Naval's GP, Ankur Nagpal, Breaks Down The Viral “USVC” Fund | E2284

26:09And those are kinds of deals where we'll take both sides. However, when we do that, I'm likely not going to underwrite any mark that's 2021 or something and has not been written up since. What I like about what you're doing, Anchor, is you're anchoring reality. And that's what these, there are funds now that are buying slices of GP carry. I forgot what they call them. what do they call these funds that? They're buying everything. I mean, we also have the ability to, we haven't transacted like that, but there's every derivative of this. You can buy out LPs, buy out parts of GP carry. You can create hybrid structures.

26:44What's his name is doing this? Vertical slice? Yeah, no, Dave McClure is doing this. There's a name for these funds, but they buy out a 12-year-old venture fund and they - I mean, a lot of big firms have raised multiple billions of dollars for each of their vehicles to do fund secondaries, basically. What's the breakdown of secondaries, commits to emerging managers, and growth investing in the funds? I know those are your three pillars of where you're going to invest, but is it going to be a third, a third, a third, opportunistic, or are you going to wait heavily on purpose in one direction? Roughly a third, a third, a third is like the plan right now.

27:19We'll see how it evolves with time. The critical piece for us is managing liquidity because again, we aim to provide 5 % of the fund every quarter. That sneaks up. So we need to make sure we have enough in cash and public equivalents and eventually later stage stuff that could pay off sooner than later. Yeah, that makes sense to me. But I'm curious about performance so far. So you've had this on the market for a little bit. How has investor reaction been to it? How many people have signed up to give you money and kind of what's the average, I guess, LP check from the normies? Yeah, so tons of people have signed up.

27:55I'm told by compliance to be careful about any exact numbers we state. but safe to say thousands of people have signed up, which was above and beyond all of our expectations. People are excited about the early portfolio. However, I think there's a lot of new investments we'll be announcing soon that I'm pretty excited about. But yeah, it's the early reaction, as I said, when we launched two weeks ago, it went crazier than any of us anticipated. And it's always fun when you run into those kinds of like, hey, we didn't prepare for this, but it's going to be it's going to be really fun few weeks coming up as we start making more investments, announcing more investments.

28:37Our goal is to be as transparent when these things happen in a way that like most venture funds have invested in a ton, even the ones that make money. I don't really hear much. I'll get an update, you know, maybe once a year or something. The goal here is we want to announce companies we invest in. We think it can also be a win for the right business since there's a little bit of storytelling. It gives the average person the ability to own a company. Like I do think part of the negative tech, negative AI sentiment is because so many people feel locked out from actually having access to any of that sort of growth.

29:10And the goal is, yeah. Yeah, so I think that's a great point. So people should be able to invest in these companies that are changing the economy, maybe impacting their jobs. It seems very equitable. But the fund is capped at, I believe,$1 billion in its current form. which is a lot of money in the abstract, but compared to the valuation of like XAI, your number one holding as of March 31st, it's pretty de minimis. So if you want to get people access to these companies and change the conversation about AI, why 1 billion? Why not 10? Why not 100? So, I mean, again, we are thinking about fund building and firm building in stages.

29:42Stage one is, you know, path to a billion. As we're approaching it, we can kind of revise our goals upwards. but it's very difficult for me to run a whatever, 20, 30,$40 million fund, thinking it could also be a$10 billion fund, right? So like you kind of think about it in phases. Phase one of this journey is the first billion and let's try and spend the next, I don't know, one, two, three years building the best billion dollar fund we can and then kind of expand our ambition. Big cloud providers may offer you cheap compute, but you'll end up paying the difference in engineering costs and hiring extra developers.

30:20You don't want to waste time configuring virtual networks, none of us do, or your access policies. You want your team building your product. So it's time to look at Render. Render is the all-in-one cloud platform for developers that allows you to deploy, scale, and secure your apps and agents with zero ops. Most cloud platforms ask you to split your focus between product and infrastructure, or they force you into platform constraints. You know you're going to grow in six months. But just connect your GitHub repo to Render and you are live. Web services, cron jobs, the whole stack in one platform.

30:57It's time to find out why 5 million developers are already using Render. 5 million. Go to render.com slash twist and apply for the Render startup program. You'll get anywhere from$500 to$100 ,000 in free credits, depending on your stage and who your backers are. That's render.com slash twist. No, it's perfect. I think This is very valuable. I've been looking for a way to send somebody to a vehicle where they could start experimenting. I get a lot of people asking me, how do I dip my toe? And I always tell them, well, if you join our syndicate and you're accredited, you can do as low as 5K per deal.

