The venture market is in flux. AI is booming while the rest of venture is in the dumps. Venture capital as an asset class is evolving and retrenching. Fewer deals are happening at higher prices. New marketplace types are emerging. AI is impacting marketplaces in unexpected ways. New technologies are coming of age. In this episode, I cover it all!
For your reference I am including the slides I used during the episode.
If you prefer, you can listen to the episode in the embedded podcast player.
In addition to the above YouTube video and embedded podcast player, you can also listen to the podcast on iTunes and Spotify.
Transcript
Hi, everyone. I hope you’re having a wonderful week. So, two weeks ago, I was in St. Tropez and we were doing the FJ Labs offsite. And for that offsite, I basically prepared an update of how the venture market is going and realized there are actually really profound trends and changes that are afoot.
From, the bubble and AI and kind of the tale of two cities between venture as a whole, like doing terribly except AI, that is like an absolute bubble to new technologies emerging to frankly, the, the nature of venture capital firms and venture capital funds changing profoundly. And as I was putting all this material together, I figured, okay, needs to be shared because things changes or afoot.
And so without any further ado, let’s get going. Welcome to Episode 50: Venture Market Update.
So the first thing I wanted to do is basically cover with you guys what’s been going on in venture capital and the venture market as a whole. And if you look at like what’s been going on, from a macro level, a global level, it feels like venture capital is recovering from the doldrums. So this is global venture investing.
You see Q3 to Q4 and massive growth and Q4 to Q1 and massive growth. But actually, when you start delving deeper into the details, you realize it really is a tale of two cities. We didn’t see an increase in investing in Europe, in Latin America, in Southeast Asia, in Asia writ large.
It’s all in the U.S., it’s all late stage, and it’s basically all AI. On a global level, AI, represented 53 percent of funding in Q1 and 71 percent in North America. So, AI is completely dominating the investment and all the other categories. It’s really hard to get funded right now. And, if you look at what’s been happening, basically, there are fewer companies that are raising, more money at higher prices, right?
So this is the median. So, the mean is higher by the way, because the extraordinary deals, the bonkers deals are moving everything up. But the median seed deal is being done at 18, and this is pre. The median series A at 55 and the median series B at 131. But the bonkers deals are so like two standard deviations of the right of the mean and there are more of them happening than ever before in a way, mostly in AI category, mostly with repeat founders or people that were at Open AI, et cetera.
And AI is really eating the market. So this is the premium in valuation if you’re an AI company in every vertical, and it goes from 69 percent to 314 percent premium just for being in AI. So basically, AI companies are eating the world from that, from that perspective. Now, what’s really cool is FJ Labs itself has actually remained disciplined in a way we’ve been contrarian on AI.
I actually covered our AI thesis in a prior, Playing with Unicorns episode, so I’m not going to rehash it here. But the one line summary is instead of competing on the core LLMs, we’re investing in applications of AI at reasonable prices. But if you look at actually at our history, even in 21, we remain disciplined, or median, and this is U.S. only, obviously, because if I included other countries that would be skewing the results. But our median valuations have remained, you know, like around 16 at seed around like 40 at series A. Both in 21, by the way, and in 24. And our series B, [even 20 and, whereas like in the eighties to hundreds, I mean, smaller sample size, so it moves easier, but we remain well, well below, the U S media and partly also because we haven’t been focusing on the super-hot, AI ideas.
Now, the exit market has been recovering a little bit, and there’s been a few more IPOs, and of course, recently we saw the Chime and the Coinbase IPOs, but for the most part, relative to the exuberance and the massive number of exits in 2020 and 2021, the markets are still much more close, and we were hoping, I suspect, that in 2025 we were going to see the M&A markets, the IPO markets open up, but between the geopolitical tensions, the tariffs, et cetera, it hasn’t really happened yet.
Now I’m hoping that the political turmoil calms down as, they try to get reelected for the midterms next year and maybe stop doing as much disruptive things. So TBD, if that happens by now, I’m hopeful that things will improve for 26. But you know, TBD, this is exogenous to the market, and not endogenous to it. And so we are market takers, not market setters.
So nonetheless, it’s a little bit better and I’m hoping it’s going to continue to improve with a, you know, slightly declining rate environment and better liquidity conditions writ large as I’m sure they’re going to be pushing for fiscal expansion and maybe creating a little bit less disruption. That’s a hope, not a reality, but TBD, and hopefully it’s moving in the right direction.
Now what’s more interesting is venture capital as an asset class is undergoing fundamental transformation and I would say there’s a schism, right? Like the top 10 firms have basically raised 50 percent of all capital and the top 30 firms have raised like 75 percent of all the capital.
