're getting there no one has made us into children and then studio giblified us but uh i mean we kind of look like children in some of these it does look we do look very childish here so it's us at the desk with the polycom they nailed it and the and uh and uh and a tv in the background that doesn't quite exist it it got some of the background objects correct, but it nailed some of the podcast mics.
55:46It definitely got the idea of what the show looks like. Us at a table in these chairs, wooden desk, polycom, suits, white suits. Of course, we're not wearing the white suits today, but it looks pretty accurate. Oh yeah, Ben's showing you what we look like and then what they look like. And this is one of those iconic tech images. So we were very appreciative to be included in this chain, this thread, right above Zuckerberg with the chain. Looking good. We already watched the Sebastian vocal thing. Dwarkesh enjoyed his Studio Ghiblification. Dan Romero shared the PayPal Mafia in Studio Ghibli form.
56:24You got Peter Thiel, Keith Raboy in there. Such a cool image. It's a great image, iconic. And did you see this image of this beautiful house? I mean, this in Studio Ghibli form just looks fantastic. It looks very warm inside, very welcoming. I wish I could stay there. And guess what? You can on Wander.com. Find your happy place. Find your happy place. Book a Wander with inspiring views, hotel-grade amenities, and dreamy beds, top-tier cleaning, and 24-7 concierge service. It's like a vacation home, but better. I like that Augustus gibbified himself for his Florida Senate Rules Committee hearing.
57:08He testified. He looks great there. Looking a little surprised. And I'm sure he was surprised because... By some of the comments that the audience were giving. By someone who shared the video with a little takedown of Augustus. What a username. We're not going to say this. Hustle, B-I-T-C-H. It says, ever wonder who's really behind weather modification? this guy is the CEO of one of the most powerful geoengineering companies in the world. That's right. A rainmaker from El Segundo. He showed up to Augustus Dorico, folks. Augustus Dorico. Coming on down into the temple of technology. We'll have to have him back on the show to discuss this and put this guy in the truth zone.
57:53I mean, perfect marketing for Augustus. Really, it's great. I mean, you can't pay for this. For every 10 haters, you get one diehard new fan. So I wouldn't worry about it, Augustus, but maybe put on a bulletproof vest because they're coming for you, boy. He just showed up to defend cloud seeding, chemical spraying, and even electric weather manipulation, and he's fighting to keep it going in Florida. Why is he pushing so hard for this? What are his motivations? Why would he want to do this? We all want to know. Yeah, Augustus. You've got some questions to answer, man. and there's already a community note on it We'll see how that bites out in the comments.
58:37I think it's proposed community notes. Yeah, proposed. Because we're X-Hall monitors. Yeah, we are. If you ever need help community noting, we're happy to. Community note, he's the best dude ever. He loves White Monster. I love him. He's the best. Well, we have some breaking news today. Okay. Should we finish up with two more slides? Sure. I was going to say we have Ahmad joining in less than a minute. Okay. We're already there. But what was the breaking news that you had? Oh, just that Ahmad is raising a trillion dollars. Oh, yeah. One T. One T. And that we're going to have him on the show. And we're excited about that.
59:20Do I have a slide? What's the actual announcement? We should read through that to introduce him. He is the founder of Mercury. Do you have the... I'm pulling up his post. We actually discussed this on the show when it was breaking or when it was leaked, but the news is out now. They're announcing a$300 million Series C in a mix of primary and secondary funding and a$3.5 billion valuation. Congrats to the whole team over there. They're excited to bring on Sequoia as the lead investor. I like the soundboard's better than the physical thing. Here comes the boom, ready or not. How do you like me? I'm awesome.
1:00:01they got spark capital marathon mp co2 crv and andreason coming in and participating in this round uh they say since our founding in 2017 we've been on a remarkable road to reimagine banking to truly serve we got him in the waiting room of ambitious founders welcome to the show ahmad thanks for joining yeah there he is having me uh it's always cool when you listen to a podcast and now you're on the other side and you can't i love it you're live i'm so glad you've been listening i I hope you've enjoyed the content. Great to have you here. Can you give us a breakdown of introduce yourself, introduce Mercury, introduce what's going on today with the round?
1:00:38Yeah, I'm the co-founder and CEO of Mercury. Mercury provides banking to startups, e-commerce, and professional services. We started in, I guess we launched in April 2019. So relatively young company, but a ton of startups use us, especially after SVB and that kind of failure. and then we're just announcing our CBC. Sequoia is leading the round with Spark and Marathon coming in as new investors and all of our previous kind of VCs are participating. So it's a 300 million round with a combination of primary and secondary. We're having Andrew Reid on the show later. Who at Sequoia did the deal? Who's the ring maker?
1:01:17Yeah, Sonia Huang. She's been doing a ton in AI, including Langchain and Glean and she knows a lot about FinTech as well. excited to have on board. Talk about what it's like now raising these rounds that are in the hundreds of millions of dollar range. Is it still kind of feel very similar to the early days where you're just like hey let's grab coffee and hang out and you talk about the business. I haven't built a spreadsheet in years. It's a surprisingly similar process to an earlier round right. I did this in about four weeks. And, you know, at least as a founder, you're really still like pitching the big vision, like, how can this be, you know, another 100x, right?
1:02:02Like, that seems hard to believe, like, that's 350 billion. But like, I want to do that. It's the bit that's different is there's a lot more support, right? Like when I was doing a seed round, it's all by yourself. Whereas now I have a finance team, I have a data science team, and like, a ton of stuff after that first meeting is done by them it's like you know you have to have a great data room you have to paint this picture of like where's the finances going in the next kind of year or two uh but you still always have to tie it back to like how how is this going to be like you know even bigger and what's the long-term vision and and paint that picture yeah just talk about uh this moment for you and the team i mean the svb crisis was obviously net positive it was you know tumultuous but you guys benefited in a big way from it.
1:02:49But then there was also an entire, maybe it felt like a two-year period where it seemed like fintech was just over, everything in the public markets was so bad, but at the same time, interest rates were way up. So you guys were benefiting from that. Does this feel like you guys feel, you know, finally kind of like validated in this moment to say like, hey, we're getting, you know, repriced well beyond where you guys were at in 21. And you're sort of like here to stay, it feels like. So we've been profitable for 10 quarters. And I really want to build a long-term company. I really like the time where no one cares, right?
1:03:27Like I think that's a great time to, yeah, it was a great time to hire great people, right? Hiring in fintech was rare. There was amazing fintech talent out there. It was a great time because we didn't have that many competitors, right? Actually, I'm always worried that, I think actually this is surprisingly uncompetitive space for how big the opportunity is. Like banking in the US makes$2 trillion. Like this is an insanely huge market. And I don't think it's like as competitive as like fricking B2B SaaS, right? Like that's actually a smaller market. So I like those quiet times, but obviously it's nice to get like a big valuation boost.
1:04:02Our last round was at 1.6 billion valuation. This is 3.5. And you know, we made a ton of progress. We were a relatively small company. We had 140 employees back then. I think we announced we had about 40 ,000 customers. Now we have 200 ,000 customers. We're almost 860 employees. So we made a lot of progress. It's nice for, I guess, the world to see that and us to be able to talk about it publicly. How do you think about Mercury is in many ways the bank of X, right? Like it's almost like it'd be hard to have been on X the past few years and not be aware of Mercury. Like you have a big following, all this stuff.
1:04:39how do you think about reaching founders that maybe aren't on on x and and kind of like scaling up right because every single day there's entrepreneurs that that walk into banks and they are building a startup and they just walk into chase or bank of america and they create an account and how do you think about kind of reaching that next generation that that maybe isn't sort of hyper online or or uh already in the community yeah we've been kind of you know i've always thought about Mercury as targeting digital entrepreneurs. So that's beyond startups, that's e-commerce, that's professional services, like anyone that spends all their time in front of a computer to run their business.
1:05:15And these people, you know, most entrepreneurs have friends that are entrepreneurs or they are in communities. So it's not just on X, right? Like it's Facebook groups, it's YouTube, it's Reddit, et cetera. And like Mercury is all over those things as well, right? Like we, more than 50%, about 60 % of our customers come through organic word of mouth distribution. And that's not just in the startup space, that's in all of these spaces. And our NPS is 82. So customers really like Mercury. No one's out there saying like, go use Chase is the best customer experience ever. Like you'll love it, right?
1:05:49Like Mercury kind of stands so far apart from kind of these incumbent banks that we are still growing from that core word of mouth. But we do invest in other things as well. Like we've built a bunch of kind of virality into the into the app and the products. So last year we launched invoicing and bill pay. And part of that is that those are like viral payment methods, right? If you send an invoice via Mercury, often you're sending it to another business. So obviously they can just pay it, but they can also sign up to Mercury. And the next time they have to pay an invoice, it's instantaneous and like super easy to do.
1:06:20So trying to think about these things as like creating this kind of payments network that like every business and now consumer as well as part of creates this kind of like added. uh yeah talk about focus maybe i think people had been asking for mercury sort of personal banking for so long and it was an obvious opportunity for you guys it felt like something you could have sprinted towards and launch how did you you know what was kind of the thought process there was there a certain scale that you wanted to be at before you launched personal or what did that look like yeah it was definitely like the biggest feature request for us for a long time uh yeah it's my it's definitely a matter of like execution doing it at the right time uh you know we did it with a relatively small team.
1:06:59I'm a big fan of kind of doing these new things with initially fairly small teams and seeing if it works and then investing. So I think we had a five-person team. One of the drivers was we found this amazing guy, Alexei, who previously had his own kind of startup, Neobank, in the consumer space, and then he helped Albert launch their consumer bank and then did the same at Mercury. So yeah, having great leaders, what's the term, Keith Reboa likes having shotgun barrels or whatever, is part of like knowing when the right time is. But yeah, we do a lot, but we also try to be really focused, right?
1:07:35Like banking and doing an amazing banking experience for startups is the core. And that's where we have like the biggest investment. Can you talk a little bit about underwriting at the earlier stage, these online businesses? How is that different? Is it riskier? How are you thinking about the risk profile of like a non-brick and mortar customer coming on the platform? Yeah, I mean, we don't give them, loans so like underwriting isn't necessarily about like credit risk but it's about like KYC and KYB like how do you really know someone it definitely startups actually easier right like people have a pretty big digital presence they have better websites they have a LinkedIn they have x etc and you know their descriptions their VCs like if it's you know for Sequoia funded company shows up you can underwrite that very quickly I think it's a lot trickier when it is like a professional service consultancy and they don't have online presence and you have to kind of learn more about it quickly.
1:08:33There we think of like a tiered approach, right? Like if we understand less, we'll ask you more questions, ask for proof of address or that kind of thing. And then we end up setting like kind of lower limits and things like that as well. So a startup will have a lower risk tier. We'll give them a lot of flexibility. A company that we know less of will initially they won't be able to send like a million dollar wire they'll have to kind of earn their way to building up the understanding and the kind of like uh features we give them yeah how do what do you think is the future of uh venture debt you know over the last year it's been in the news because of some sort of high profile shutdowns but obviously it's an important tool is that is that like it has you know become a meaningful part of the business over time or is it just something that you guys want to have available, but it's not a huge focus?
1:09:25Yeah, venture debt is actually like a small-ish market. People talk about it in the VC community, but even SVB, which by far was the biggest venture debt provider, I think they had like a$7 billion portfolio when they failed. So I think there's a good way to do it and a bad way to do it. A good way to do it is like your revenue-generating company and maybe there's some cash flow need because your AR takes a while or you have some equipment, et cetera, and then you do it, that's great. I think the cases where I don't love it is you're some AI company that's not making any money today and you want to go extend your fundraise and runway by 20 % or something.
1:10:09I think in those situations, when you run out of that, as you start getting into that 20%, it reduces your options. When you have a fundraise, a VC fundraise, and you're in that last 20%, sure, you do layoffs, you kind of reprioritize, you cut your founder's salary, you do whatever the hell it takes to survive. But if you have this fixed burn rate that is caused by venture debt, that's not a healthy position to be inside. I think if you use it in the right way, it can be quite beneficial. My last company, I raised a million dollar venture debt and it saved the company and actually got us to profitability.
1:10:44So it's really important for us. And I think people kind of, with most things on the internet at least, Like people tend to say these extremes, like never do it. I think it's like, is the right thing at the right time? That'll be the capture we use on this clip. We do provide venture debt. Yeah. When we post this clip, it'll be like breaking news. Never use venture debt ever. Watch for the whole reaction. I had a question about interest rates. Do you want to go first? Go for that. Yeah. I want to talk about how interest rates affect your business and how you think startups could prepare for this.
1:11:16I mean, high interest rates, you're making more money on interest. Low interest rates. there's more business formation, more venture capital flowing in? Does it just net out where you don't care? Do you have a preference? What's your view on interest rates right now? I mean, obviously, we can't control it. So under every circumstance, we'll try to win. I think the ideal interest rate for us is something like 3%. I think there's lots of VC funding and business formation at 3%. But we also have the opportunity to make money on our deposits better at 3%. And, you know, for us, and I think most startups, like, I think you want to operate, like, whatever the macro economy is, and like, try to set yourself up in that way.