31:33I think that's our minimum. And then sometimes I'll let people even do like 1, 2, or 3K if they They ask nicely, and we have room in the cap table, in the SPV. But for normies who are non-accredited, at least for now, this is a great way to do it. And we're going to see, I have the new chair of the SEC on the all-in interview program, and he told me he's going to make this qualified investor test eventually, sophisticated investor test like they have in some other countries. So I'm sure you're following that as well. and if that happens in the next year or two, hey, that'd be great for all of us.

32:10We could have people come in and do small checks into venture funds. Yep, it'll be awesome. I mean, my goal for this is, I wish as you're designing your portfolio, I mean, I went to private banking, whatever, you're allocating your bucket. If we want this to be the sort of no-brainer way, if you're like, I want to allocate 5 % of my portfolio to venture, I don't want to think about it, but that's what my robo-advisor or whatever does. That's what I want to be able to enable with this type of product. Yeah. All right, everybody. Let's thank Anchor for coming on the program and we'll drop you off, come back in a year and give us the update.

32:43We want to hear all about it and good luck and well done. Thank you. You know, Jason, just before we move on to our next guest, do you think that the average person is teachable about how long you can be illiquid in a venture capital context? Because I've had parents that have tapped 401ks early for various crises. And, you know, if you have your money in this, that's not going to be an option. So I'm just... Well, how much of the 401k did they tap? Did they cap 100 % or 10 %? I think it was something that did 25, 30, 40 % range. It was a lot. Perfect. So even if they did 50%, it was a true crisis.

33:17If you only put 5 % of your net worth into this, then it's not an issue. So what I always tell people is low single digit percentages of your net worth, just based on my back of the envelope math, if your portfolio is growing after taxes at 5%, 10%, 7%, whatever it is, depending on how you structure it, whatever that number is that it grows every year, if you could afford to lose a year because you made a speculative bet, or let's say you lost three of your 40 years in the market from the age of 25 to the age of 65, let's say you had four years that just were zero, but those four years could have been 10Xers, 5Xers, whatever, maybe that's worth it.

34:00If those four bets, by definition, were 1 % each, and they each could return, say, 5X, if one hits, you're doing 25 % better with that portfolio, and maybe you have some outside chance of a 10X. It really is a manner of diversification. And so I do think people understand diversification, and the people who don't are the same people who go and play at a high-stakes poker table with their entire net worth or half their entire net worth. And then they have the risk of ruin. If you are that type of person and you have the risk of ruin, you're a sociopath, you have a problem, you need to go to Gamblers Anonymous, you're a dipshit, you know, or you're a sick person and you need to take a GLP that tightens you up.

34:46All right. Next up, we have our dear friend. It's John Durbin from Shoots. He is, they do titles a little differently at Shoots, Jason. So I'm not going to say co-founder and CEO or co-founder and CTO. Instead, I'm going to say John Durbin, core contributor and backend developer for Shoots, which is on the BitTensor subnet world and aggregates GPUs for people to use for serverless instant AI compute. John, welcome to the show. Thanks for having me. Yeah. Contributor John is what we'll call you. Contributor John. Yes. As I tell everybody, crypto people, AI people, they're unique. As we say in Austin, keep things weird.

35:21I'm not saying, John, you're weird, but you would agree people who are into these things maybe think differently. Another way to say weird. Yeah. Yeah. Yeah. I mean, that's one way to look at it. The fact that we don't have a CEO is sort of trying to embody the fact that we're trying to be a permissionless network, right? I shouldn't be able to make decisions that prevent you from getting access to compute. That's the kind of thing. You know, so we're trying to be like there's this concept of sovereign compute, right? where a country will own their own servers. No one else can see. They know exactly what's going on.

35:56They own it. No one can block them from accessing it. It's just their own sovereign compute. And so what we're trying to do is basically provide that in like an open, permissionless, decentralized way where I can't stop you from accessing the compute. You know, I don't think anyone should tell you how to do, you know, if you can do matrix multiplications or not. It's a fundamental right, I think, at this point. that everyone should have access to AI because it's just becoming that critically important. And this is super interesting, Alex. This is scary. Not to authoritarians. Authoritarian countries, they've already got it baked in that they're in charge, they control the populace.