So it feels like there is a market where on the one hand you have like massive capital accumulators in Andreessen Horowitz, Thrive, General Catalyst that are accumulating billions and billions of dollars. I’m not even sure I would call them VCs anymore. They’re investors across the life cycle, the companies they invest, they stay active investors when the company is public, and so I’m sure that the IRR is that they’re going to be getting is more like 10/15 percent than the historical 20/25/30 percent that people were shooting for in in venture capital. And on the other hand, the emerging managers that have less than three vintages they’re getting less and less capital, and all of them combined are getting like 23 percent of the fund, which is a 10 year low.
So we’re seeing more and more mega funds that are over a billion, and that are capturing more and more of the capital. And then you have like minnows and small funds. And so you need to be hyper specialized and small or you need to be big. Everyone in the middle is not doing well. And as a result, we’re seeing a massive decline in new funds and a 48 percent decline in venture funds rate, new funds, 68 percent decline over a three year period.
And as I said, nine firms raised 50 percent of all venture capital. Essentially 2000 VC firms closed since the peak in the last three years. And we’re seeing a lot of GPs leaving, retiring, et cetera. So massive consolidation, venture as an asset class is retrenching writ large.
And the reason is, of course, IRRs have been declining. Now, part of the reason IRRs have been declining is people came in and way too high prices in 2020-2021, maybe early 22. And companies did not grow in these valuations. And so the returns are declining, especially since there have been no exits, right, the markets, the M& A markets, because antitrust have been closed 22/23/24/25. And there haven’t really been many IPOs. And as a result, DPIs are very, very low. So the LPs, the investors into the funds have not been receiving distributions. So they have no capital to reinvest. And as a result, you know, you haven’t been getting the benefit of investing in privates.
Back in the day when you’re investing in a venture capital, you would get a premium to the public markets. But over the last 20 years, Nasdaq has actually done better, obviously mostly driven by the Magnificent Seven and the AI companies in the last few years. But you’re not getting the liquidity premium.
Now, of course, this is venture as an average class and the top quartile is doing, doing a lot better. But even though the top quartile is doing a lot better. It takes forever. What we’re seeing is the best companies are staying private longer. And so what is really happening from a mega trend perspective is companies are saying private 18 20 plus years, right?
I’m an investor in Founders Fund II which invested in SpaceX. I think in December 2007, like an 80 million valuation or so. It’s 18 years ago and obviously SpaceX still private and that’s compounding is doing amazing, but venture funds were marketed as 10 year vehicles plus one year and one year extension.
So maybe 12 year funds. Yeah, now we’re getting to 10-12 years into the life of the funds, and most of the value is still not distributed. And so exits are taking a lot longer than before. So top quartile still doing well, but distributions are taking a lot longer, the best companies are staying private a lot longer.
And as a result, a lot of the LPs feel we haven’t had distribution, we have liquidity, we’re overexposed to the category, we’re retrenching in venture as an asset class. I think now is the very best time to invest in venture, the same way that the late 2000 and, 2008, 2009, 2010, during the Great Recession, was the best time to invest in venture as an asset class.
Because the companies were being created, raised at lower valuations, so there was less competition, so when they won, they won the entire category, and that’s just when like Airbnb and Uber and Twilio, WhatsApp, and Instagram and Slack all came of age. And I suspect, outside of AI, so non AI companies that are raising right now are going to be the same defining companies of the 2020s, because same thing, they’re raising lower valuations, there’s less competition, and the people that are earned there are in it to win it.
So what I think is going to happen, and frankly is happening, is secondaries are going to become part, a larger part of the playbook of venture funds. So far, 73 percent of funds have never participated in secondaries or sold through secondaries. So they either wait until things go public or get acquired. And that’s it.
But as liquidity has dried up, and these things are taking longer and longer, I think more and more people are going to explore having, doing secondaries. Now, there are multiple ways you can do secondaries. you can be buying or selling on like the platforms like Forge have brokers.
The issue is, this is really limited to the very top venture funds names out there like SpaceX and Stripe and OpenAI. And Anthropic and Databricks. And these have liquidity, but everything below doesn’t really have liquidity, and these that have liquidity, I’m not even sure you want liquidity. For the companies below, secondaries happen more likely around a round that is happening.
so what we’re seeing right now is the, the evolution of like three types of secondaries that are happening. One is, people are buying and selling secondaries in these platforms around rounds. Number two, there are funds that are being created to buy LP positions. And so I think what is a very interesting asset class right now, is buying a fund that’s been 8, 9, 10 years in its life.
So it’s closer to liquidity. You understand the underlying assets well. And you buy them knowing or expecting that there will be liquidity in the next few years. If you know the names well, you can buy this LP position probably at a 50 percent discounted NAV, you know, leading to very good returns.