1:11:56So, you know, we, we grew a lot when interest rates were zero, especially during kind of the big, big bubble. And, you know, there was definitely like a struggle in terms of growth during like some of that 2022 period. But, you know, since then, it's picked up, and we've obviously grown our market share and things like that as well. you've invested in hundreds of other startups. I imagine the first of every month is pretty chaotic in your inbox. You're just getting flooded. You're like, great, I have another 300 emails to get through. Talk about your sort of personal investing process. I imagine you have such an amazing, you have one of the craziest networks in startups because so many of the best companies in the Valley are running on Mercury.
1:12:40But talk about your personal process. You're busy running a multi-billion dollar business that's doing hundreds of millions of actual EBITDA. At this point, are you ever making investments where it's just a good friend or a founder you back says, hey, you should invest in this company and just email them and say, sure, I'm in? Or is it, what does your process actually look like? I talk to every company I invest in. I probably do like 20-ish this year. I've done about 300 altogether. The main reason I do it is I just like talking to entrepreneurs and hearing like these crazy fun ideas. Like I've done a bunch in space and defense and things like that where, you know, it's very new to me, very different from FinTech and it's always like a great learning experience.
1:13:21And I really want to give back to entrepreneurs. Like I've been doing startups since 2006 and I think it's a great way to actually be helpful to people. But, you know, if I invest, I really want to believe in them. Like my rough criteria is if this person texts me at like 5 p.m., am I going to be excited to like do a call with them or am I going to be annoyed that I have to talk to them? So like, I definitely want to do that 30 minutes to like, uh, have like a bar of saying, okay, you know, I will be excited because I like, I love this space and I want to help this person out. So, uh, yeah, that's definitely a key.
1:13:51Uh, and yeah, uh, yeah, it tends to be a broad set of kind of things and I'm still an active, active investor. So hit me up if you, if you're raising. No, it's amazing that it's actually, most people have to like really justify, oh yeah, I'm angel investing and it's gonna, you know, benefit me in these ways and maybe get some return. but for you it's like hardcore customer development feedback, talking to customers and maybe you invest as well. So very cool. You got anything else? Yeah. Can you talk about how AI is impacting the business? I'd love to know. There's the boring stuff, which I feel like is just LLMs are great at text processing.
1:14:25I'm sure there's a lot of stuff that's no longer paperwork. Maybe you're scanning documents, OCRing them, throwing them through LLMs. Is that working? And then how does that compare to just having your engineers be more performant because they have cursor, clog code, and devon, cognition in the mix? What does the AI strategy look like right now? Yeah, I think the most exciting kind of short-term stuff for us is the back office stuff. So, you know, my part of the thesis of starting Mercury was like, what if the marginal cost of serving the next user is zero, right? If you think about a normal incumbent bank, you're walking into the bank branch, you're talking to someone, everything is actually quite expensive for them to serve a new user.
1:15:06And that's why they have these on-risk fees for small businesses, right? It's just hard for them to make money from a small business. For us, it's like, how do we get it to be a database row? But it turns out there is a lot of customer support, there's a lot of compliance, there's a lot of onboarding, and a lot of that is text-based stuff. So how do we make it so if someone submits a utility bill to prove their address, we can like automatically process that and not just it's not just cheaper it's also instantaneous right like instead of waiting three days for a human to look at it we can give them a great answer straight away uh so lots of applications and compliance risk uh and onboarding and uh support and lots of it we have rolled out already lots of it we are rolling out i'm really excited about it i think it's like a genuine improvement for customers uh you know engineering productivity i i haven't seen as much of a boost as i would like to see i think it's really good for making an app from scratch.
1:15:57When you have a huge code base that's all interlinked, it's just not as good. We also built in a weird programming language, Haskell, which has been great. We're amazing engineers, but I think a lot of these coding LLMs are not trained on Haskell as well, so when we were a late adopter, at least on the back-end side. And then lastly, I think the user features. There's some interesting stuff coming. I'm not a fan of just shove a chatbot in front of it as an interface. I think it's easier to have structured data and be able to do things. But we've done some stuff around search and OCR. And I think in the long run, if we have your finances, if we have your financial workflows, we have all that data.
1:16:38There's lots of ways to help you automate your business, help you speed up getting insights and things like that. So, yeah, I think there'll be lots of really meaningful ways to incorporate AI that customers like. yeah on that last last note there are you guys thinking about uh llm uh result optimization you guys benefit you said something like 60 of your growth and new customers are from word of mouth obviously llms are just another form of word of mouth because they're just sort of like taking all this you know content and then you know sort of organizing it and channeling it how are you guys thinking about that or is it just happening by default you mean like a founder that goes to ChatGPT and says, what bank should I use?
1:17:20What bank should I use? And you hope it says Mercury. Yeah. Yeah, I did a tweet or ZEAT or whatever they call it now. And, you know, we've got a graph of like how number of signups is going with LLMs. And it's like fully exponential, especially in the last two, three months. Still a small base. But, you know, if the growth rate continues, it'll be significant. And I'm using ChatGPT to ask about services and what should I use. And Mercury actually just naturally gets shown up, right? Yeah, that's what I would have expected. The content is like this kind of community-based. So if you ask for it, like, what's a bank I should use for my startup?
1:17:52Like, you know, Mercury is like likely to come up. One thing that I'm thinking about is like, how are you even more active on these communities, right? So obviously I'm on X and I'm very active there, but a lot of our conversations that are happening on Reddit are not like with me or someone at Mercury kind of participating. So I do think you can change the sentiment and like improve your visibility by being active in communities. and I think that's like a big part of like how you influence LLMs. I think like you want to basically create like real content, right? Like it's a new type of SEO. This is like community-based kind of SEO and recommendation.
1:18:26And I think that's like, that's like probably genuinely aligns with like building a good online community anyway. That makes no sense. Well, thanks for stopping by. Congratulations on the round. Big milestone. Massive milestone. Happy for you and the whole team. And yeah, good luck to you. We'll talk to you soon. Talk soon. Bye. On the Fed rates issue, he wouldn't take a stance, but thankfully that's why we have Polymarket. Polymarket has an open market on what the Fed will decide at their May 7th meeting. This is the FOMC meeting. It's 41 days, 16 hours, 40 minutes and 27 seconds away. I know you've been counting down, Jordy.
1:19:07What do you think Polymarket's expecting? Right now, the options are 50 basis point cut, 25 basis point cut, or no change. The market's pricing, no change at 86%, but they are expecting a 25 % basis point cut on the next meeting, June 18th. Yeah, it's wild. The volume on this is pretty substantial. there's almost nine million of volume just on this one polymarket. Wow, yeah. Useful tool if you're monitoring rates. I think Ahmad's probably sleeping pretty well these days with an extra 300 million in the bank. That's right. He's probably sleeping on an eight sleep, folks. Go to eightsleep.com. Nights that fuel your best days.
1:20:01Turn an e-bed into the ultimate sleeping experience. Do you have a snore sound on there or something? Or just the gong? Okay, wake up to the gong. Wake up. I bet there's a way to add a custom sound in here. And we also got the Studio Ghibli photo of the 8th Sleep. Looks fantastic on the previous slide. I really enjoyed making these. Yeah, they should throw that up on social. Throw it up. I'll send it over. I got a 95 last night. My 8th Sleep autopilot made adjustments to the temperature of the bed overnight to boost my REM sleep by a whopping 17%. So double digit increase. You're doing great. Thank you to Eight Sleep for supporting the show.
1:20:45You can get an Eight Sleep by going to eightsleep.com slash TVPN. Get$350 off your pod. Just do it! Totally clipping the audio. Someone just had to take out their headphones for that one. Rough. uh well uh let's let's let's move through some of the timeline that we got more guests joining in about nine minutes uh salma says okay gpt4o cooked i fear i generated this creative ad for bic pen the prompt was create bic creative ad add the pen as a hair accessory used to tie models hair up compelling message wow they didn't even she didn't even say what ad copy to use yep and it says twisting hair since 1945 and that's a pretty good ad now with these things you know you're in this like uncanny valley where it's possible that the prompt was a little bit more like you never know if somebody did a little bit more work on the front end right this is this is what's happening with like the vibe coding thing where people are saying oh i just vibe coded this game and it's like yeah they actually spent more time on it but i think this is real i i think this really was the prompt and i think if you went and tested this right now you would get there and it's photo reel not uncanny valley at all and it looks great and uh yeah if i was bic pen i'd be printing this out and putting in the new york times send it to the printer send it to the printer um anyway uh it's a knockout drag out fight for the next format of studio ghibli web weaver deep fates is advocating for uh rick and morty anime which is terrifying in my opinion i don't know if we can show the full image, but it's very funny and it does look exactly like Rick and Morty.
1:22:28I like this Studio Ghibli of the New York Stock Exchange on the next slide. And it reminded me of our sponsor, Public Investing, for those who take it seriously in multi-asset investing, industry-leading yields and they're trusted by a million folks. Go over to public.com and sign up. Fantastic. I like that one. Theo says, Sir, another Gibb-lified historic photos. I wasn't done doing an ad for Public. Oh, you were? Okay, keep going. Public.com. Okay. You got it. All right. That's enough. Go to Public.com. Thank you to Public for supporting the show. People are having a lot of fun with these Studio Ghibli photos.
1:23:10There's George Bush receiving the information about 9-11. I actually don't think... Good effort. I don't think they did Dabia that well on this. It just doesn't really look like him. Yeah, you're right. It's missing some of it. I've seen it botch the likeness thing a couple times. Well, the Fiverr illustrators still have a job, I guess. Yep. Mimetic Theory, Mass, who has been on the show before for that fun swing project he did. if chat got if chat gpt has rejected gibbifying multiple of your ideas you're not trying hard enough you were running into that yesterday yeah big issues there someone gibbified his his uh his swing which is very nice to see um sam altman says tremendous alpha and images and chat gpt right now it's very true the new version of images and chat gpt is still rolling out so please try again later today if you don't have one he says it's an incredible technology product i remember seeing some of the first images come out of the model and having a hard time uh thinking that they were really made by ai we think people will love it and i think they have what they're really yeah this is why i was saying this is you know very clearly sort of sputnik uh uber for dogs kind of moment for ai yeah exactly it's like sputnik of iphone moments little deep seek moment sprinkled in sprinkle a little deep seek moment in there yeah uh should we talk about nikita beer he's making moves he's over at solana now officially an advisor yeah officially an advisor not a full-time employee but uh cool moment i think he had been a pretty big critic of crypto historically he went on this sort of uh generational uh shit posting run where he just kept saying like i'm gonna do it i'm a logic coin lunch coin he never did i don't think he ever did okay well maybe it's coming soon then and uh anyways uh big big pickup for solana yeah uh solana especially after the rough ad it's just like a new it's a new surface area for nikita yeah it's it's you know undoubtedly one thing that we know is some of the best social products or many of the best sort of social products are providing entertainment and crypto is very good at providing entertainment, right?
1:25:32Anybody that's sort of invested in one of these tokens, it's like you're watching the chart all day long, you're following it, there's sort of this team-oriented element to it, so. And yeah, I mean, when you think about the hope for pretty much every cryptocurrency project over the last decade has been, let's cross into the average users on the consumer side. Not purely just as an investment vehicle, but let's help developers create crypto projects like Polymarket that really, really broke through during the election cycle and was just a news source for people basically. Even though there were crypto rails underneath, I'm sure Solana wants to power that.
1:26:10And that's what Nikita has been writing about online for years, talking about how to grow consumer apps and specifically take advantage of a bunch of different APIs and iOS, figure out how to link contact books and whatnot. All techniques, growth hacks that should benefit Solana if it's building its own apps or developers in the ecosystem. And so Grok summarizes it saying this development is viewed as a significant boost for Solana, particularly due to recent favorable changes like regulatory clarity, increased app store openness, and the upcoming launch of Solana's mobile device Seeker. The community is highly optimistic seeing this as a critical moment for Solana to expand in the mobile app market with expectations that consumer crypto applications will see accelerated growth.
1:26:58And so Nikita writes, while the jury is not out yet, Solana has the fundamental building blocks for something to break out on mobile and certainly many apps are making headway. So today I'm joining Solana as an advisor to help select companies launch and grow their apps. So let's hear it for Nikita. Some personnel news on the timeline. You love to see it. Some people making moves. Spoke with Nikita last night over DMs. We're going to get him on the show at some point, which will be cool. And speaking of crypto coins, we have massive drama going down on Hyperliquid involving jelly. So for those that don't know, Hyperliquid is a sort of decentralized exchange.