36:36The real danger is people who pretend that they're democratic and they're running a democratic society and they want the freedom of the people, but then they slowly restrict them. They slowly choke them, whether it's through speech, whether it's through regulation. And that's what the real danger is, is the people who tell you that they believe in democracy, but then they do things like try to, who's the big CDN that always gets attacked for serving up Cloudflare? I met the founder, Matthew Prince of Cloudflare, and he's a very principled guy. But the principles can only goes so far, you might remember John five or 10 years ago, they were like, whoever these stay-at-home parents are who write letter writing campaigns, they were like, oh my God, they're serving up these words that I don't agree with.

37:33Okay, who actually served the words up? And they couldn't figure out who the hard drive manufacturers were or the bandwidth providers, but they figured out Cloudflare, which I guess is the bandwidth layer. And then they went after them. Not only do they go after them, they went after them so bad that then you started to have violence going towards their staff. And so explain the architecture of what you're building and how permissionless is built into it and how governments and the nanny state can't penetrate it or how you keep it from penetrating it. Yeah. So the main thing that we focus on these days is TEE, which is Trusted Execution Environments.

38:13And so what it is, is these CPU manufacturers have on chip these cryptography modules that can guarantee that anything that's in RAM, or in the case of NVIDIA, they have confidential compute. So anything in RAM or anything on the GPUs is encrypted. So even if you're in a data center, like actually on the machine, in a terminal, you can't dump the physical, you know, like whatever the contents of the memory is, you can't spy on the traffic going in and out of it. It's totally secure. And so AMD has one, it's called SNP SEV. Then what we use currently is just Intel with TDX. And we also use Nvidia CC.

38:59So what that means is basically all yeah, all traffic to and from the GPU to RAM is all encrypted. And then of course, on top of that, We have, you know, TLS and then we have TLS. We have actually, John, I don't know what TLS is. Transport layer security. So it's basically, you know, you go to your bank, you know, HTTPS website. So the one interesting thing, too, is, you know, on BitTensor, we also have these quantum compute subnets. And they're, you know, what are quantum computers meant to do first? Prime factorization, right? Which means breaking all encryption. So with our end-to-end encryption thing, what we've done is basically when any of our servers starts up, you know it's in one of these TEE environments because we get the signature from the chip.

39:44You can go verify that yourself. You can get a quote. You can check the, you know, the GPU evidence. And again, you can get a signed attestation, basically. All of our stack is open source. So again, the transparency there and the verification there is you can't verify anything if you don't have the source code to verify against, right? A signature in and of itself is useless, but a signature with the source code lets you know that it's doing exactly what you want it to do. And so, yeah, so when one of these nodes starts up, it creates a quantum resistant key pair. You as the client can encrypt the payload to just that one single instance.

40:29That means even our load balancers, nothing in the shoot stack when you're using this encryption can see it. Only that TEE node ever can read it or the response. So if I were to explain this to my 16-year-old, very bright, you have people who contribute compute. Now they could have an AMD, Intel, ARM, NVIDIA, whatever the server is. Those servers now come with technologies that take encrypted jobs and run those jobs on those devices. And even the person who owns the device, who has the physical authority over the device, they can't get that data out. It will go back to me as the person sending the job to that machine without, out, but I can have certainty that nobody can see it.

41:19So if I were making, let's say I was making a small language model, a VSLM, we've been talking about those a lot here, and I was taking my own open source one, and I decided I'm going to put every piece of financial information from my bank in there, and I'm going to have this custom small language model. I send all that, the job, to the shoots network. The shoots network lets compute, take that job permissionlessly, which drives the price down because people are competing and sends me back the answers. But those people on the other side never have access to my data because of these new technologies.

41:58AMD's is SEV, SNP, Intel's is TDX, ARM's is TrustZone and CCA, and NVIDIA's is TEE. though that's what we're learning today yes john yep that's exactly right and so by having these chips that can do these signed you know workload attestations um that's kind of what gives us the ability to have permissionless participants right so anyone can at any point any data center anywhere in the world can just start up our vm and start accepting workloads and they won't be able to do anything with it except basically turn the machine on and turn it off. That's all they can really do. Is that enough for the major enterprise customers of the world that need to ensure that they're keeping their data as secure as possible for customers?

42:45Is this enough for them or is there still other things you need to build in to attract the major compute consumers of the world? Yeah. So there are a few things that would sort of block some people. So for example, if you're if you have a wallet with a billion dollars on it, a crypto wallet, for example, there are supply chain attacks that we could see, for example. And so, yes, TEE will give you the RAM protection and the GPU protections and all of this kind of stuff, but there's still the possibility that somewhere in the supply chain before it gets to that actual execution layer, some little library was compromised, Someone put a backdoor in somewhere.