You can also invest in the GP, at a very big discounted NAV as the GPs are in dire need of liquidity, if only to make their own capital calls for the future funds that they’re creating. So secondaries are coming of age. Now, of course, not everyone can always do secondaries, right? If you’re, a venture fund and you’re on the board and you in 25 percent of the company, if you sell in the secondary, it probably kills the company because any potential investor is going to say, wait a minute, what is it they know that we don’t know?
And as a result, they wouldn’t want to invest. But when a fund like FJ Labs, part of the reason we’ve done very, very well. Is we own only a couple percentage points of the company and so it’s not a signal when we sell and it sells just selling is just part of our strategy as we’re recycling capital to invest in new investments on a go forward basis because we really are venture investors.
We want a 10 X. If there’s only a three X left, it’s probably an amazing company and it’s doing very well, but like maybe it fits more in the P. return profile than it is like our early stage venture profile. So it’s better off for us to sell and reinvest early. And often, by the way, the founders are asking us as a favor to sell.
They’re like, hey, Greylock, Sequoia, Andreessen, they all want in, but they all want 15%. I don’t want 45 percent dilution, but I want all three of them in because I don’t want them to fund a competitor. So please, you know, why don’t we take 25, 30 percent primary and would you mind selling part of your stake and the strategy we’ve done here, by the way, is we’ve typically sold about 50 percent of our holding, , and again, we only sell the winners.
So it’s kind of the anti VC strategy when we feel that they’re fully priced. and 50 percent it’s kind of a no regrets philosophy. If it goes to the moon, we still have 50 percent of right. It’s amazing. If it goes to zero, we’ve made 10 X on the 50 percent we sold and we’re very happy as well. So kind of works.
No regrets philosophy. This is going to become more common. So easily, a lot of things are happening in venture. We have the mega funds that are asset accumulators. We have smaller funds that need to be beautiful and specialized like FJ on network effects and asset light businesses. But you have like a lot of dedicated small funds, a couple hundred million that could do well or like someone like Benchmark, will continue to do very well.
LPs are changing. Obviously, University examiners were large LPs. But of course, with the idea that they may now be taxable, and we’re no longer getting donations from the federal government, this app LP base is probably disappearing, or at least shrinking dramatically. And they’re even considering selling part of the position.
A lot of the firms are considering, as they’re becoming asset accumulators, going public. So General Catalyst is exploring an IPO, and there’s rumors that Andreessen might do the same. And in a way, it’s kind of following in the footsteps of the PE firms that have done it before, like Blackstone, KKR, Apollo.
And now you’re seeing also permanent holding vehicles. Companies like Sequoia and Thrive are just holding the assets forever, right? So there’s no idea of like needing to create new funds on a go forward basis. So basically the summary is, venture is changing. A lot of funds are dying.
There’s like a complete implosion of the asset class with a clear segmentation between large asset accumulators that are brands that are very likely going to become a combination of permanent holding vehicles and going to go public. The general catalyst thrives and reasons of the world and beautiful, small, specialized funds of a couple hundred million and affects slow box group – FJ, Benchmark, and those will do well and everyone in the middle probably will die and will suffer as a result. So profound shifts in venture capital that are happening right now.
Next thing we’re to cover is like some of the cool investments we’ve done in Q1. That are highlighting like interesting trends especially in marketplaces, of course. So I already did a marketplace trends update in January, speaking about like a lot of things happening in B2B marketplaces, speaking about things like live video shopping company like Palm Street that was selling rare plants, but now I’m going to cover the investments in the first half of the year, which are highlighting other trends that are emerging, in marketplaces writ large.
So we invested in a company called Garage. Garage is a fire truck marketplace. We’re talking to like 30K AOV. And what’s interesting is before garage existed and the founder is amazing. The product is amazing. Basically the fire houses would, would list the trucks on Facebook marketplace.
And they’d have very little liquidity, you know, it’d be complicated for them to shut. So the way they unlock the marketplace, they said, you know what, we’re going make these end to end seamless experience. We’re going to manage the shipping, we’re going to manage the payments, we’re going to take care of everything.
And by the way, on average, these trucks are being shipped 1200 miles. So they’re being put on a flatbed and driven away. So there’s all the complexity. So adding a service layer on top of Marketplace can often unlock a category that didn’t exist before. So fire trucks didn’t really exist in the Marketplace.
And this is something we’ve seen before. I mentioned in the last, podcast presenting the Marketplace trends, Tetra, which is a heat pump installation marketplace where basically they took the entire hassle of like replacing your heat pump out, but like you didn’t need to get a general contractor like you basically show what you had and they would get the person for you. Or Alpagga, which is a restaurant equipment marketplace, which did installation and shipping, again, unlocking the value of the marketplace.