1:27:41It's very popular right now. It's already publicly listed itself. It trades at around a$14 billion valuation. uh and so this one is in there's some drama involving uh jelly jelly which is a token that was launched by a slow ventures portfolio company so sam lesson was was helping them out with it uh which remember we covered it sort of ran up to molt i think 300 million dollar valuation in 24 hours or something crazy like that it came back down to earth um it's been at a 20 million dollar market cap and sometime in the last 24 hours a trader opened a massive six million dollar short position on jelly jelly oh yeah i just need to pause you because like when you say like as i process this like jelly jelly like sam has posted some images of what the app is supposed to do this new social format i'm like you know could work it's kind of ridiculous like i i like people taking shots at wild apps you tell me it's a 20 million dollar market cap company i'm like yeah like that's awesome like that sounds perfect for this like it's a seed stage company like go for it guys like i'm in full support if you tell me the token's trading at like 20 billion i'm like okay let's maybe let's pump the brakes this is real um yeah but it's so funny that you can have a 20 million dollar market cap startup essentially like a yc stage startup in terms of size and then someone can come and take a six million dollar short position out on you that's like it's only in crypto is this possible and it's very very weird dynamic but take me through what's happening you know they're getting a slack message hey somebody's shorting us with size yeah we're actually just a pre-seed company.
1:29:14Fortune 500 companies. Yeah, we're being shorted with size, actually. So the trader deliberately self-liquidated themselves by pumping Jelly Jelly's price on chain, essentially forcing himself out of the trade. Weird. Hyperliquid was left holding the toxic short position, currently at a massive loss to negative $10 million P &L. Jelly Jelly pumped from a$10 million market cap to$50 million in under an hour due to the forced squeeze. Hyper Hyperliquid could face full liquidation if Jelly Jelly reaches$150 million market cap, which would be a 3x away. Market is watching closely. But is that that big of a deal for Hyperliquid?
1:29:54I mean, we were talking about this earlier. It seemed like it's a$14 billion organization or coin. It seems like, yes, obviously losing $10 million or even more would not be great, but it seems sustainable. This doesn't seem like a true run on the bank situation. Yeah. And I'm looking at Jelly Jelly right now and it appears with the market cap that they are still at around 20. So liquid seems to be doing just fine. So, uh, massive drama, but not a full crisis over at Jelly Jelly. Very interesting though. Yeah. We'll have to talk to Sam about it and get, uh, before we get our next guest in here, uh, uh, we have, he's actually ready.
1:30:39He's ready. Let's bring him in. Let's bring him in. There he is. Hey, what's going on to the show? What's up guys? How are you? We're great. How are you? I like your bananas. Oh, thank you. Yeah. We were inspired. We're doing the morning routine. We got the Saratoga water right here. We're taping our mouths. We can't tape our mouths during the show, but it's been good. Uh, how are you doing? Can you give us a little, uh, Brief intro on yourself, your company, what's new? Yeah, so I'm Stephen Schwartz. I grew up learning how to program pretty early. And I basically spent all my childhood going Facebook group to Facebook group and selling different people software and met some people really early on that I'm still working with.
1:31:21And you can kind of consider us like chronically online kids. And we figured out a lot of ways to make money on the internet and And segue that into a platform called WAP where we help people make money on the internet. And a lot of shit is going on right now. I think we just released a lot of new products and someone leaked our fundraise a year ago. So I think there's been a lot of people reaching out to us and things are really busy right now. So it's getting pretty crazy. Where did the name come from and why didn't you release the funding information? Yeah, the name came from, we actually just purchased a ton of domains throughout our early years.
1:31:58And we just, every project we had, we picked a random domain off the shelf. And we'd always buy like auctions ending soon and domains are about to expire. And we had WAP.io and looked like a cool phonetically sounding name. And now it stands for we help offers pop. And I think it was super random. So that's where it came from. And we didn't announce the fundraise, I think, because like our goal is not to raise money. And I think our customers don't really give a shit about how much money we raised or anything like that. And I think it's more distracting than not. And even the people that we want to work with and whatnot, I don't think that the good people really care so much about all the buzzy stuff that you might want to talk about online.
1:32:36So I think we were trying to just stay pretty low key. It's more distracting than not. Cool. Yeah. Talk a little bit more about the team dynamics. sounds like you just worked on a ton of different projects with the founding team. Was it was it fated that you guys were all going to work together forever? Like, yeah, talk about that. So I met someone in Cameron in a Facebook group. We were like 13 years old. He made a post saying he's looking for an iOS developer to help him build some projects. And I slid into his DMs and we started building different iOS apps together in like 2013 and selling those those apps to people in the Facebook group.
1:33:10And from there, we built games. We built chat apps. We built marketplaces. We had our own agency and kind of just like so many different things. And I think during COVID, it became clear that everybody in the world, they didn't really like their job. They didn't want to sit behind a desk all day. They either got fired or they didn't really like it. They wanted to travel. They were at home and people were pretty addicted to the Internet. So it's kind of like the whole world is waking up to what we had done our whole life. And when that happened, we found a lot of forums were actually meant for selling software.
1:33:38And we started selling software in there just for fun. and we was trying to make a little bit extra cash. And Cameron found one and we started to make like a couple thousand dollars a week from it. Like, what the hell is this? People are cash apping each other, selling each other for desktop software. And it was the same desktop software we built when we were younger. So we were like, why don't we just start selling it? And it made a lot of money. And we were like, let's build a platform where we can actually sell it and structure the delivery of the software and accept payments on it. And people liked that a lot more than the forums.
1:34:07And pretty quickly people asked if they could also sell on the platform. and we added the necessary functionality for that to happen. And over the past couple of years, I think people were asking us, hey, can we also sell this? Can we also sell this? Can we also sell this? And I think we started to build a lot of that. And now people are selling all sorts of stuff. I think maybe for the first couple of years of our business, it was really like we're building a better Shopify for more digitally native businesses. And maybe you can kind of think about it like retail Stripe. And then before we knew it, we had a couple hundred thousand people coming to the site each month and buying something.
1:34:37And we were like, how can we make this a little bit better where we reduce the barrier of entry for people that are trying to sell something. So you don't actually have to program software. You don't have to have a Discord server where you have to come plug it into our website. Let's just build Discord and at least the parts that matter. And we started to build things like chat and forums and live streams and all this different ways to engage and connect with your customer base. And that's where we are now. And yeah, so I think on one end, we have like retail Stripe, retail Shopify. And now we have a lot of consumer functionality to chat and message and interact.
1:35:08Yeah. Question going back a little bit to the raise and then I have some stuff related to the product. Have you guys been profitable this entire time? You guys are the make money app. I imagine you guys like, you know, running a good business yourself. Maybe talk about that. Yeah, so when we first started, we were bootstrapped for like the first six months, then someone tried to acquire us and we said no and raised our first money from a couple people. And since then, we've definitely had a couple profitable months. I think usually before each fundraise, we had a lot of months that were cash flow positive and to take it up a notch when we raised a little bit more money.
1:35:41And I think after we do that, we obviously ramp up spend and start investing in some pretty cool areas. So we've had profitable months right now. We're definitely not cash flow positive, but I think we could be if we wanted to stop investing in the places we're growing. And I mean, our model is lent itself pretty well to making money. And I mean, we only make money when people make money on the platform and people are making a lot of money on the platform. So we're certainly making a good amount of revenue. And we're also investing a lot of that back into the platform. Talk about hype generally.
1:36:14You don't care to announce fundraise. gets hype in the in the venture world but at the rate that you're growing i imagine at some point you know i can imagine wap as a public company and you know you are going up against giants right shot you know shopify you are i'm sure directly competitive in in many ways uh and i'm sure you're going to enter other areas will there come a time when uh you want to you know you know do the tech podcast circuit you know you're on our show which is which is awesome but it seems like you've been very intentional about. I remember hearing from Carried No Interest about WAP.
1:36:50Last year at some point, he's like, WAP is the most underhyped, just absolute monster. Nobody's paying attention to them. They're absolutely crushing it. Do you ever see a moment where you'll flip the switch? Or are you guys just going to stay focused on your customers forever? I think we're definitely going to stay focused on our customers forever. I think we do the right marketing to the right customers. And I think that it's no shock that the people in tech don't necessarily know about WAP so much because they're not really the ones who are starting many businesses, right? They're trying to go for the super scaled up businesses.
1:37:23And I think we have generate millions of views probably per day at this point across social media for WAP. And that's mostly targeting our core customer base, which is super scrappy young people that want to do something that's entrepreneurial and start a business. So I think we're definitely allowed where we need to be. And I mean, we have a YouTube channel now that generates several hundred thousand views a video and we're definitely allowed where we need to be. But I don't think that I mean, recently we released a new product that's like getting a lot of attention that people are in tech are also finding a lot of value on.
1:37:54And I think that's going to become more relevant to go on. What's that product? It's called content rewards. It's basically kind of think about it like a probably a five to 10x cheaper CPM than any other platform that exists right now. And um all ugc doesn't really feel like an ad it's not exactly an ad and um people are getting a lot of views from that and and um we certainly are experiencing a huge lift from that that's like productizing the sort of clipping which has been happening right you know remember i think people first realized that like andrew tate was doing it because you know you you you'd see videos of him that just were made by him but then also you know shared by completely other people that were funneling into his program monetizing.
1:38:34And is that, is that the, you know, we've talked about kind of the different eras of making money online. It was like selling Pokemon cards or trading sneakers and is like clipping. Is that like, if you're in high school today and you have no skills, but you just want to hustle, is that like how you see that as like, that's the sort of it, uh, entrepreneurial activity? Yeah. So I would look at it as almost like a big funnel, right? So maybe you start with that and I think you graduate towards maybe operating your own business one day, even on the platform. So I mean, what's happening right now is we're having, it's the fastest growing way people can earn on WAPs right now.
1:39:08And we already have thousands of people that are earning money, some actually a pretty significant amount. I think we've paid out some people even for our own marketing over$10 ,000 total. And those are definitely people in high school that are really good at just making videos and they know the trends on social media. And I think the future isn't really about like everybody kind of knows an ad when they see it. And it's not really as effective as it used to be to just run an ad on meta and expect people to resonate well with it. So I definitely think it's the future. And I definitely think that we'll see a lot of the people that are clipping and making EGC content right now, go on to build really awesome businesses because they won, they have a little bit of a taste of making money on the internet.
1:39:44They have a KYC payments account all set up. They know how to market very effectively. They know how to, what the content rewards are. They know what they're marketing and they're pretty in tune with everyone that's actually paying them to market. So they know it works. And absolutely, I think that's a, it's a huge part of the future. And we look at it as maybe like the first step in earning your first dollar online. Can you talk a little bit about the differences between Gen Z and millennial entrepreneurs these days? Like what's changing? What's better? What's worse? Yeah, I think I would say Gen Z's prioritize fun.
1:40:14And it sounds kind of silly, but I think like nobody really wants to work on something if they're not having fun. And I think that even if you give somebody a lot of money, it's difficult to actually get up out of bed every day and start to do the work. And I think that we're seeing that all over the place where people are just fine making$3 ,000 and traveling throughout Asia. And that's like a very, very different mindset where it's like, I need to go to college. I need to make$200 ,000 a year. Otherwise, I didn't do it. And I think people in Gen Z are kind of just doing it. And they're having a lot of fun with that.
1:40:44And I mean, we look at WAP as like, WAP is, we are building the future of work. And I think that the future of work is very fun. It should be engaging. People are making a lot of money by making funny memes. And like, that's a, that's a complete shift of like value creation. And I think the Gen Z is all over that and, and, um, uh, they, they must bring fun into work. And that is, I think the key difference of what, uh, millennials versus Gen Z. Talk about live streaming. You guys support live streaming. Uh, it's been something that's been big in Asia. It's probably big here. I'm not, I haven't purchased a product through a live stream myself, so it doesn't feel big yet, but is it, is it already?
1:41:23you feel like it's already mainstream like you're you're obviously bullish enough on it to like integrate it into your platform uh i'd love your take on it yeah i mean i think the saying is like you can say a picture is worth a thousand words and then maybe a video is worth a thousand pictures and i think that um a live stream is like the most effective way to communicate with an audience at scale and um it allows for people to repurpose the clips afterwards and it takes very um little effort to to go live to thousands of people and maybe tens of thousands or hundreds of thousands So I think live streaming is definitely the future of how you connect with an audience at scale.
1:41:56And it's probably the future of shopping as well. And I don't think it's necessarily that different. I mean, you look at QVC, and that's obviously a long time ago. You look at HQ Trivia even back in the day. And now you look at TikTok Shop, and even Amazon's integrated a lot of live shopping. And nobody cares more about something than what's happening right now. And that's what live streaming is all about. So it's very, very, very bullish on live streaming, not only for actually the point of sale, but also what happens after the sale. If you're running an e-commerce brand, I'm sure you'd want to have a weekly seminar where you actually go over with your customer base and say, hey, do you guys like the product you just purchased?
1:42:29Why don't you drop the comments right here? Maybe we can do a demo live and show everybody how to use the product. So I think live streaming has a lot of good use cases, not just for the point of sale, but even after to engage a customer base. Is there a cohort on WAP that you think will be particularly resistant to the relentless march of AI? There's a lot of people worried about, oh, ChatGPT, the next version is going to put my business out of business. what do you see as being like, yeah, this person on WAP is going to be here for two decades or more? But I think the people on WAP are really special because what happens is as the trends evolve on social media, the people don't just stop doing what they're doing.