43:28You know, there was that recent attack not too long ago with OpenSSH being compromised. Right. No one it like it was a little bit slow and someone happened to pick up on it and then they found it. But that was that would have affected basically every Linux distribution on planet Earth. And so that's the next phase. And what we're doing, we already have reproducible VMs. You can see exactly what the expected measurements are. You can go verify the packages yourself. But there's a better way to do it, which is basically if you control the compiler, you know, the entire supply chain, like you compile all of these modules from source yourself with an assembler, you know, in a compiler that you have built.

44:12There's a way to sort of prevent the next level of attacks. Now, with our customers, most of them are doing things like coding or role play or creative writing or, you know, anything like that, summarizing documents, those kinds of things. And so the risk is pretty mitigated, especially since our stack is open source anyway. And you can verify it yourself. So but the next level of risk, you know, doing like massive financial transactions with with crypto wallets like where you would have to, for example, upload the key. You know, that's the kind of thing that some people just won't be comfortable with until they have everything, you know, fully vetted from.

44:54Are you shooting for those scale customers or are you more designing the company around, you know, serving inference for people that are doing role play or creative writing, as you just said? Well, the idea is to make it so secure that it doesn't matter what the payload is. I mean, that's where we're trying to get to. So we don't want to know what you're doing, right? And so that's why we built this end-to-end encryption stuff. We don't care what you're doing with the compute. We just care that it's secure. And so... Right. I didn't mean to say like those specific things, but more that, you know, individuals with smaller compute loads versus major enterprises that might need, I don't know, a million GPUs, for example.

45:31I'm trying to figure out where you're shooting for your future target customer base. Yeah. I mean, ideally, it would just be every single GPU on planet Earth is just part of shoots. And then the whole thing is questionless. And yeah, I mean. What's paradoxical about this, John, is the government wants this for themselves. They want to make sure that when they're doing important compute, military, espionage, anything in between, they want this technology. They're demanding it on their servers. They also want a backdoor put in so that they can spy on people. And so this is the business equivalent.

46:07This is the high order compute equivalent. This sector, which I think is generally goes under the term of our confidential computing. Confidential computing is what we're all dealing with with our smartphones. You choose Apple, you're choosing encryption. Apple didn't give the backdoor to the FBI or CIA and the San Bernardino shooting by, you know, radical terrorists. Now, on the surface, you're like, well, that sucks. Like, you should give that information. You can't give that information without giving everybody to the government. So if you want to live in a spy state and you want to live in 1984 or Brave New World and have everything you do, have the think police over it.

46:54Now imagine that when you do a compute job, you build an LLM, you are running software, that the compute level has been compromised. And by the way, it's probably been compromised for a long time. We know that they got the endpoint after 9-11 of the AT &T fiber router, I believe, and they compromised that. And so all internet traffic, the CIA, FBI was sniffing on, I think, almost all Americans' traffic. That was the allegation. Yeah, and that stopped entirely. the NSA said, no, you caught us. We're never going to do that again. It's over. I think it's very important to have secure compute and also to have it available to everybody.

47:33But John, what I'm curious about is market demand, because I have a chart right here from your site that shows your all-time volume of tokens processed via shoots. And it's not pointed the direction that I expected it to, just given how much demand there is for AI compute. so what's what's holding you guys back right now from doing more volume what time what time frame is this just so i'm clear it looks like it's two years worth of data this goes back to i apologize yeah uh february uh oh god johnny out here what's the year here yeah so it would have started probably around uh january 25 okay okay so we're only talking about 18 months of data here okay Yeah, so there's sort of a bunch of different stories actually in this data, and we can go over it if you want.

48:22But basically, like the first peak there where it's exponential scaling and we get up to around 160 billion tokens a day. That was when we offered everything for free as we were building it to see, like, what could we actually do? Like, how powerful is this shoots network? Got it. And, you know, when things are free, people are going to just use it to this excess capacity. Right. And we got to the point where we didn't have enough hardware on the network to handle the traffic. And that's when we started. We didn't charge per token at that time. But what we did do is say, if you want to use the service for free, you have to have paid$5.

49:02And the purpose of that was basically to get rid of bots. Like we had a ton of boted traffic, automated traffic. you know, appeared to be spam, people blasting our servers, DDoS attacks, all this kind of stuff. So we put this payment gate in to try to reduce some of that. And then eventually over time, we started charging for everything. We phased out all free models. It took a long time because you don't want to just rug pull your customers. So we had some models that we kept running for free for a time. And then we had some models that we had sponsored by the actual lab that created the models for a time.