So basically integrating a service layer in marketplaces, is unlocking new verticals, and Garage is the latest example of that. Next, interesting example is, P2P rentals is finally becoming or coming to the front. So when Rent the Runway was first created, they actually tried to be a peer to peer rental marketplace.
Of course, it’s in a way, the margins are better, you’re doing a lot less work. But the reverse logistics, didn’t really work. People didn’t trust each other. And so they needed inventory. They needed a warehouse. They needed to do dry cleaning. But of course, it’s a much more capital inefficient, or capital intensive and capital inefficient business with much lower margins.
By now, a company like Pickle has completely cracked on that. People are comfortable doing it. And maybe it’s just because now they’re comfortable, you know, riding in Ubers and seeing people’s places in Airbnb. And so the behavior pattern has changed and people are comfortable with it. But obviously they’re renting from each other and they’re doing amazingly well. And there’s a very high NPS, you get better pricing, you get better inventory, and everyone’s been extremely, extremely happy with the way the market is going.
And this is, by the way, the 50th time that I’ve seen this tried. There’s a company called Z Lock in France. There was one doing power tool rentals peer to peer in San Francisco, and none of them worked. But here, there’s recurrence, there’s INPS, people are loving it.
I mentioned last time as well, B2B recommerce. Obviously, recommerce is big in the consumer world, but it’s finally coming to the fore. I mentioned a company called Ghost last time, where excess inventory in fashion, was being sold. So historically it was being sold to big outlet stores like Century 21. And for the first time through Ghost, they were able to like sell smaller units such that retail stores could benefit from buying excess clothing at like 90 percent off and a lot of consumers would benefit of it.
And Rebel is another example of that, doing it for baby gear, and creating a marketplace in the category. So again, that B2B recommerce is not coming and doing it at scale and working wonderfully.
And the last big trends, it’s kind of a mega trends. Kind of an evolution of the old search funds where now you’re seeing a lot of like little roll ups where people are buying these old school companies and bringing a tech layer or bringing AI to make them way more efficient. And this is happening in kind of every major vertical of offline businesses. People rolling up laundromats, people are rolling up barbershops, nail salons.
Every major category you can think of bringing it to clear, making operations more efficient, getting a P.O.S. Getting skill in the inventory to improve margin. Slow ventures has been the main proponent of the strategy in the venture side and with AI allowing companies to be more efficient, have lower costs.
It’s absolutely, a mega trend that is only at the beginning.
Next thing I want to talk about is the impact of AI marketplace. So as I said, we are not investing directly in the LLM companies, but everything we do is touched by AI. And the main concern that people had when they talked to me in their AI, in their marketplace companies, they’re like, is AI going to disrupt us?
Or is the discovery going to go to the LLMs and we will no longer exist and we will be fully commoditized, pure execution machines? And I thought long and hard about this and ultimately decided that that’s not the case, right? If you think about how people or shopping, there’s three behavior in marketplaces.
So one, there’s shopping as entertainment. You don’t know what you’re looking for, and you’re looking for like fashion items. So this is the way people shop on whatever classified site or Vinted. They basically go there and they browse through listings. They may not even be in the market for something, but because items are 30 euros or 30 dollars or 40 dollars, they see something they like, they buy it.
There’s a very minimal role here for LLMs because there’s no point in it. We’re not trying to get efficient. We’re just trying to see cool things. And do I think that, like going to open AI and telling them, Hey, create for me a shopping interface for fashion that matches my needs is something that’s even in the top 1000 priorities for Open AI, Anthropic or Grok. No, absolutely not.
And by the way, even if that type of funnel was taken by the LLMs, Because these are winner takes most, it’s not winner take all marketplaces in the back end, most of the value would still remain with the actual marketplaces. Neither am I worried that it’ll disrupt the top of the funnel here, nor am I worried that it’ll margin compress.
The other behavior people have when they go to these marketplaces is they know exactly what they’re looking for, and so they put in the search results. So, when people go to Amazon, they typically type in, you know, LG C4 65 inch OLED, and they get that one item. And actually, eBay is also a searcher in marketplace.
Now, here, same thing. Is there room or scope for an LLM to really profoundly change that, given that you know exactly what you’re looking for? Not really. Now, can you improve the search results? Probably. Can you propose related items? Probably. And the reality is, in a way, these searches also somewhat happen on Google.
But what’s interesting is, if you go to Google and you type these items, for instance, in Electronics, almost all of the results for eBay and Amazon because they have such a large market share in these two categories. Amazon and eBay have 43 percent of the U.S. e-commerce on it combined. And in some verticals, we’re talking like 90 plus percent.
So, can I see a world where LLMs replace Google and capture the same value as like Google search in terms of getting top of funnel? Yes. Will they capture more? No. Because again, the fulfillment, the inventory management, the shipping, the insurance returns, all of that will be done and you will continue to have your core value.