1:43:04They don't just fall off the face of the earth. And I think that they adapt really well. And I think a lot of people on WAP start one business and then second business and third business, and then all the way still 10 businesses later, they're finding their niche in the market. And just because the niche change doesn't mean they're going to stop trying to provide value for people on the internet. So I look at AI as actually a pretty awesome driver of our business. And it makes it, unless we're going to hypothesize that everybody just stops living and doing anything with AI, which I think is certainly maybe some people's perspective, but people are still going to exist and they're still going to be here.
1:43:34And it's not like they're going anywhere. So when that happens, it's like, I need to figure out how to make money for myself and maybe how to have a purpose. And I think that WAP is all about that. So I actually would say WAP as an entire platform is very, lends itself very well to AI, or maybe AI lends itself well to WAP. And I look at, we have a lot of AI products on the site and people talking about AI, doing AI courses, AI software, AI in their live stream. We have some video features that you can send people, avatar messages and creating ads with AI. So I look at AI as a huge plus for almost every single person on WAP.
1:44:08That's awesome. Makes sense. Last question. How do you think about hiring? I'm basically an army of like 100 ,000 people that are sort of your users that also want to support WAP because if they can make WAP better and bigger, it's good for their individual businesses. But yeah, what does your team look like? Yeah, our team is actually mostly like users of our platform. And I mean, people are some of our earliest users. We were our first users and some of our earliest team was our first users. And I think that the best place to recruit from companies broadly is actually a user base. They know the product well, they're building for themselves.
1:44:42And I think that we have a saying, it's build, you're your own customer. And we don't use anything else other than WAP. We don't use Slack, we don't use Discord, everything's on WAP. And I think that's the best way to hire also. We have a great community of hundreds of thousands of users that use the product every single day. And I mean, they're giving us the best feedback. And oftentimes they're programming for us and building little features in the site and doing our marketing as well. So the way we hire is people that are in our space, on our product, and that really are passionate about our mission.
1:45:09uh i can uh random thought i can imagine you know uh a boomer vc talking to you saying you know what i really think what you're doing is the future of work and you're like bro this is the present of work we're living we're living it uh but but it's funny how these sort of platforms emerge where i look at wop and i hear you talk about it it's like yeah i believe this is the future of work in many ways it's people being entrepreneurial like finding you know new ways to uh you know like you said, create value on the internet. And it's just very cool to see. So I love it. This is such an awesome congratulations on the progress.
1:45:44I enjoyed that. Hopefully we don't create too much hype for you. I think another, another year, but, but your days are numbered, dude. You're, you got maybe another, you got, you got basically another six months, uh, you know, relatively low key. And then people are going to be like, wait, these kids are doing like, you know, hundreds of millions of revenue. Uh, but, uh, it's great to have you on. We'd love to have you on again. yeah thanks so much for coming on this was fantastic well awesome guys have a great day bye-bye thank you bye yeah that's great uh well yeah what a fascinating company dude that's completely under discussed that's the final boss of of gen z entrepreneurs uh i would uh i would love the charger uh oh yeah sure we share one charger hey you don't become the most profitable podcast while spending money on duplicate technology.
1:46:31But we got Underdog coming in. We got Jeremy from underdogfantasy.com. He's got a beautiful background. How you doing? What's up? You're ready for the full TV. Love the setup. Looking great. We got the background. I love it. There you go. Can you give us just a quick intro? Who you are, what your company is, what's new? Yeah, for sure. For sure. Company is underdog. We just turned five years old. We focus on building games for sports fans in America. I've been doing something much the same my whole adult life for 15 years now. Started a business called Star Street in 2009. Actually started as a sports stock market, but it was very early real money fantasy sports days.
1:47:16And I've been trying to make sports more fun for 15 years now. That's awesome. Talk about the news today you got a new yeah we just uh we just announced our this the first close in our series c um spark capitals leading 1.225 pre-money um valuation there we go there we go i got this new uh i got this new soundboard and uh some nice validation and acceleration for what we're doing size gone sorry i got this very excited we got a soundboard we're excited about it today i don't know if you can actually hear it but uh there's some great great sound effects i don't hear it i wanted to get it no so so my first thought goes one uh you know super validating get this round done but i remember around the time when you started this company it was every round i don't know about every but at least the first one or two rounds that you're raising were like very hot like people were you know chasing it they got priced almost to perfection right i i don't even remember i remember it was like my recollection it was like a almost like a seed precede getting price like well well above 50 i don't know exactly the price but a lot of people when they see rounds getting done like that that are just like super early and like expensive end up saying like well this company has to execute perfectly and you know clearly you know it's hard to execute perfectly but you executed very well um how do you uh you know talk about kind of that journey in terms of like making sure that if you want to go from whatever, you know, wherever is initially priced to 1.2 in five years, you kind of have to like hit these sort of perfect milestones all along the way.
1:48:56Yeah, definitely. It's definitely not been perfect. It never is, of course, but it's gone fast. Look, I think we had the luxury of the founding team at Underdog was seven. We all worked together at a prior company, Draft. And we built products that people really loved. So we kind of had, in a way, the customer base, the fan base when we started the business. And we built games that ultimately through some M &A got taken away from them. So that was the kind of start. And that allowed us to move fast. I've spent my whole adult life in this community. So we had the community and supported the community early.
1:49:32And that's something that's been really important to us is the people we build for, the customers we build for, making sure that they love our products, they love our games, and they've supported us the whole way through. So that really helped. Those seed rounds, the first ones weren't as glamorous. I think our first valuation, the first money we raised was at a$6 million valuation. Oh, wow. So not that glamorous, but it went first. So I was not early. We did the first, then a second one at a higher valuation, then a third, and then a year later, I think was our Series A. The partner who led our Series A was my very first investor in my first company 15 years ago.
1:50:06He then was a partner at SBA Angel, Kevin Carter is his name. He runs the capital now. He was the first person I called when I was having to raise a big round. He had been in every seed round, obviously, and he stretched kind of well outside his fund to lead our round because that's not what his fund was meant to do. But he's kind of seen this for 15 years and knew the journey. So he's been a great backer and a great asset to the company. Can you tell me more about the, I mean, you mentioned it's like very community led. What's the marketing mix right now. I see a lot of the big sports companies partnering with Pat McAfee or ESPN or the Dave Portnoy universe.
1:50:40Where have you found success? Yeah, it's a real mix. Look, the top acquisition source for us always has been our existing customers. Almost a third of our customers come from directly attributed referrals from other customers. It's always going to be the best source, right? That's a sign of us building products that people love and that's our whole opportunity. Yep. Um, walk who is just on, we have partners on walk. So that's great. They're a big partner of ours. Um, and, and partners in general, we, we do a lot of our own content as well. Um, two years ago, we, um, we made the decision to start a show and we built a set in Gilbert arena's basement.
1:51:16That show is now Gil's arena. Um, and that show is, we think by, by kind of any forms of measurement, the fifth largest daily sports show in America right now. Wow. Wow. Massive success for us. And that crew, Gil and the whole crew have been amazing to work with and amazing partners to the business. We have a couple other shows as well, and just continue to invest in trying to meet sports fans where they are and add value to their lives. That's awesome. Talk about building in a hyper competitive space. I think you don't even have to be aware, you don't have to know anything about, you know, sort of like anything sports gaming related to know that you're competing with some heavily, heavily funded, you know, sort of public companies that can outspend you on kind of every dimension.
1:52:01But clearly, you guys have been able to like, break through and become a real player, despite, you know, being able to invest far less. How have you guys, is it like raw execution and just building the product that you know, you guys want, so you know, your customers are going to want? Or what do you think has been your guys's kind of winning formula to date? Yeah, look, we started Underdog with kind of a really simple thesis. It was, there's so much more to be built for sports fans in America. And if we can be the best at building product, we'll build the biggest company in this space, right?
1:52:31That's always been the belief five years in, we have obviously more confidence in that than ever before. And look, we can be the best at building product and can, with a lot of confidence, say in the space, we are the best at building product, not just because it's our DNA and what we prioritize and what we look for in people and what we really care about. But also just by the kind of function of how this space came, almost every other large player is largely reliant on legacy tech that was initially built for a different market, for different customers, for the UK and Europe. And we had the opportunity when we started this business to really play a long game and say, hey, we're going to build this for the future.
1:53:07We're going to build to build the best product. So we should build all our own technology. And that's what we've done, both fantasy sports, now licensed sports betting technology, And that gives us the ability to build a differentiated product, to build a better product, and to just have the fastest product velocity. I think the whole company is sick of hearing me say it because every two weeks when we do our big company meetings, the two things I make sure I say every two weeks are it's product and it's people. Those are two reasons we're going to win, and we have to have the fastest product velocity and the best people here.
1:53:38And that's what we focus on. Can you talk a little bit about regulation? I know the products are legal in, Wikipedia says 31 states. I don't know if that's exactly accurate, but obviously there's a march from the whole industry. And even though you're competing, you're probably working together in some ways. Is there a lobbying group that you're part of? Like mechanically, if you're part of an industry and you got to kind of team up with your rivals, what does that look like day to day? Yeah. So there's a lot of different, there's kind of a lot of different pieces to it because there's, we offer games right now in 41 states.
1:54:14Okay. We offer our core game, our pick 'em game in 36 states. Where licensed sports betting in one state, when Missouri launches later this year, December 1st, it will be a second state. Or so we hope we obviously need to get licensed by the regulator there. And for us, it's all about, we want to always build games within the laws and build the best customer experience we can within the laws and kind of the wants of the regulators. um this space is definitely highly political uh there's a lot of competition and a lot of people who can't really don't have the means to compete or the ability to compete on products so they try to compete in other ways um that's definitely something we faced and when we do that we always try to team up with the people who care about customer experience care about delivering the best products to customers and say hey that that's how people should compete that's how people should win it should be all about customer experience how do you think about focus specifically with the team i'm sure that everybody or not everybody i'm sure a lot of people that join underdog do it because they like they're just obsessed with sports and they love uh the products that you guys make but at the same time like you know if you've got a lot riding on a certain outcome it can be distracting from actually building the underlying product and i'm sure for you and and other people on the team there's also a lot of you know distractions around some athletes like hey you want four seats to this game you know like they're a fan like how have you guys stayed focused because clearly clearly to accomplish what you have in the last five years like you've had to be pulling some you know very late nights and and really leaving it all um on the court on the court there we go landed it um it's funny we're we're maybe signing at least for a new office and we have this vision of putting a basketball court in there so it will really be on the court then there we go look we're really lucky we get to work in sports um sports is fun sports brings people together i've been a sports fan my whole life that's obviously kind of how i got into this um and i just think there's so many wonderful virtues of sports and then the community the communities it creates and the bond it creates um i just think that that's a blessing for us look we it's funny over the course of my career people and not as much anymore but people used to say oh we don't invest in hit in games businesses because it's history of businesses and it's tough have to predict hits.
1:56:32Like the sports are the hits and the hits keep coming. We don't have to worry about that. Yeah. Can you take me on a little bit of a historic, like historical tour of the industry? Like when I was a kid, it was like gambling. It only happens in Las Vegas. Then in college, people were playing online poker and then there was like a poker apocalypse or something where all those sites got checked out. Can you give me some just historical anecdotes? What were the key turning points in the industry? Yeah. Let me start post Black Friday. I believe that was 2006 okay and that's when poker really kind of went away online in america it's now come back but just in a few states and nowhere near to kind of what it was then got it um real money fantasy sports as we know it really started in 2009 um fanduel that's actually when star street was born as well um fanduel kind of being the largest of that early generation obviously and have continued to be the largest.
1:57:27There were a lot of incumbents who weren't so pleased to kind of see their success. And in 2015, there was kind of this regulatory storm that hit the fantasy sports industry. For a period, 15 attorney generals said the business was illegal. They were under an FBI investigation. They were under a DOJ investigation, only like a six-month stretch. The industry then got together and said, hey, we need to clarify the laws. We believe what we're doing is legal. but we need to clarify the laws and got together and passed legislation. There's now fantasy sports legislation in, I believe, 22 states. So that happened then.
1:58:02Then there's always kind of been fits and starts of, hey, maybe legalized sports betting is going to come to America. But it never really did until 2017, a court case called PASPA. Went to the Supreme Court. It was being kind of pushed by New Jersey who wanted to offer sports betting. Went to the Supreme Court. in 2018 when PASPA was overturned by the Supreme Court that made sports betting no longer federally illegal in the U.S. And from there, states can legalize it as they so choose. Somewhere around 28 states, I believe, have now legalized online sports betting or sports betting in some format.
1:58:39And so the companies that were best positioned for that early set were FanDuel and DraftKings because of the fantasy sports advantages, the customers they had, the ability to operate a digital business. And that's kind of wave on a sports betting in America. Now, when PASPA was repealed, there were 31 days until New Jersey was launching sports betting. And then a quick wave of six states came soon after. So every company that wanted to be live for the market open had to take tech that was already built and operational, that was built in the UK and Europe, where there was a way more mature market and bring it to America.