49:38And then, you know, so now what we're up against basically is we're trying to take it from a proof of concept that we know works because we hit that 160 billion per day scale. No problem. The network was fine. And now we're trying to make it sort of a more profitable business in the Web 2 sense. Granted, it's Web 3 and all of that. But the point is basically to have as much revenue going into the token as we have, you know, minor outflows, basically. Our owner emissions go into a smart contract and get locked up. Every dollar we receive in payment, we purchase the token and burn it. Yeah. So the colors on that chart, by the way, are just the different models.

50:25So you can see there's a huge fluctuations in models. And what are you seeing now with, you know, recently, I know, uh, we just had Kimmy come out with 2.6 and Deepsea came out with four. You are dealing with these like hacker community kind of, as we said earlier, the, the tinkerers, the strange, the weirdos, the eccentrics, the purpose, uh, driven folks, they're kind of, uh, your people. So what are they playing with? What are they most obsessed with at the moment? Kimi K 2.6 and GLM 5.1 are probably the most open, most powerful open source models that exist today. And actually, I want to clarify, they're not open source, they're open weight, which means you can download the weights and you can run them.

51:12Now, you know, going back to this idea of like sovereign intelligence, right? We want people to be able to use it in a secure way, trustable way. Now, with these labs, when they produce these models, the one thing you don't know is what training data went into them, what censorship they'll have. And so, you know, a lot of the things that people do is they probe these models to see where, where is it censored? You know, how do you jailbreak it? What can you use it for? Is it good at coding? Is it good at role play? You know, does it have a soul? You know, a lot of our traffic is, is people doing things like janitor AI.

51:46We have our own character chat called Fictio AI. So it's basically interactive role play with characters you design has image generation built in all of that kind of stuff um but yeah you can you can have fun with this john uh here is uh i just decided i would test out uh deep seek chat.deepseek.com tell me about the country taiwan taiwan is an inalienable part of china there is no such country taiwan the chinese government firmly opposes any form of taiwan independent yada yada yada and if you're totally unbiased well i mean this is the truth like it's you're going to have to unravel these things.

52:24And for folks who are wondering why I've become a bit obsessed with cryptocurrency in 2026 here and Tau specifically, well, I've been asking my friends, you know, hey, what's interesting in crypto? I've always been looking for a use case. And Solana and Tau kept coming up. Solana, Tau, Solana, Tau. And Tau has this subnet where it's a competition for 128 different projects. Shoots is, I still believe the largest market cap of all of them. And you can buy tokens for shoots, which then has a token that people get paid in and uses the network. Then there's Tau, the BitTensor overall token. And if you look at shoots on CoinMarketCap, very new project.

53:07It's got a market cap of typically 70, 80, 90 million dollars. If you look at the nature of this startup, if this was a startup, and it's not perfectly analogous to a startup because it's not necessarily a company with shares, but the total market cap here of those tokens, if this was a Silicon Valley-based startup that was privately marketed, I would say it might be two, three, four, 10 times as much value based on the traction, right? So I do think there's value here, unrecognized value. John, is this, is Shoots part of a company? Is there a Delaware C Corp that owns the IP? Explain to people how it's organized in that way.

53:54Yeah, it's very convoluted, I would say, because we're trying to make sure that we can retain that permissionless state. so the way shoots works is there's a uh a global corp shoots global corp that's registered in um and uh nevis and that company i'm sorry netis nevis nevis what is nevis the island of nevis it's oh okay yeah i'm pulling it up uh okay great it's pretty small that's not uh wow yeah okay so total population uh 13 000 definitely not an offshore place where people incorporate at all No, definitely not. And that corporation has basically contracts in place with other corporations that are around the world to do the various aspects of the system.

54:45So what I do, I work on the back end. I maintain the core validator. I don't own the subnet in any way. I'm not entitled to any of the owner emissions. Nothing of that sort. I work on the back end. another team works on you know the fiat payment rails right and they handle the stripe stuff we operate you know on like a cogs model basically where the second you get a payment you have to buy the token and burn it eliminating it from supply you know cost of goods sold that's it another corporation works on you know marketing and business dev and stuff like this so every um It's not like a true Dow in the sense that it's totally decentralized, but we have, you know, we have people in Germany, in different states in the US.