And that’s true in every major vertical. So I’m not worried about like LLM destruction of marketplaces. The ones that are search driven. And in fact, I can imagine that you can use them, use the LLMs to actually help people find product discovery. And so I think the way, in fact, I’ll talk a little bit later on the next page, how I would implement, and I use LLMs in a second.
The third one now is more open for discussion. So there are categories where you don’t know exactly what you’re looking for. It’s a considered purchase. You’re buying a car, you’re buying an apartment, or you’re buying high end, whatever, ski equipment, and you don’t know much about the category. So there, typically, the journey you would have gone in the past is you would have done a lot of research.
You’ve gone online, you’ve spoken to whatever car dealers, you would have test drove the cars, etc. So here, there’s definitely a role for an LLM to help you iterate and think through what is best for you. And in fact, I’ve used ChatGPT for exactly that purpose as I’m considering moving from Turks and Caicos to Antigua or Nevis.
And it’s helped me think through, first of all, that these were the right destinations, find the proper properties, find brokers, understand the valuations, etc. So it’s been an amazing tool for that. And I think it’s part of the reason companies like curated dot com, which you or which uses human agents to make recommendations.
I’ve actually sold a reasonably low valuations relative the amount that they’ve raised, because I think there’s a risk and fear that perhaps they’re going to be disrupted by the LLMs. But even here, it’s not obvious to me for these considered purchases that GPT is the one winning it versus an LLM on the site itself, right? Like so imagine Carvana can probably have the very best car suggestion engine because they have all the data that they understand exactly what people are looking for used cars, and they’re probably better positioned to give you advice on what car to buy that then GPT, maybe, maybe not TBD.
But even then, even if it was ChatGPT that made the recommendation, ultimately, the transaction happens in Carvana, they still capture a lot of the value. I’m I’ve been using Rufus, which is the AI in Amazon, for instance, to ask for product ideas, and it’s actually really, really good. And so one of the things I can imagine most marketplaces doing is actually adding a layer where if you want recommendations, it’ll be there.
In fact, in the long term, the idea that you have two search engines, one for search results, and one for like conversations about products makes no sense. I think it’ll be one box, and that box will be flexible. You can say what you’re looking for, or you can ask for advice, like the long form questions, and you will get the answer, and the search engine will be good enough to know which one will be it.
And so people have been asking me, okay, if I’m a marketplace, should I be indexing myself in the LLMs? And so the answer to me is, is if you are indexing yourself in Google, then the answer is yes. Which means given that 99 percent of the people were indexing themselves in Google, the answer is absolutely yes.
You should be doing AEO, which is the way they’ve called like these LLM, engine optimization. And so you answer engine optimization. You should absolutely index yourself there because it’s kind of free traffic for now, and because many people are not doing it and not doing it well, you’re going to be at an advantage of capturing a larger share of this free traffic.
Now, the case is where I can imagine you don’t want to do it. Imagine you dominate your category. You have like 99 percent market share, and when someone wants something in this category, they’re coming to you directly. They don’t go to Google. You’re then I can make a very good case for not indexing yourself in Google and therefore not indexing yourself into the LLMs.
But if you’re indexing yourself in Google, index yourself in the LLMs. So in other words, the LLMs, they’re not a threat to the marketplaces, but they are a profound existential threat to Google. But what I’ve noticed in my personal usage is I no longer use Google, whatever I’m searching for, I just ask ChatGPT.
And obviously, I’m n of one, but it’s so much better to have one answer than have a billion answers and to be able to iterate and thought, think through, which is why for Google, like getting Gemini right is existential, but also it’s an existential threat to their business model, right? Right now, there are no ads in, in, in the LLMs and it’s just a certain product.
So it puts the core and underlying business model of Google address. So Google when may have a large share in the AI space, but if I was them, I’d be focusing all my efforts in that because OpenAI is very much at risk of being disrupted by OpenAI and this is what would be keeping me up at night, all night, every night.
Versus whatever, a Meta, an Apple, an Amazon, an eBay, I think you can use it to do interesting things, but like, it’s not an existential threat to you, basically. So, instead of worrying about LLMs that are disrupting your top funnel, which I don’t think will happen, here’s what I would do.
First thing is, people have actually been using AI. For the first time ever, do proper cross border trades. So it used to be, when you were in Europe, you’d launch a site for France. Then separately, you’d build a completely different site with different items, different liquidity, in Germany. Then a different one in Poland, a different one in the UK.
But now with AI, for the first time, Vinted has created a truly Pan-European and frankly, even global marketplace. The listings are translated automatically. The conversations between the buyers and the sellers are translated automatically. Now, it only works if, because they also have integrated shipping and integrated payments.