1:59:14And that's why if you look around the space today, you kind of see products that look almost all the same, right? There's different colors, there's sometimes different prices, but the core experience is the same. We believe, I believe that's not the right product for American customers. It's a very hardcore product. It's very transactional. It's not that much fun. We believe that the games we build and the games that should be built are to make sports more fun for someone who's watching a game and wants to increase their enjoyment with the game or who wants a reason to watch the game or something to do with their friends.
1:59:47And so that's specifically what we build for. An analogy I use quite often, and I've never been a fan of using a company's analogy for another, but I think it just fits so well for what we're about and what we believe the opportunity is. Let's think about what Robinhood is to stop brokerage. That's kind of what we are to sports book, right? Simpler, more intuitive, more approachable, more fun, less transactional, less about big dollars and thinking about it from financial perspective more put ten dollars on on an experience tonight on your opinions express your opinions on sports and have fun with that cool very very cool uh congratulations massive milestone it's great having you on congratulations looking forward to uh looking forward to the next markup i'm sure thank you we're working on it yeah awesome dude good luck thanks for coming on see you bye that's great we got rob coming in soon another size gone hits uh who led jordie well how did jordie wind up with the soundboard this show i feel like i would be a little bit more responsible with it but oh you can trust me with the sound yeah i'm just hitting it 25 times with incredible uh sound effects so uh we got rob mower joining right now hey rob how you doing what's up jams how are you guys doing you're doing great doing great it's great to have you uh why don't you you introduce yourself uh before i do and do a silly one more uh Yeah, definitely.
2:01:14So I'm not sure if you guys are getting any introduction here, but I created the Huberman Lab podcast with Andrew. I now run the podcast along with the media company that Andrew and I created called SciComm Media. The long and the short of it is Andrew, as I'm sure many are familiar neuroscientists at Stanford, natural born teacher who has a real gift for explaining how to better your physical health, mental health and performance. So my job with Huberman Lab is basically to get out of his way and just make sure that he has the tools he needs. And I'm running the business to make sure that we're putting out, you know, really a high quality episode every Monday.
2:01:53And then also now on Thursdays with our Huberman Lab Essentials, which are our 30 minute episodes that go out on Thursdays. Was there a key turning point in the business where it was taking off or was it gradual? Actually, start and maybe go a little bit earlier than that. I've heard it. I've heard it, but I'd love for the audience to talk about the story of just how it all kind of started. Yeah, let's start there. That's great. Okay, yeah, definitely. So let me bring, I guess, best starting place in 2019, I met Andrew. We had lunch together because I was helping interesting people get booked on podcasts.
2:02:34After about 10 minutes of talking to Andrew, it was clear to me that he was one of the most knowledgeable people I've ever met on topics that were of interest to me. He, you know, like whether it was how to manage stress, how to optimize sleep, he had studied David Goggins in his lab researching fear. He had done MDMA tests on mice studying the like thermogenic effect of that compound on them. So basically, I was like this, this, this guy knows more about things that are of interest to me interest to my friends, and probably of interest to folks like, you know, Rogan, Tim Ferriss, and other podcasters.
2:03:13And my goal from then became, okay, how do I help in terms of brokering some of these introductions? And just selfishly wanting to hear him on some of these shows, chopping it up with folks like that. In 2020, when COVID broke out, Stanford basically encouraged Andrew to get out and start talking to media. And that's when Andrew and I reconnected. just given the fact that his lab and a lot of his research was around fear, circadian biology, things that people were struggling with during the early days of COVID, being kind of like locked down indoors, a lot of stress going around. So it just kind of seemed like a natural fit to go on a lot of these podcasts.
2:03:55And after doing a bunch of them and, you know, getting just really positive response from audiences, Lex Friedman actually encouraged Andrew to start a podcast. And then, uh, Andrew and I talked about it, you know, end of 2020 launched the show in 2021 and it was kind of off to the races from there. Uh, did you guys have instant product market fit? It's sort of historically extremely difficult to grow podcasts, right? Like people don't really realize they see you guys at the top of the charts. They just imagine you get a couple of viral clips. All of a sudden you have a big show. It's very difficult to grow.
2:04:31I'm sure you guys had the benefit of like getting Andrew on these shows and, and maybe having those other podcasts promoted or whatever, but how, how quickly was, was, you know, was it a rocket ship or, or maybe slow and steady, steadier than people might think? Yeah, it was definitely a bit of a rocket ship. I think, you know, a lot of times, um, people will look at the success of the show and kind of see it as this, you know, uh, overnight success in some ways. But I think what, uh, is important to remember is that at this point, Andrew had, had 30 years of researching biology and speaking to it from a deep love in terms of understanding the mechanism of how these different systems in the body work and how to optimize health based on those, and had at least a decade of teaching experience, doing lectures and things of that nature, and then just putting the camera in front of him and his ability to speak to these topics that are, you know, rooted in science, but kind of coming at them from the angle of, you know, how people can better their health in both the short term and long term through using them, I think is where the kind of like product market fit was where, you know, whether it be on YouTube or on the pod, you know, RSS podcast platforms, people were interested in these topics and Andrew could speak to them from a place of, you know, deep experience and understanding.
2:05:57And yeah, just kind of thinking it through in terms of what topics are going to be of interest to folks, but really just kind of perfute her kind of following the path of where his interest lied as well was was key. Give yourself some credit too, because you had a like a decade, like, you know, basically, you know, managing, you know, different various brands and whatever. How do you guys think of just sort of like distraction and avoiding distraction and focus. You guys spent a very long time just doing the show. Sort of simple. You've branched out more recently, but it still feels very focused.
2:06:36And then at the same time, I'm sure if you guys said yes to all the great opportunities in front of you, you would never even be able to be in the studio. So how do you think about what's been your framework for deciding what to do, when to do it, when to say now that kind of thing yeah i would say our default answer is no um basically we look at it through the lens of uh everything is a uh is a if we look at it through the lens of our goal is to put out the best possible episode every monday morning um everything that is competing for our time against that really needs to be fantastic for it to make sense in terms of us pursuing.
2:07:24Andrew wants to spend all of his time prepping for either guest interviews or solo episodes on the podcast. So basically, we see everything as competing with that. So our goal is always to weigh opportunities to not just foolishly pass things up, but really to think about it from the context of this is pulling us away from what our main goal is, which is those Monday morning episodes. And if we have the bandwidth to take on something that's not going to interfere with that, and maybe even will be additive to that, then fantastic. Otherwise, it's a no. And I think, you know, you hear a lot about this from folks like, you know, Steve Jobs and other leaders where it's, you know, the challenge is not saying no to things that are easy to say no to.
2:08:16It's saying no to things that seem like fantastic opportunities and ultimately are going to pull away from, you know, what our bread and butter is, which is just, you know, creating that best episode every Monday morning. Can you talk a little bit about the evolution of the product with the introduction of the essentials show and then distribution and how you think about repurposing the content and everything that happens downstream from that one really canonical piece of, you know, Monday show. Yeah, this is something I think a lot about. And it's something that our team is very focused on. We view it as like, we put out these, you know, three hour plus episodes, and then all the way down to a tweet that Andrew does.
2:08:59And it's like, how do we fill in the various kind of like subsequent like time domain elements that uh span that spectrum so it's like uh everything from you know full-length episode to essentials episode which is 30 minutes to clips on youtube or x or wherever we're kind of you know putting those out on social media um to a newsletter and basically you know all the way down again to you know a tweet and the goal is always to get the information out to as many people as possible and help people, regardless of how much time they have to allocate towards consuming the information, get the information.
2:09:41So, you know, there's obviously the like monetization strategy that kind of spans across those two, because at the end of the day, we are a business. But the real goal is like, okay, how do we just get this information out there to as many people as possible? And the Essentials episodes were a genesis of that, where we're looking at our content saying, you know, a lot of people just do not have time for a three hour episode. We were hearing from a lot of folks, this is actually pretty funny, that they'd love an episode, but they were like, if I'm sharing this with someone and it's a three hour episode, it's almost like I'm giving them like a to do or like an obligation where we were hearing, especially from like a lot of, you know, CEOs or business leaders.
2:10:24They're like, listen, you know, I loved this episode, but I'm not going to share this with my friend because it's basically like, hey, I'm going to suck three hours of your time away or basically be requesting this of you. So the Essentials episodes were kind of like birthed out of that concept of like, how do we make our content more shareable? How do we make it more digestible for those who have limited time? What do you think the future of RSS is, the actual underlying technology? As we've gotten you know we're a new show but working with apple podcasts and spotify and all this stuff it's just genuinely hilarious and feels wrong how little it seems like apple cares about just podcasting despite it being such an important media medium do you think it all kind of like shifts more towards streaming or or you know what what's your take broadly on on rss is it sort of the cable technology and and maybe there's a future or we stuck with it forever oh man i have so many thoughts on this especially the apple piece um you know it's funny because like i i got in a podcast in the very very early days where it was like you download on itunes and you kick it over to an ipod to listen to it and i just became obsessed with podcasts again this is like probably 2010 time frame um and obviously at that point it was like apple was the only player And then, you know, they launched their podcast app.
2:11:52And slowly over time, other players have entered into the space. And Apple's kind of just like been caught flat footed, maybe. I think probably some of it has to do with the fact that it was a part of their business that was, you know, never earning money. So if you start with a zero revenue stream, then, you know, anything is a benefit. Whereas a company like Spotify or YouTube, they are very focused on P &L and looking to do things to dominate the space. To get to your question directly, though, it's interesting. I think this is how you guys view your show and how we view our show. Even calling it a podcast is weird because a podcast is just the medium through which it's being presented to an audience.
2:12:42and we see ourselves as content creators and we're always going to put ourselves on the platforms that make the most sense, whether that be RSS feed and people are getting it there. And I do think that there are a lot of people just who listen to podcasts while they're doing various activities and whatnot and are just listening to them. That having been said, I think there's a huge percentage of most shows now that offer video where, you know, I would say half of our audience is either watching our content on YouTube or on X. And, you know, another half is is is listening. It's the the benefit of the RSS feed is it's far more consistent.
2:13:34So we'll have episodes on YouTube that get like 15 million views, but then we'll have other episodes that are kind of in that 300 ,000 to 400 ,000 range. Whereas on audio, it's a much tighter window, so it's like a kind of more consistent basis. That having been said, RSS feed is set up in this weird kind of like – it prioritizes recency over relevancy. So it's basically like, here's the most recent thing as opposed to serving it up in a meaningful way like YouTube does. YouTube is really good at understanding the algorithm and serving it up in terms of, you know, what's most relevant as opposed to just what's most recent.
2:14:14And I think RSS feeds need to figure that out. I think Spotify is working hard on that. But, you know, we haven't seen much movement from Apple on that front. I put a tweet out recently on the fact that I believe X is going to be a dominant player in the podcast space in the future just by virtue of the fact that it's inherently social. Like if you think about it, we'll put out an episode on Spotify and on Apple and on YouTube, and we in large part are announcing it to our audience on X. And we do that as a means to broadcast it. And then people oftentimes, we also put our episodes on X, but oftentimes people are then leaving the platform of X to then, you know, pay attention to the podcast elsewhere.
2:14:58And I think as X gets better and better at allowing users to easily navigate content and watch full shows on their platform, there's no reason why people should be then migrating off platform to pay attention to the content. So I think that that's a major thing that Spotify will have to contend with. I think that YouTube has a pretty dominant place in the market with just their monthly active users. I want to say it's like 2.5 billion or something was the most recent number on that. Whereas X and Spotify are kind of in the like 600 million range over the course of a month. So a lot of ground to make up.
2:15:40But I think it's a lot harder to build a social platform than it is to build a content hosting platform. So like I think X would find an easier job of creating the ability for people to easily consume content on their platform as opposed to something like Spotify trying to build a social network as part of their app. Yeah, that makes sense. Last question for now. Quick 30 seconds of advice on how you guys have handled AI impersonation. A lot of people have an incentive to generate an image of a video of Andrew saying, I love this product. I take it every day. It's like it helps with my tea. And then it's just completely fake.
2:16:25Any words of advice for people that might be dealing with impersonation or just sort of fake AI content as well? Yeah, I think being vocal about it. So like we've made sure that Andrew's been pretty vocal about it. We had this thing where it was the Jawserciser, which was a complete nightmare on our end. We basically filed legal action against him numerous times. They kept like shut. They'd basically like file bankruptcy and spin up another company and kind of like all this craziness. A lot of people to this day still believe that Andrew was promoting this plastic ball that you put in your mouth and chew on where basically a large part of that was AI generated or clipped out of context from a podcast.