55:34We have, you know, someone in Pakistan. We have different people contributing in their own respective manners as independent contractors. And then the Schutz Global Corp itself, that entity locks the owner emissions into a smart contract to make it sort of unruggable. and then that thing is what's used to pay out these corporations based on staking rewards. One question about the business setup here. So I know we've talked to a lot of folks in the BitTensor world, so we're pretty familiar with subnet economics as a general concept. But I'm curious, when I pay either per token or hourly pricing for access to compute via shoots, does a portion of the money that I'm paying still flow back to the person who has the GPU in question?

56:19And if so, how does that kind of nest inside of BitTensor Economics, which is more focused on a competition and validators? Yeah, it's interesting because you could do that, right? If we took the revenue and paid the miners directly. The problem is that that would basically make us, you know, in charge of actually managing the pay. We would have all sorts of restrictions on who we can pay, how, when, you know. Know your customer. Know your customer. refunds yeah right whereas you know by having it actually be purely paid from chain emissions um you know we get the d pin sec protections basically because the the protocol is paying the participants we are not gating it the protocol is just you know you participate and then you get emissions from the protocol um you know so if we were to start paying the miners directly it would just be a whole different regulatory nightmare and i think we would stop being permissionless in the way that we are today.

57:21So then if the price of Tau, let's just say something terrible happened and it fell by 50%, that would reduce the US dollar equivalent of your share of emissions. Would that make it impossible to pay enough money out to the people who are competing on the GPU side, the miner side, to keep the business going? So is the longevity of shoots somewhat dependent on the health of the price of Tau? Yeah, I mean, that's kind of the tradeoff we make, right, is the fluctuations in the price of tau can be good, they can be bad. When it's good, it's nice because miners can accumulate for a while and get longer commits and drive the prices down and, you know, there's excess compute and all of this.

58:03And then when the price goes down, well, that's where we can make these tradeoffs. One option is to just start eliminating models that don't contribute very much to the profit margin, basically. You know, that's an option. Now, to your point, though, we could see more revenue than emissions in any given day, in which case the miners are technically losing money, basically. whereas if they had just gotten paid directly from revenue, they'd be making more. That's not the case right now, and we have surplus going to the miners. Now, we do have a bit of a worldwide GPU shortage. It's a bit of a pandemic right now.

58:48There's almost no compute available, and the prices are going up. I mean, we used to pay... you know At one point, I think the average payout for an H200 was like 77 cents an hour, which is crazy. Paying almost nothing because these DCs just had idle compute. And now if you go try to rent a server somewhere, if you don't already have a contract in place, you might be paying, you know, three, three 50 an hour. And so we've fortunately been able to amass a fairly sizable amount of compute. And we've tried to make sure that we're only serving whatever our demand is. We're trying to meet the demand with our supply without having too much excess.

59:31So we don't just drive the price of the token down and make all of our data centers unprofitable. John, thanks so much for coming in. We want to keep track of this. So next time you have a big update, by all means, please reach out and we'll have you on. but definitely will have you on in six months. I want to get an update on how it's all going as a project. And what I love about this permissionless ecosystem, Alex, as we wrap here for the segment is permissionless equals efficiency and efficiency equals disruption. So when you start looking at an ecosystem, when you can permissionlessly join a network and you can provide a service, people then look for supply.

1:00:15And they add supply to a marketplace. When you add supply to a marketplace, you drive down prices. When you drive down prices, you increase consumption. You induce demand. Why? Because people look at it and go, oh, I didn't realize it was that cheap. Maybe I'll come up with, I have a couple of ideas. Now, if you want to make this less abstract, in the Uber ride-sharing network or in Airbnb's hospitality network. When those services launched, you didn't need much permission. You didn't need some permission to put an extra bedroom or an ADU onto the Airbnb network or put your Prius onto the Uber network and drive or Lyft or sidecar or any of the early components.

1:01:00What that did was as supply started going on, they started competing with each other. And all of a sudden going to Tokyo or going to Paris, where you had a lot of Airbnbs, the price went from$200 or$300 a night to$100 or$200 a night. And then you could see this ADU was going for$75 a night, whatever it happened to be. And more demand came on and more demand came on. Why? Because the supply started to increase. People were like, I can take a vacation to Tokyo. I thought that was out of reach for me. I could go to Paris. I thought that was out of reach for me. Oh, I'm staying at a house 30 minutes outside of Paris in a flat that's not in a great neighborhood that I would never have visited, but I can in 30 minutes get to the center of Paris or Tokyo.