But for the first time ever, you can actually have a true global marketplace and a true Pan-European marketplace by having auto translated conversations, auto translated listings with integrated shipping and payments. So obviously you do need to nail that, but if you have that nailed, it allows you to create.
And so cross border is a mega trend. Ovoko is an amazing car parts marketplace in Europe, sourcing car parts in Eastern Europe and selling them mostly in France. Even in B2B, we’re seeing CarOnSale now as 30 percent of their cars being sold B2B cross border. So cross border is growing and mostly helped by AI because of automatic translations and translated listings.
Number two, if you’re marketplace, you can use AI to actually massively simplify your listings. So it used to be the listing process was. You would take a whole bunch of photos, then you would write a description, then you would write a title, you would pick a category, and you would select a price. All that is a lot of work, and frankly, you are not necessarily in the best position to understand what is the correct price for this item.
With AI now, if you have enough data, you can take a few photos of an item, and boom, the AI should say, oh, this is the item, this is the category, the title, this is the price, and boom, you’re done, so. So many of our marketplaces have done this. So we’re investors in Rebag, which is a handbag marketplace. And they have this AI called Clair that will do image recognition and let you sell your handbag.
We’re in a company called Collx, which is doing it for trading cards. So you take photos of your trading cards, it immediately identifies the highest value ones and helps you sell them. So this is happening vertical by vertical, completely improving the listing process. number three, improving the quality of the listings.
So we’re investors, a company called PhotoRoom. So PhotoRoom, you take photos of the items, it identify the type of item it is, where it’s most likely to sell and we’ll change the background automatically to increase the sell through rate. So in some cases it’s just creating white background, but some cases putting it in front of niche or whatever.
So massive again, mega trend of using AI to not only simplify the listing process, but improve the listings. And then last but not least, everyone, should be using AI to improve productivity internally. So everyone now, all the programmers are using Cursor, GitHub, Copilot, and improving programmer productivity.
A lot of companies are using customer service AI to like help deal with returns, answer most basic questions. Frankly, in a cheaper, faster, better way than if you’re talking to actual agents. And people were also using it for sales and marketing. Other thing, as I said, I would be using it for is I would probably add a an item recommendation engine to every marketplace.
So in addition to having your core search engine, I would have like, “Hey, what is the best item I should be buying?” So I’d probably add this as a E that I didn’t put in the slide, but it makes a lot of sense. And as I said, I would integrate, integrate yourself and index yourself on the LLMs.
Other cool trends that we’re seeing out there that are not related to marketplaces, but like that are finally coming of age. So, one is Humanoid Robotics. So this is, a company that we’re investors in, called Figure. And, many people ask me, wait a minute, why would you want Humanoid Robots when you can create specialized robots for things?
Well, there’s tens of trillions of infrastructure that has been invested. And for humans to operate in right warehouses, et cetera. And so you can either rebuild all this infrastructure for non-humanoid robots, or you can just build humanoid robots. And so Figure has created like an amazing AI and amazing robots with like amazing hand dexterity where it can actually replace humans and like, you know, last mile picking and packing warehouses, and they’re being deployed in the South Carolina BMW line, and replacing much more expensive machinists that are working because the robots are working like 20 hours a day.
And so we’re at the eve of this revolution. It’s only just now beginning, but I think they have orders. Maybe it’s just an LOI for like, I think they’ve mentioned like a hundred thousand robots. So I can see a world in the not too distant future, like four or five years in the future where millions of these robots will be, in the workplace.
especially in things like last mile, like last mile picking, packing warehouses. and in, And frankly, starting to become in the home. Now they’re still too expensive. They’re not super functional yet to be in the home. I think we’re going to start seeing the first version. So this is Figure 2, when Figure 3 comes out and maybe a year, I don’t have exact timelines, but I suspect based on the speed at which these things are happening.
And probably more common when Figure 4 comes out again, this is speculation. I don’t have any information on that, but I’d say three, four years down the line. Yeah. And so I can see a world five years down the line, and again, it’s going to follow the typical form. The first few ones are not going to work really well, they’re going to be very expensive, they’re going to be adopted by, wealthy families, and as the costs go down, ultimately it’ll be the Jetsons.
And everyone will have one, and it’ll be massively deflationary, it’ll improve massively your quality of life, it’ll give you free time to do all the things you like to do. And now if you like cooking, cook, but if you don’t like cooking, you can cook for you, go buy the groceries and it’ll free humans to be able to do the things that they love.
And we’re the very eve of that of that trend. And companies like Figure, and by the way, Figure is not the only one here. We have a bunch of Chinese companies and obviously you have Tesla with Optimus that are in the race, that are going to be transforming this race. But just putting it out there that robots are coming.