2:17:07No affiliation with the company. So we've basically been vocal about like, listen, if it's not being put out on a Huberman Lab handle or it's not listed on our HubermanLab.com slash sponsors, it is not a company that we have any association with. We also very intentionally have very few sponsors. Most of them you're going to be familiar with. They're the sponsors that are basically always featured in our episodes over the course of a month. And we're highly, highly selective with who we allow to be a sponsor of the show. and you know if you ever question it just go to hubermanlab.com slash sponsors that's where the sponsors we work with are and you know if you're seeing andrew say something ridiculous and it's being put out by you know some bizarre company then it's probably ai or it's probably just like you know uh faked in some manner believe half of what you see and none of what you hear that's right words of wisdom well thank you so much for coming on the show and we got uh last thing we got uh andrew himself in the chat on x he says rob is the secret weapon behind human lab and no one can steal him he's a loyal bulldog and he's also super generous with knowledge to help others and he's a great triathlete we didn't get to cover your athletic career but uh would love to have you on again uh you know anytime there's so much more to talk about i mean i know i'm the same business i'll see you back in malibu you're the man thanks for coming on yeah thanks andrew Andrew's the best.
2:18:32We'll talk to you guys soon. We'll talk to you soon. Bye. That's great. We got another Andrew coming on. We do. Andrew Reid from Sequoia Capital. Creator of Reid's Law. Are you familiar with this? Yeah, yeah, yeah. We got to try to find that post. Oh, yeah, we got to find that post. It was too good. Rune was saying, sometimes the person who popularizes a phrase accrues more value than the person who creates the idea. and Andrew replied and said, I call this Reed's Law. And here is none other than the creator of Reed's Law himself. Andrew, welcome to the show. Thank you for having me. Thanks for being here.
2:19:10How are you doing? I'm good. Yeah, that's going to be my tweet that's going to live in infamy. Hopefully. There's been various attempts from people that get that as on Wikipedia page. But I think there's an actual Reed's Law that keeps canceling it out. All right, well, we're going to work on it. That's our new mission. Yeah, get it on Wikipedia. We got to make it happen. Very blue today. I love it. You're really owning it end to end. You're verticalizing blue. It's great. Great intro. But great to have you on. John, you want to kick it off? Yeah. I mean, we wanted to have you on to talk about this post from Harry Stebbings.
2:19:49He says - Oh, I'm hoping that's why you asked. I was hoping that's why you asked me to come on the show. Yeah. He says, through my role at 20VC, I studied investors for a living. Andrew Reid must be the best investor of the last five to seven years. Eleven Labs, Vanta, Figma, Odo, Bold, Klarna, Zapier, or Zapier, Circa, average ownership of say 5%. Andrew will have made 3.75 billion for Sequoia. And he ends it with an emoji that has star eyes. So we wanted to go through those companies. Tell us how'd you meet them? What'd you like? What is the business doing well? And kind of go through what your process is like.
2:20:24So maybe we could kick it off with Eleven Labs? Well, sure. I guess the timing of that tweet was particularly good because it was, you know, I think on Wednesday of last week was when the Wiz deal was announced. Yep. And, you know, yet another$3 billion gain for Doug Leone. And, you know, of course, you know, everyone at the time is saying, you know, that guy Andrew is just so amazing. Oh, yeah, you know. So I spent like a few days in the typical Sequoia wallowing of self-pity of like, how is it possible for Doug to have yet another one? And I'll never live up to people who came before me. And then Harry's like, you know who's been amazing is Andrew.
2:21:06That's great. I love it. I was like, yeah, everyone's been saying that this week. You know, that's been the real talk track at Sequoia. That's awesome. Do you want to talk about it? Before we dive into all the individual companies, how did you end up at Sequoia? You started as an analyst at Goldman. And then how did you transition into venture? Was it something you always wanted to do? And yeah, just break that down for everybody. Sure. Let's see. So I was 2012 from school. I started my career at Goldman Sachs where they have the blue background profile pictures as the originators of that. I spent 18 months at Goldman and left to join Sequoia.
2:21:52I had one of the all-time amazing happenstance introductions. It was Sarah Guo, who's now at Conviction. Back then, she was in my class at Goldman. And she was then dating and now married to a guy named Pat Grady. And Pat's not my partner here at Sequoia. And Pat, have either of you ever met Pat before? I haven't. So Pat is like one of the all-time great technology investors, massively underrated. And he is a very interesting guy. He's one of these guys who has – the way I like to describe Pat is his brain is so structured and framework-oriented that it ends up like when ideas enter his head and ping-pong around for a while, at the other end of his system spits out very creative ideas, despite having not one...
2:22:49He's not a creative-minded guy, but he ends up with very creative ideas, which is why he's a great investor. Anyway, his idea at the time was to have a class of two associates at Sequoia, one guy who was maxed out finance and one guy or girl who was maxed out startups. I was the finance guy. I was, you know, at the top of my class at Goldman. And the startups guy was Matt Huang, who now runs Paradigm, who had previously started and sold a company to Twitter and did a bunch of angel investing. So they put me and Matt on a desk together. So we were the class of two associates February 2014 at Sequoia.
2:23:32And it was, you know, it was a very interesting time to be entering the VC world. I think for those of us who are fortunate enough to enter at that time you sort of had a few years of earning your stripes and learning and by the time you were able to start doing your own investments you know you still had many years of this like big cloud wave ahead of you and you caught the tail end of mobile so you were able to get a few really good investments under your belt early on and then you know since then it was kind of off to the races but um and Sequoia is a great place obviously to start and invest in career because you look around and you know it's like legend and there's Doug and Mike and Jim Getz and Roloff and Alfred and Pat and Carl you know it's you kind of like all the first name basis investors are pretty cool to be around you know so it's great yeah there's uh there's this meme that like founders want to raised from other founders.
2:24:29I think the reality is like founders want to work with very successful people that can help make their business more successful. Uh, in the early days, did you have any sort of, uh, self doubt around sort of like coming into this industry? Like, you know, a lot of investing is just like, you know, being a clear thinker and like, not, you know, uh, just believing all the hype and like, you know, uh, likability is a big factor. Like there's a bunch of different factors, but did you have any self-doubt in the early days prior to this absolutely generational run? Hopefully, you don't have too much self-doubt anymore.
2:25:04I've totally stomped out all the self-doubt now. I wake up every day just feeling so good about everything I'm doing. You should try Hubris. Hubris is really popular these days. It's fantastic. After investors that try Hubris after a generational run. It always works out. Yeah. Ask Masa. Do you have a crystal ball yet? We have a crystal ball here. There you go. Yeah, you need a crystal ball. Yeah, yeah. Yeah, this is great. That's perfect. It's funny you mentioned the self-doubt thing and kind of getting started. I actually think, you know, I was – I think I was the first person at Sequoia to actually tweet.
2:25:44You know, it was like this was my – and the origins, this is probably 2017. I remember thinking, I started in 2014 and I was doing well, I'd sourced some investments, I'd helped out on things. GitHub we did in 2015, the company was acquired in 2017. So I felt like I'd kind of gotten my feet underneath me, but then I imagined, imagine you are a founder and you are invited into Sequoia and you have your pick of Mike, Doug, like, Roloff, Jim, you know, Pat, Alfred, you know, why would anyone ever choose to work with me? You know, I was 27. And I was like, well, like, the one thing I do have is, like, I think I'm reasonably likable.
2:26:30I have a pretty good sense of humor. And I am, I guess, less self-conscious than many people when it relates to, you know, the internet. So I started, you know, with my 14 followers. So I think I'll make some jokes on the internet. But – and it felt so countercultural at the time, you know, because I think this is kind of before – VC was not yet as online as it is now. I think the pandemic was sort of a – you know, like the Clubhouse era pandemic, you know, last dance tweeting phase of venture was, I think, a bit of a paradigm shift. Anyway, so the origins of that was like, how can I possibly, you know, be like relevant to founders?
2:27:11and the internet is a fairly open playing field where if your profile picture is pixelated enough, no one really knows how old you are, right? And so then the first investment that I was truly led for Sequoia was Robinhood in 2017. And anyway. I have a question about that famous Sequoia memo that went out right as the market was collapsing. I believe this was during the housing crash. And it's this really deep macro econ analysis of what's going to happen. And I'm wondering if, is that macro DNA still alive and well at Sequoia? Do you guys still look at consumer confidence and what's happening in the credit markets and all of that?
2:27:59Or is that less relevant today? No, it's, you know, Sequoia is an interesting business. You know, in the sense that there is people talk a lot about, you know, multi-stage VC firms and what that means. And, you know, Sequoia, we have this, you know, thriving seed and early stage business that is sort of the heart of Sequoia dating back to 1972. And, you know, a growth equity business that spans series B's up through crossover investing into the public markets. We have this overlay Sequoia Capital Fund vehicle, which is the permanent capital behind all of these funds. We have Sequoia Heritage, which is a multifamily office.
2:28:48It's this independent business affiliated with Sequoia. We obviously see them all the time. Sequoia Capital Global Equities, which is, I think at this point, the largest TMT long-short hedge fund in the world. so you put all that together it's you know one of the benefits we have I think you know for being quote unquote multi-stage is you get you know different bites at the apple as companies scale and you know you try to catch everything at the seed and if you miss a few you can do the A and etc etc but also you do turn you know it's hard to really lose your mind on things if you have the forcing function of the public markets you know around the corner every single day and And also it occasionally does provide us with interesting insights.
2:29:32I think the RIP Good Times memo, which is the one I think you're referencing with the pig with the knife. The tombstone, I think. Actually, that's a great – that style of presentation, it's all substance, no flash. Every page is its own independent study. and I think that's held up so well in hindsight
2:30:00is I think something that Sequoia we try to do those sorts of things speak from the voice of Sequoia very rarely so we did this Black Swan memo right at the start of COVID which was kind of a similar warning but yeah it's a nice thing to have And obviously it helps both on the upside and the downside, spotting new interesting things and also making sure we don't just lose our minds. Yeah, the Black Swan memo was obviously really key because everything was about to change. But it did feel a little bit lighter on the macro econ. And I was wondering if that was because of the structure of the fund and RIA regulations or anything like that.
2:30:41But I don't know. Maybe it's just different style. I mean, it was very it was happening a lot more suddenly than the previous crisis, certainly. So there's less. Yeah. Well, I also I think, you know, going back in, you know, in 07, um, you know, like we're, like, we're out in Silicon, it's out in Silicon Valley and, you know, not every founder was reading the journal every day. Right. So there's a way in which you sort of had to shake everybody and you know, it's this, oh, there's this, you know, it's a New York problem, right? It's like, you know, it's a housing problem and sort of it was shaking everybody and and saying, no, this is a you problem.
2:31:17I think 2020, everyone's on Twitter all the time. Nobody was like, wait, this is the China virus. Like, it's over. I think we all remember when it hit Italy, when it hit Washington State, you know? So anyway, it wasn't different in that sense. Yeah, what you got, John? What's your approach to working on boards sort of broadly? You've sat on a ton of different boards, you've been board director at a number of important companies. is like, do you have a playbook and a system now that you come in and you say like, you know, you know, kind of grab the reins or like, what does that even look like? What is, what is partnering with you look like at the board level?
2:31:56Every, um, you know, every situation is just like, you know, every good board is, you know, is the same and every bad board is bad and so unique in an unhappy way. No, I think like the, in general, a paint by numbers approach to anything in technology and anything in investing is bad. And I think the worst board members tend to try to be the people who come in and take the reins and run their playbook, run that company through their playbook. Because playbooks age very quickly, especially in a world of accelerating change. So I think the first thing you have to do when you join a board is just take the time to learn what business you're actually in.
2:32:40And I think this is one thing that I've found is you can get to a pretty good sense of a company and its market at the end of your diligence process when you're writing the memo, but you haven't seen the people in that company respond to adverse events. you haven't seen what the real bottlenecks in the business are and you can have opinions but you sort of um it's much more easy to feel those opinions when you have dollars at risk so i think in general the first thing i try to do is like just figure out what business we're actually in and where the bottlenecks actually are um i think one thing i like to do pretty soon after leading an investment is do an executive search with the founder.
2:33:30I think it's like a very nice way to get from the opposite sides of the table when you're negotiating an investment. You know, the first thing you want to do is make sure you feel like you're on the same side of the table and, you know, go on the same side of the table and grinding a recruiter on how many leads we have in the funnel for our VP of engineering search. It's like a great way to align. And then there's this concept that Roloff talks about a lot of being a shock absorber on the board, which is something that I really, really believe in. When things are going well, many board members love to do the rah-rah cheerleading board meetings.
2:34:08And one of the nice things about being at Sequoia is in the portfolio review, if you're beating your number but your page is next to the whiz financials, it's like, okay, there's levels to this. And I think it's helpful to remind the best companies that there are current generation companies that are performing even better than this, and here's what they're doing. And then similarly, when things are going poorly, and poorly is almost always relative, right? But you can have a company that in a vacuum is doing amazing. We come in at 80 % of plan, and everyone's ho-hum in the board meeting. It's like, oh, and then we need to change our strategy.
2:34:50We need to change this executive. we need to, as always, we need to fire the VP of sales. Like, you know, I think oftentimes the best thing you can do is just like cause everyone to take a deep breath. Yeah. We're actually in decent shape. We have a lot of runway. You know, look at these metrics. Here's what's going well. You know, it's probably not a VP of sales thing. It's probably a product thing. You know, that sort of work is good. And then every now and then, maybe once every two years, you have a major thing you got to deal with. Yeah. an M &A, the capital raise, et cetera. Can you talk a little bit about the importance of concentration?