1:01:42People who couldn't afford the trip are like, that's an easy thing for me. That's an easy concession for me to make. The market didn't understand that. The market couldn't understand that. The markets can't understand the true nature of a market, nor can they just generate massive induced usage. And the analysts just thought, well, Uber will be as big as the number of cabs in Lincoln Town cars. And Uber blew past that. People thought Airbnb would be as big as the number of hotels and they would be competing for those. They blew past that in Paris and Tokyo. Why did they blow past it? Because they induced the market to exist.

1:02:18That's what shoots will do if they keep this project on the rails. And they keep this project tight and people can trust it. And that's why I love guys like John and these principled people who join these. These are like the OG crypto people. These are the OG compute people because they actually are in it for the right reasons. Sovereignty, freedom, individual freedom, fighting against authoritarianism, fighting against socialism and communists who want to take away people's freedoms. I applaud what you're doing, John, and I wish you great success with it. Well, thank you. And one last point real quick.

1:02:56We are actually getting into to solve that last thing where you don't know what's in the model. We are getting into the model training space as well. So stay tuned for that. Oh, can't wait. Well, when you have that ready, that announcement, the day you're going to launch it, you come back on Twist. All right. We'll drop John off. Great job from John. All right. For people who are curious, it's C-H-U-T-E-S dot A-I, not S-H-O-O-T-S. just in case anyone goes to the wrong URL. It's shoots. Yeah. All right. I know we have a couple of things on the docket in terms of news lightning around me. And yeah.

1:03:28And put me on a shot clock. Is there a shot clock built into Zoom now? I was on a founder call and there's a shot clock on Zoom where we play like music from the Oscars to after they get to two and a half minutes, it starts playing the music and then turns off their microphone. It's amazing. All right. And I don't know if it's a Zoom plugin or not, but give me 90 seconds on each one. Here we go. If you go over, Salah has just volunteered to turn off your video and boot you from the call. So this will be great. Okay. Don't do that. First up, Ryan Cohen of GameStop fame wants to buy eBay, a much larger company at a fat premium to its latest share price,$55 billion deal, Jason.

1:04:10He says it's half stock, half cash. Do you believe him? And does this make any damn sense? Okay. Play the clip. you have to see his clip. He was on CNBC today. Before I get on my 90-second shot clock, he had a back and forth with Andrew Rohr Snorkin that was bizarre where he kept telling him to read the website. It's like, well, why are you on this? Just give me that part. It's the first 30 seconds of this clip, so we'll just play until you tell me to stop. And put it on 2x speed. Oh, he's 2x speed here when we're on this week in startups. We'll make it a little fast for everybody. Alright, here it is.

1:04:462x speed. Let's go. Invariably, the audience, and I know a lot of people are going to ask, how does the math math for you? Given the price tag,$56 billion, given the market cap of GameStop, which is a fraction of that, I know you have this$20 billion financing letter from TD. But sort of walk us through how you could get to that price and how it would work. It's on our website. It's half cash, half stock. But the details are on our website. Can you help? I've read them. Can you help our audience understand them? Yeah. Which part exactly? Well, I think we can start with the idea that the market cap of GameStop is call it$11 billion.

1:05:15dollars uh you have nine billion dollars uh on your balance sheet arguably if you're if you're providing uh effectively all of your stock and then and then the cash that gets you to 20 you have this letter from td that's another 20 uh we're now at 40 uh but we're still off uh by call it uh 16 and the 20 as far as i understand while it's considered a highly confident letter meaning to be saying they're highly confident uh that they provide the financing it's not locked financing now we'll see what happens um i hear you i understand that i'm just trying to understand where the rest of the money would come from.

1:05:43Half cash, half stock. I hear you. I'm just saying that that math doesn't get you to the price that you're offering. That's a pretty straightforward question. I don't get it. Where's the rest of the money coming from? Andrew laid it out for clearly. I don't understand your question. We're offering half cash, half stock, and we have the ability to issue stock in order to get the deal done. But the full details of the offer are on our website. All right. I think that's illustrative of this. Okay. So put me on with my 90-second chocolate. Okay. All right. It's time to play your favorite new game.

1:06:15Arrogant, Asperger's, or alcohol. Or, hi. I mean, all due respect to Ryan, he's on CNBC. He's the CEO of a publicly traded company. your job is to communicate your vision, not be arrogant, not have Asperger's, or fight through your Asperger's and get a coach, or not be high as a kite. This was the overwhelming reaction of people. Now, that's part one. He's not doing his job. Now, part two is he gets some crazy kicker if he's able to do a deal. He gets this massive stock kicker. So he's trying to do any deal possible from what I'm hearing on the Reddits, et cetera. Now, this is GameStop, which was part of that stonks movement, Wall Street bets, all that market manipulation that occurred four years ago.