It’s early, but it is a megatrend.
Next thing is the real use case or killer application of crypto or stable coins. Now, of course, the biggest asset in crypto right now is Bitcoin, which has a use case as digital gold. And clearly, if you’re living in emerging markets, and you don’t have an amazing savings process, and you’re living in inflationary regimes or regimes that haven’t had a tendency to confiscate your assets, you know, think of Venezuela, I’m thinking Zimbabwe, I’m thinking Argentina.
Saving in Bitcoin makes a lot of sense. And the fact that it’s built in deflationary makes sense, but actually stable coins are the real killer app. In these countries where you need a means of exchange, Bitcoin is not a great way to pay because it’s deflationary. And by actual stable coins where you can be, and they can be dollars or other currencies, but for the most part with USDC, USDT, they’re in dollars, and amazingly, not only to save, but actually to transact, to pay, etc.
And again, the main use case has been emerging markets, but they are coming of age. And now with the introduction, with the circle going public, I can see a world where USDC becomes more acceptable as a means of paying even in the West as well. And now you’re seeing that the Stablecoin transaction volumes have reached and surpassed, like, the Visa transaction volumes.
And it won’t be very long, but where Stablecoins will have a higher transaction volumes than Visa, MasterCard, AmeriCorps. So the big mega trend is like the emergence of Stablecoins. And in fact, I’ve created a stable coin company myself called Midas, which is allowing you to basically generate yield on your stable coins and it’s doing very well.
We’ve grown from like zero to 150 million TVL, but like no marketing, no KOL, no points just by having amazing user interface and great product market fit and actually in a fully regulated, legal bankruptcy remote way. But we’re at the very beginning of the stable coin revolution.
Stable coins will come much more common in the years to come. Other thing probably worth mentioning, and this is really playing out in Ukraine is like the future of warfare is really drones and seeing companies like Endurall that have, like, gotten a lot of, press in the US and they’re doing very well and we’re investors, but I think the more interesting ones are actually the ones that are coming out of Ukraine, which are at much lower cost per drone and the cost per kill, which is the only way they’re going to, you know, the West read large in the future is going to be able to compete with the massive manufacturing advantage that China has if we ever are in a hot war, and I hope we never get a hot war, but currently we’re clearly in a Cold War II between Russia, Iran, North Korea and China on the one hand, and then the West on the other hand, and they have a massive, because of China, manufacturing advantage.
And so I think the only way we win is by out competing them and being more efficient and more cost effective. And right now, it’s playing out that way, at least in the Ukrainian conflict, where Ukraine is using these inexpensive drones, very effectively. And there were a few examples of that, where both the airfield was bombed by, like, having a truck, deployed drones in Russia.
The drones flew and destroyed all the planes, and one of the drones also took out a big bridge, I think a year ago or two years ago at this point, a jet ski type drone that cost, I think like 20k, sank a 150 million battleship. So, the future of warfare is definitely drones. And I think the vision I have here, that I can see for Ukraine on a go forward basis is maybe it can become, once the war and the conflict is over, the magic, the defense manufacturing base for the West to arm Taiwan and frankly the U.S. and the rest of the West, writ large, in case we need to, again, I’d rather carry a big stick and not have to use it, to be ready for conflict. Because I think right now, militaries, writ large, have always prepared for the last war, so we’re still investing in overly expensive aircraft carriers, manned planes, tanks, et cetera, that are completely ineffective and would be totally ineffective.
And so this is definitely the future of warfare, where you want to minimize cost per kill, and keep these things highly efficient. And I guess last but not least, you know, people have been talking about self-driving for a very, very long time. And self-driving kind of worked already a decade ago, but for a number of cultural reasons, technological edge cases, and regulatory reasons, it didn’t come to the fore.
But actually it is finally here. the way most surpassed, lift in market share that it was as of November 24 in San Francisco, I think it’s expected to be bigger than Uber by the end of this year and it’s happening in LA, it’s happening now in partnership with Uber in Austin, but self-driving is here, it’s here to say, and so the future is definitely self-driving electric and it’s finally come of age now, it’s going to take a while for whatever, the New York Taxi and Limousine Commission to accept it. It’s going to take a while for the costs and the pricing to decline such that it’ll start displacing, many more drivers, et cetera. So people have often worried, “Oh, this technology revolution is different. All the jobs could disappear.”
These things happen slowly. And this is the best example of that, right? Like people were worried that we’d have the top job category in the US about truck drivers, think about 3.7 million truck drivers. The replacement is going to take 20 years, 30 years. It’s not going to happen overnight. These people will all find jobs.