2:35:29I feel like just observing from the outside, the Sequoia playbook has basically been like, find the power law winner in whatever market, but then also get the most concentration and the highest ownership percentages. And we see that with the S1s go out and a lot of VC funds are saying congrats and Sequoia is the one showing up at the top of the cap table. Has that been something that's been kind of a drumbeat internally? Or has there ever been like a moment when it's been kind of harder to establish that foothold because of like crossover investors coming in or overpaying or valuations or anything like that?
2:36:05There's the WhatsApp example, which is like really famous, but I'm sure there's a million others, right? Yeah, I think our North Star, we want to be the largest outside shareholders in the most important companies of tomorrow.
2:36:19and it would be amazing if we were the seed investor in every single most important company of tomorrow. And we actually do a decent job of that. And if we don't do the seed, we try to do the Series A. If we do do the seed, we'll often try to do the Series A. And at any given round, you know, there's a relative risk-reward and also we can't forget that we're in the net multiple of money business. We're focused on investment returns. And if being the fourth biggest investor of the most, most, most important company is a good investment, we'll do that too. But I think the ability to double down and triple down on companies that we really believe in and hold them for the long term.
2:37:09That's the other thing that often gets lost. There's a famous example of Sequoia and Apple. Do you guys know how much money Sequoia made on Apple? No. Okay, well, I'm not going to tell you because it's not nearly as much as you might think. Oh, right. Now I think it's coming back to me. In the single digit, I think Sequoia made a 40X on 150K. Wow. So single digit million dollar gain. And when you have experiences like that and then you realize, you know, Sequoia owned 10 % of Google at the IPO, you know, the list goes on and on and on. So this idea that, you know, we don't necessarily need to be the first investor in every single company, but we should be the longest term investor.
2:38:02and for the really special companies, scaling up with them and compounding. You know, that's like the business that we want to be in. How do you balance both competition and partnership, both within the firm? Like you want to put up the biggest and best numbers, right? Like I imagine you're competitive and then externally too, you're oftentimes competing with other firms that are sort of your friends to win rounds at certain moments, but you're a multi-stage fund. So sometimes you're going to do a round and they'll do the next one and then you'll do the next one and then you got to be on the board with them.
2:38:36So like, how do you approach, you know, you have to sort of show up at some point to be like, I'm going to win this deal. And then at other time and then the next day you're back to being, you know, boys. Yeah, well, it's funny. There's been, you know, there's this one of those like really wonderful things about Silicon Valley. I remember one time I'm sitting in a board meeting and I'm sitting next to somebody who I know has a term sheet in on a company that I also have a term sheet in on. I think he must know and I definitely know, but we're not going to talk about it.
2:39:14Anyway, so I think it is actually quite wonderful, this ability to sort of – or even like it's constant, right? Like, you know, we're co-investors in this one company, but we're investors in the direct competitors in this other situation. To me, the lesson is just, like, always try to do business with good people and trustworthy people and, you know, try to find front stabbers, not backstabbers is a good rule of thumb. And, you know, like, if you do that, by and large, people are, like, good and ethical in the Valley. Like, it's a pay-it-forward place. It's a repeat game. I think there's a reason why a lot of the bad behavior comes from people who aren't from Silicon Valley.
2:39:59You know, it's like, yeah, it's actually a wonderful place to do business in that way. But Sequoia, you know, we are as cutthroat as it gets, right? We are frontstabbers. We don't hide that in general we want to do usually as much of a round as we possibly we can. We'll tell you what we're willing to pay. We're not externally, necessarily, with investment banks or with search firms known for being the easiest group to deal with. But here often, like, oh, Sequoia, I've heard it's really tough internally, really sharp-elbowed. And I mean, the stakes are really high. It's a small team. Expectations on performance are insane.
2:40:44Right? And it's a consensus. It's a consensus investment process as well, where you don't need to just basically close you. You've got to close, if you're an entrepreneur, you've got to close all your partners, correct? Yeah, exactly. And the partner meeting, I've seen founders, some founders really rise to the occasion in the Sequoia partner meeting room. I've seen people just totally wilt. But it's an amazing, it's actually an amazing culture inside the building. Because of that, I think like because the expectations are so high and the team is so small, right? I think like this is one of the things that, you know, Doug has this presentation, he calls it the laws of physics, where one of the laws of physics is that fund returns are inversely proportional to team size and also fund insurance inversely proportional to fund size you know we've been very disciplined about keeping the number of people on the team and keeping the funds to like a relatively stable size and then working together to kick ass that's the that's this can talk a little bit about paths to venture you mentioned that in your analyst class there is the high startups, maxed out on startups versus maxed out on finance.
2:42:03Are both of those paths still viable today or have they shifted over time? How do you think about hiring new people? I think with venture, um, and it's funny, I, I read and, and, you know, I do, I help lead growth stage investing business. So like venture, like true venture, you know, see it's a different thing. From my perspective, it sort of doesn't matter at all what you did before because nothing you do before joining a venture capital firm prepares you for how truly multifaceted long-term success is in venture. You can divide the job into sourcing, picking, winning, and company building. Those are probably the four core competencies.
2:42:52And what we try to do, or at least what I try to do, is I want to find somebody who can really hit the ground running on one of those things. So with finance, you'll have people who come in and they can be really good on the picking, like the investment process, due diligence, have a real point of view on why a company should succeed. But maybe you've never taken a meeting before one-on-one with anybody in your life. That was me. When I joined Sequoia, at Goldman, I was the headphones on behind the Excel sheet. I literally had never done a 101 meeting before in my life. I got to Sequoia, I turned to Doug, and I said, Doug, is there a sector you want me to cover?
2:43:32And Doug looks over and he goes, that's what we hired you for. Noted. So I like to find somebody who can hit the ground running on one of those things, but clearly has like the potential to max out on all of them if given enough time and then we just really try to invest in young people and give them opportunities to try things and take meetings and sponsor investments and um you know if you look at a lot of it a lot of the people at sequoia started in their early 20s here and apprenticed you know were nobodies for five six years and And then all of a sudden, people started recognizing them when they had amazing portfolios.
2:44:16And that's sort of what we try to do. Last big question I have for you. How do you think about underwriting generative AI investments today? There's been a lot of chatter on the timeline this week about what's ARR, what's not. And you're in a lot of the companies that feel like they will be these sort of power lodge and ai winners like even figma itself is like such a sleeper right and that like they have like most of the important creative people in tech in the app all day long they should be able to launch a ton of different products that that end up dominating in some of these categories but what's your what's your just sort of like broad approach to sort of and the way that you sort of look at all the revenue that's sort of just like popping up in all these new novel categories and will probably lead to big businesses, but some will kind of evaporate as well.
2:45:11It's a terrific, yeah. Figma, stay tuned everybody. The, the, it's, you know, it's the most transformative, it's the most interesting time in technology that I could possibly imagine. I actually think moments like today on the internet, you know, when you have the giblification of, of, of, you know, all of the last five years of memes happening at once. It's like a nice reminder of this, unlike prior technology waves, which were distribution extending waves where a small group of people had this at first, now a billion people have access to chat GPT and are logging on and doing the same thing.
2:45:52But turns out we have like, I don't know, a zillion computers in Virginia just making cartoon memes today, which is, it's like a small power plant being burned. These are coal-powered memes, everybody. Another one of these? Are you sure? We can do so much. So it's amazing, but because of that fact, everybody is on the internet all the time, and most of these technologies are available at the get-go to everybody. You can see revenue ramps like nothing we've ever seen before, right? Because it's transformative technology with full distribution from the get-go. And some of that will certainly be easy-come, easy-go revenue.
2:46:38I think some of it certainly won't. It makes each individual investment decision nuancing a Plex in its own way. I think the important thing is going back to that law of physics presentation that Doug gave when I first got to Sequoia. There are some things that never change in business. and there are some things that change all the time. And I think the important thing is knowing which one's which. Powerful. Powerful. I can see why you closed so many deals. You're great. You closed us. Yeah. Great having you on. This is fantastic. Yeah, thanks, guys. Thanks for having me. Come on. Come on with.
2:47:19We're going to have Dylan Field on, hopefully, at some point. You guys should come on and just we can all hang out. It'll be fun. That sounds wonderful. Like other good old days. Thank you guys. Awesome. This is great. We'll talk to you soon. Cheers. All right. Yeah. You know, he's in a quiet period for some of the... Thank you, Jordy. We heard your feedback to the fans. Jordy is keeping the soundboard and getting wild with it. He's in a quiet period on Klarna because they're about to IPO. And I've been saying that the SEC should ban quiet periods because I just think that people should be able to yap 24-7 regardless of what's happening in the public markets.
2:47:55So that's what I will be lobbying for the next time I'm in Washington, D.C. Ban quiet periods. Also ban lockups. Anyway, we have our next guest. Welcome to the Temple of Technology. What's going on? How are you? Hello. Thank you for having me. Long time listener. First time call up. Fantastic. Let's go. Amazing. Well, we were just talking about Klarna. We were talking to Andrew Reid. We were talking about how he's in a quiet period. Can't talk about it. but maybe you can, maybe you can take us through any of the interesting S1s you've deep-dived. But first, introduce yourself. What do you do and why do you deep-dive these companies?
2:48:35Sure. I'm Tanay. I'm a partner at Wing, which is an early-stage venture firm. We invest at the seed in Series A in primarily B2B companies, have the emphasis on data and AI. My journey with S1s dates to even prior to being an investor. I think I've always been just a huge business note, I guess. And I view S1s as the time when a company kind of lifts the curtain and gives you a peek into the business. And typically these are good companies, right? Because they've made it to the point of publishing S1. Maybe once in a while you do get like a WeWork or something out there, but by and large, these are generally good companies.
2:49:12And I think in every S1, different ways to read it. Obviously, there are public market investors that are making a decision on, hey, should we invest in this? How should we value this thing if we would invest? All of that. I think the lens that I generally take on it as someone that's not necessarily actively trying to make that call, at least in a professional setting, is more what are the lessons from this business that, one, you can just learn to learn from about business in general. But then, two, I think there are a lot of interesting things about many businesses that could be applicable even at the earliest stage, both in terms of what's going on today.
2:49:46of example, NES won today, a lot of AI washing in them, whether or not AI is related. I think it just tells you where things are at in the current state of the market, which is, I think, an important thing to know just given how quickly things are changing and how there's so much kind of noise and unclear where the signal comes. I think you do get some signal from it. And I think the second is like, depending on the kind of company it is, there's a lot that kind of similar startups maybe that are even 10 years earlier can take from it in terms of almost like lessons and playbooks of, oh, did like, you know, how did they solve that?
2:50:16How did they grow their dams over time? How did they figure out reaching these customers that are maybe difficult to reach? And so one good example of that is a company like Service Tied In, which went public recently. I think any company building a vertical AI, any founder building a vertical AI company should be understanding Service Tied In to figure out how do they expand all the products that they have? How do they grow their ACVs within that customer set? Because it's a problem and a question that they're going to get in every financing route, but a problem they're going to face as they grow that business.
2:50:46And there are some good kind of lessons in there about that. I remember when, I think when Snapchat published their S1, it revealed how much they were spending on the cloud and that might've moved cloud budgets. Have you ever seen an S1 that's been almost like a bull signal for a startup in the sense that like you can see a line item on some company that's going public and hey, they're spending a lot of money on this series B company. We got to go talk to them and see if we can do the next round? Oh, I think there have been a few like that in terms of it just tells you that some spaces are ripping as well.
2:51:18And I do think like this core we want is going to lead to a lot of other interest in, for example, the data center more broadly, because I think what's happening is that these basically all the big tech companies are spending$200 billion a year of CapEx on on that buy. Right. And obviously, NVIDIA is like the big, big, big winner and the beneficiary of that. But reading through Corwee, I mean, Corwee is like such a small pie, but it's like Microsoft's like 60 % of that revenue in like a year, right? And they're growing that revenue from 20 million to 2 billion in like two years. Like, you know, what companies have done that in the past?
2:51:49And I think what it tells you is that there's just so much demand at that layer of the stack right now that if you have a compelling solution there, even if you take up like a small, small pie of like the budgets, you can race into these hundreds of millions of revenue very quickly. So I actually think this one is going to be like thought of is going to renew some of the interest. And there has been a lot of interest in that chip companies and all of that. But in other areas and almost like that data center stack, just because there's so much spend and so much build out still happening. And so I think that's actually a good one in terms of in present day even.
2:52:21Yeah. You got some? It feels like CoreWeave is actively positioning itself to make sure that they're trying to get in front of the core questions, which is like, hey, you have some of these big contracts. They're long-term contracts, but they're not super long-term. Revenue concentration too. Yeah, revenue concentration issues. They've been working on this big open AI deal. Do you think that investors will even care? Obviously, you're not a sort of a professional public markets investors, but do you think the market just wants like a pure play stock like CoreWeave? Is that part of why they should just get out?