And that’s true, frankly, of every other AI disrupted category. These things take a lot longer than people expect. But the good news is, it’s finally here. And by the way, if you go to China, drone delivery is also finally here. The U.S. made very stupid decisions in the early days where they were like, you needed, the FAA said basically you needed a pilot, you needed a line of sight, you needed to buy air rights to fly over things, so basically making drone delivery essentially illegal and impossible in the U.S. In China, you have actually in Shenzhen, which is the highest density, one of the highest density cities in the world, you have drone food delivery. There’s been like over a million drone food delivery. So they go, the drones, or they deliver not to your apartment window, to be clear, they deliver to like a delivery point, a block or two from where you are, and then you go pick it up there.
But it’s like common. And if you can make it work in Shenzhen, you can make it work everywhere. So this is not just for rural areas where you can do drone delivery. So drone deliveries are here. not in the U.S. yet, but I’m hoping that we change the way we regulate the airspace to make it legal because the drones are capable and it’s going to lower the cost of delivery dramatically, you know, further allowing this revolution to happen.
For me, tech writ large, AI, the internet, it’s always about doing things better, cheaper, faster. And I want to do that, increase people’s purchasing power, to increase people’s productivity, to allow them to do more, to have more free time, to have a better quality of life, and actually to be able to pursue what are the things they love to do, whatever those things may be.
And so, I can’t wait for a world where, you know, you no longer need to own a car, and you can just have, like, these self-driving cars that are, that are taking you, that are taking you everywhere at a much lower cost per mile than marginal cost per mile because they’re electric. And self-driving where all the things you’re ordering are delivered for you for a much lower price because they’re delivered by autonomous vehicles and drones.
So the future is starting to be here. and I’m extraordinarily excited for, for what is to come. So that’s kind of it. trends from the first half of the year. I mean, a lot has happened. A lot is happening. let me know if you have any questions. I’m surprised actually, but I see a fair amount of viewers, but this has been one of the few ones where I haven’t seen very many questions.
So let me know if you have any comments, questions, et cetera. And if not, I will just wrap up and, see you in the next one. I’m not as usual. I will be creating a transcript for this and, posting it alongside with a deck on the next blog post, which is coming on next Tuesday, which will be the summary of the episode.
Okay. LinkedIn user, I have a question. I’ve been thinking for a while in the era where AI technologies make it easier than ever to produce hyper realistic, but false imagery, deep content, fabricated reviews. IDC Marketplace is evolving the strategies to safeguard customer trust, especially when authenticity, verification, brand credibility, and visual integrity are so central to the buying decision.
Systems or innovations you believe will be critical to ensuring the marketplace remain reliable sources and digital environment increasingly flooded by manipulated content.
This is by the way, true writ large, right? It’s true not just in marketplaces where you can have fake reviews, but it’s also true and social media and ads like the deep fakes of people saying and creating fake news if you want.
like all these things, it’s an ever ending, arms race. between, the, the bad guys and good guys and, and that never ending arms race, happened and whatever viruses and antivirus companies. So that the hackers versus the white Knights.
And it’s happening here. We’re seeing a lot of companies coming out with deep faked detection technology. And so, more and more companies are going to start implementing these basically safeguards where they’re validating that the review is actually posted with someone whose identity has been verified that that kind of content appears as authentic. So we’re seeing right now a trend. We’re being pitched countless deep fake detection companies. Whether it’s for media, whether it’s for marketplaces, et cetera. I’m sure some of them will emerge will become very good. And so the ever ending like deep fakes become better versus companies attacking them ever becoming better.
That war will keep fighting. But I suspect that as per usual, that the good guys will always remain, a step ahead. And you’re going to be able to figure it out. I mean, right now, despite the massive improvement in quality, you know, you can You can still at least get a sense for like, Oh, these reviews, like, like, don’t have heart.
They don’t feel like they were written by real humans, et cetera. But yes, is it going to be an ongoing challenge the same way than the early days, the ongoing challenge was preventing people trying to scam you, right? Like in trying to sell you things, and there were infinite scams at early marketplaces.
Like Craigslist is probably one of the better examples of that, where you had infinite scams there, and people found ways to deal with it.
May I be interested in Kashmir and Tanzania? maybe. I don’t know how well you’re doing, but I’m happy to take a look at it. just send me the deck, and send me a LinkedIn message with enough information about traction, union economics.
Deal terms, et cetera, and we’ll review it now. We don’t invest much in Africa. We have done some, mostly I have to admit in Nigeria and Kenya and South Africa, the bigger markets, but we’ve actually had good success. We’ve had a very big exit in Algeria, and we’ve had a fair amount of exits and success in the Middle East as well.
So, We’re happy to look at it.
Okay. I think, this show has come to an end. So thank you for tuning in and, I’ll see you in the next one. Have a great summer. If I don’t see you before then.