2:53:03And maybe the revenue concentration doesn't matter because we're in the meme market still. And in many ways, business fundamentals don't matter. It's so much around narratives. Vibes. Yeah. No, I totally agree with you. I think in this initial period, I think the how core we've trades is very much going to be about narratives and vibes and memes. And speaking of memes, I mean, this is a company that was started by three former commodity traders that, you know, basically got a bunch of GPUs together to mine crypto and then pivoted to AI, right? And this is like, this is the web three day I pivot done right.
2:53:38And it's going to be like whatever,$25 billion company in like six years, basically. So it's kind of incredible in that sense. I think on your core questions, I think so much of it is going to be around that. I think like the public market investors who are going to think about professionally are viewing this as like, do they have the potential of becoming like this kind of AI hyperscaler or this AI cloud? And they'll talk about, oh, we have the 15 billion from open AI commitments over the next seven years, six years coming through. We have a bunch of 15 billion of performance obligations from prior contracts that are already going to hit over the next three to four years.
2:54:17But then on the flip side, you have Satya going on podcasting that, hey, I'm glad that we're a leaser capacity because there's going to be a surplus of supply over three, four years. So that isn't a farewell for, is this going to continue beyond that potentially for them? And I do think it's one of those where the timing, I think it's, this is the moment in time, right? This is the time when markets, when there are some people I feel like they missed the boat on Nvidia, especially including the retail investors, right? Who are like, oh, this is now a new pure play generative AI workload company.
2:54:53And so I do think it's good timing from that. The other thing though, is they need the capital because they do lose a decent amount of money on like a free cash flow basis. I think they lost like$6 billion last year. And so they have some cash in the bank, but they do need the capital. So it feels like a good time for them to get out and, you know, play into these memes and narratives and also, you know, obviously allow some of the early investors to like that. Sorry, you can't hear it, but he just played the loudest gong sound because of how loud, how big the$6 billion is. How much money they were losing.
2:55:27We really got to figure out how to guess and can hear the sound effects. If you don't get the joke, it's not going to land. We're just cracking up. Anyway, I want to stay on CoreWeave. There's been this narrative of like, will they get out? Won't they? I saw Martin Scrawley say, oh, they're 5X over subscribed now. Can you just take me through, like, what is the emotional roller coaster that a company or CEO or even a public markets investor goes on when one of these S1s drops? There's the quiet period, and then there's the S1, and then later there's the IPO. And, like, what is the normal cadence and what should people be paying attention to?
2:55:58Yeah, unfortunately, no personal experience in that. It would have been nice to take a company public. But, no, I do think it's definitely a stressful period for them in terms of just... And especially when there are so many questions around it, like there were some rumors about Microsoft not being like trying to break free of some of the obligations, which they've now dispelled. So with this one, there's some added kind of just question marks on is this going to happen? The Cerebrus, which file last year had similar issues. And there was some CFIUS kind of investigations that have taken longer and it's kind of held back their IPO.
2:56:29And so it is one of those where now I think there's some rumors that the founders here have sold a decent amount of secondary already. So I'm sure, you know, that probably helps a little bit, especially when it's like$450 million worth. But ultimately, it still is. It does look like it's going to go ahead now, but I can imagine it's going to still be a crazy kind of first few months once it does stop trading as well, just given so much on Sony around it. Can you comment on what you think might be going out in the next year, what people are excited for, what kind of names are people chomping at the bit to read the S-1s of.
2:57:09Yeah, we saw, wasn't it Stripe? I mean, you imagine it's everything on that most in-demand secondary once you're in the Decacorn category, but I want to know. Yeah, yeah. And I'd even love your take on, do you think figures going to bring back the spec? Have you heard any rumblings? Probably can't say, but I'd be curious. Yeah, I think in terms of what people will be excited to get out, I do think Stripe is definitely one of those. I don't know if they plan to go public anytime soon, given that they seem to be tapping into the secondary markets pretty often and providing liquidity to employees, potentially to some other investors as well.
2:57:49And so maybe not an immediate need. There are a few that are rumored to be filing soon, such as Chime, which I think will be an interesting one on the fintech side. I think the markets would really love like an AI company going public. I think CoreWeave is one example of that. But I don't see an OpenAI or something like that anytime soon. But maybe if one of those kind of labs decides to do it, I think that would be pretty interesting. I think the application layer companies are probably still a bit too early is my guess. And so I do think we'll see some. And I think one of the interesting things is that CoreWeave is relatively young, right?
2:58:29It's about seven years old as a company. But Klarna, which filed, is like almost 20 years old. And then StubHub, which also I believe is five, I actually haven't gone through that one yet, is also like a 20-year-old company. And so these are some of the ones that have been like, you know, people have been waiting for these to go public for a really, really, really long time. And so it's nice to see some of them start to get out. And I do think we might see some younger companies start to go just given how fast the AI boom is. But I think it might be another couple of years before we get like, you know, like a true AI native company out in the public markets.
2:59:05I've heard a little bit of rumbling from investors that track the CFO hiring cycles at the large later stage growth stage Decacorn companies. and they say, hey, this CFO that just got hired at this company, he took his last three companies public within 18 months of getting hired. We know what's up. Is that a valuable signal for anyone? Or is that just some like kind of fun fact in Silicon Valley? I think there's a lot of truth to that. I think the timing can vary though from that hire in terms of, but generally, especially some companies don't even have a CFO. And so when they bring one in, some people believe it's like, it's like this now starting to at least think about it.
2:59:41But I do think it could be that it's imminent like a year from now it could be like it still might take two to three years because some of the internal controls and things like that like corey for example had like a note in its filing that they found like weaknesses in their internal reporting and controls which obviously is not is not great to uh from like someone who would be a prospective investor looking at that and so some companies just like to do it a little bit earlier to make sure that they are there's none of that ahead of when they start when they actually file yeah give any questions uh you got one i i i just want to have you back on the next time a big s1 drops yeah i'd love for it to be a regular yeah this would be awesome when you when you drop a new uh deep dive and i and i still i definitely want to know uh yeah i just just what the next big s1 will be it's uh we are is there polymarket for this yet i want to bet on that that would be good yeah that would be good and it might actually incentivize them to actually go public which maybe isn't the worst thing in the world.
3:00:40I'm sure you've seen the DoorDash memes about Klarna. Oh, yeah, yeah, please break that down. Here's a fun fact from the filing. So Klarna is registered, it has a banking license in Europe. And so they basically get deposits from people in Germany and Netherlands and places like that. And so that money is being used to fund the interest fee loans. And so basically what's going on is that the European savings are being used to fund like the Americans ordering burritos. Let's go. I thought it was just incredible. It's like American exceptionalism at its peak. Like the American consumer is the winner yet again.
3:01:16Amazing. Undefeated. Nothing can stop the American consumer. It's the most powerful force in the world. Yeah. Well, I mean, thank you so much for coming on. This is a wonderful chat. Give us a heads up when next time you're dropping. Yeah. Yeah, we'd love to have you on the same day. We'll slot you in and I look forward to it. Awesome. Really enjoyed this and thanks for having me. And yeah, looking forward to the next one. Yeah, for sure. Awesome. We'll talk to you soon. Cheers. Thanks so much. Bye. Yeah, Ben, tomorrow's challenge to make the show 1 % better is let the guests hear the sound effects.
3:01:48So we don't just, they're just saying like some, you know, normal thing and we're just cracking up. Yeah. It's great. Well, we should close with this timeline post from Gokul Rajaram. He says, Prediction, one or both of Canva and Figma will enter the AI builder market in the next few months, likely with a product very similar to Bolt.new or Lovable.dev, but tightly integrated into their existing products. There is no way these two very smart companies will tolerate being disintermediated by this next generation of AI native companies, which are essentially verticalized builders, design plus engineering platforms.
3:02:29It also supports their stated strategy to democratize building. I say this with zero knowledge of the roadmap or internal discussions of these two companies. You mentioned it on the chat with Andrew, but what's your take? Yeah, I mean, for me, as a Figma enjoyer and active user, it's always been very frustrating that I can design something in the app that's pixel perfect that I can even mirror on my device and it's pixel perfect, but then I can't even get the simple functionality of, I've got to go to Webflow or Framer or actually build it out in Vercel or whatever. And it's just always, it's such a natural extension for them.
3:03:12And that's going to be the challenge for Bolt and Lovable because all the, they do some of the design side stuff as well, or they'll use like templates or you can just prompt it. but there's something about the sort of precision design that end up being the most sort of like durable. creative properties or the most durable websites are probably the ones that are actually sort of designed that the sort of high quality customers end up caring like a lot about precision design yeah i i mean i'm so excited for this because i feel like the the business that we're building here with tbpn is is great and cool but it's not one that can really support a software engineering department yeah but we will have problems that can be solved by small apps and with this new wave of generative AI with better tools from Figma and integrated AI building products, we will be able to either personally in a weekend build something that makes the show 1 % better or hire someone to work on that for just a weekend or a week or even have one person on staff that can actually build and maintain systems that speed up our workflows.
3:04:28And then of course, some of those will turn into SaaS products that we'll wind up building. And we'll just say, hey, yeah the thing that we built internally we're shelving it it doesn't matter if we built it in a few days as opposed to previously uh the idea of building custom software at a media company was crazy right yeah it was a huge waste of resources in almost every scenario totally um so i'm i'm really excited for this i have a bunch of ideas for uh little apps i want to build deploy to my phone uh i was reading a post from uh andre carpathy and he was saying that he's now gotten to the point where even though he doesn't know how to program for iOS, he was able to vibe code an iOS app, deploy it to his phone, and he doesn't need to go - World's most talented programmer.
3:05:10It's like, you probably learn it in a weekend, let's be honest. Or in an hour. But yeah, he was saying he could kind of program in languages that he was familiar with and still deliver in iOS, get it running on his phone. And the really cool thing about that is that you don't, if you're just, if you're a customer, a developer of one for a customer of one yourself, you don't even need to go through the app store. You can just deliver it directly to your phone. And so all of a sudden that opens up a whole new world of things that you can do that the app store might say, hey, no, this is a violation of some sort of TOS, or we need to review this.
3:05:45We can't change this. And by the way, it's going to take you two weeks to approve this. You're going to be able to just say, hey, I have an idea. I need it on my phone. Boom. The next day it's there. And so I'm really excited about that. anyway great show today thanks for everyone to tune in do we want to read our uh our review oh yeah we have a review i'm gonna pull it up i think you're the best i got a review with an ad in it and this is from that hamelon it says the espresso martini podcast tbpn has immediately become one of my favorite listens awesome guests and great conversations also incredibly bold for being the first u.s podcast on a 996 schedule thank you one thing to improve i have no way to get notified when they drop the TBPN branded Italian suede driving slippers, they should get set up on Lalo, which powers drops for the world's biggest creators, musicians, and live events.
3:06:32Other than that, great show. I've been... Lalo.com, the drop CRM. We actually 100 % at some point will have Italian suede driving slippers. So go to Lalo. Is it Lalo.com? Lalo.com, shout out. Great domain. We will pump the stock. Go to Lalo.com. And this business just makes a lot of sense. Drops are the sort of default model for the internet. It's what the internet likes. So building a tool that sort of provides around that. It's a hassle to do a drop. If you're not set up, and even if you have a Shopify store, you're like, okay, now I have one SKU that's only going to be up for a couple of days.
3:07:09It's like, it can be a hassle. So I mean, I haven't used the product yet, but I think it's a great idea. And I'm happy that they wrote this in and sent us the message because I'm going to check it out. Yep. And it seems like there was a big client. They worked with Nicki Minaj, Rolling Loud. Speedway Motorsports. We love cars. Usher. Porter Robinson. This is great. Outside Lands. Yeah. Thanks for writing in. This is very cool. Rufus. Crushing it. Lalo.com. Go do it. And thank you for listening today. I'm excited for tomorrow. Yeah. Thanks a lot. It's flying by. Yeah. Cheers. We'll see you tomorrow.
3:07:41Bye.
From the publisher
TBPN.com is made possible by:
Ramp - https://ramp.com
Eight Sleep - https://eightsleep.com/tbpn
Wander - https://wander.com/tbpn
Public - https://public.com
AdQuick - https://adquick.com
Bezel - https://getbezel.com
Polymarket - https://polymarket.com
Follow TBPN:
https://TBPN.com
https://x.com/tbpn
https://open.spotify.com/show/2L6WMqY3GUPCGBD0dX6p00?si=674252d53acf4231
https://podcasts.apple.com/us/podcast/technology-brothers/id1772360235
https://youtube.com/@technologybrotherspod?si=lpk53xTE9WBEcIjV
- (00:13) - Studio Ghibli
- (24:50) - Anthropic Copyright Dispute
- (01:00:18) - Immad Akhund
- (01:30:41) - Steven Schwartz
- (01:46:31) - Jeremy Levine
- (02:00:55) - Rob Mohr
- (02:19:02) - Andrew Reed
- (02:48:05) - Tanay Jaipuria