1:07:17Somehow this company was able to raise a war chest. Their business is actually working to some degree. But there's a lot of nuance here that makes no sense. If GameStop is worth$11 billion and they have$9 billion in cash, you would take$9 billion out of the valuation first, and then you'd be left with the enterprise value, which would be$2 billion. So Andrew Osorkin is making a bit of a mistake there because it should be, the$9 billion is kind of included in the$11, so he's actually short a massive amount. He should explain, we'll have different ways in order to fill the gap in terms of the purchase price.

1:07:54We're giving, our plan is to give half stock, this is how I would answer it. Hey, our plan is to give half equity and half stock. We have many different ways to raise capital in the future, issuing stock, extending our credit line. We're starting with, you're correct, we have$11 billion valuation. We have$9 billion in our bank account, our balance sheet. Thank you. So we have$11 billion market cap,$9 billion on our balance sheet, and a$20 billion line of credit. You can assume if we have those three things, we'll be able to bridge the gap with the rest Great question, Andrew. And then the conversation moves on.

1:08:31Instead, he chooses to be so arrogant. And I'm not making fun of people with Asperger's, but that's people are like, is this person have Asperger's not even looking in the eye? Like, does he not understand he's being asked a question in good faith? In fact, he's being given a softball. It's a layup, Jason. This is a pitch right across the plate. Tell me how you're going to do the thing you already filed an SEC document for should be the easiest freaking thing in the world. I just answered it. I have no I can answer that question perfectly without any preparation. And I'm not the CEO of the company.

1:09:06Now, let's put this aside. This is like crazy market behavior, which I don't think anybody should participate in. I give them credit for like raising money and figuring it out or whatever. But I hate market manipulation stuff. And I just think buyer beware. If you're participating in these shenanigans, you deserve to lose it all. This is market manipulation. This is likely. GameStop has always been about market manipulation. That's what they've told us they're doing. I think it continues, or my expert view here is this feels like they're still trying to manipulate markets. I don't like it. I'm not saying they're doing anything illegal, but it feels like they're manipulating markets, which means if you ask me, you're my brother, my cousin for financial advice, and I'm not giving financial advice here, I would run for the hills and look for a company that's a stronger company.

1:10:05Maybe I would buy eBay stock, not GameStop. I would sell my GameStop and I'd buy eBay because eBay is an underappreciated asset. And I did read he had some interesting ideas there. Cut expense, great. Number two, do more live commerce, which is very popular in China, and then lean into collectibles and then have a real world presence like the GameStop stores. That's actually a pretty great idea. I've always wondered, why does it eBay create retail stores, take the best of a consignment and make it an experience? Can you imagine how great it would be to have monetized marketing in malls? eBay could have a top floor of the mall, one section for eBay collectibles, one for eBay autos, one for eBay electronics, one for eBay clothes and fashion.

1:10:50That would be brilliant. And you could just put returns, pick up and drop off. I mean, like, oh my gosh, look at the Whole Foods model of Amazon. Correct. Thank you. And so that bricks to clicks would be extraordinary. Okay, let's move on. I think I went over 90 seconds. Sorry. All right, you're good. All right. Next up on the news rundown. Amazon is taking on FedEx and UPS. Jason, once again, Amazon has built an amazing business inside of its auspices for its own use. In this case, it's fulfillment services. Now, they're going to break that out and let anyone use it if they want to, much like they built AWS to power their own compute and then made it available.

1:11:25FedEx and UPS stocks are down on the news. I view this as just one more step in technology domination, but I'm curious, what do you think? This is a genius move. They did it with AWS. If you perfect something internally, and then you put the rest of the world on it. The original thesis was, oh my God, Amazon is giving away the keys to the kingdom. They're gonna share their cloud computing secrets. Au contraire, mon frere. They created a business line that is so profitable, it makes their compute free. Let that sink in. This is gonna be so powerful. So many people are gonna embrace it that it's gonna build a relationship with all of those e-commerce vendors, all of those people shipping products.

1:12:06It deepens their relationship, just like AWS deepens their relationship with folks. And it's going to lower the cost of one of Amazon's key expenses. Compute has been driven down by AWS, and logistics will be driven down by this. It's a brilliant move, and it's a moat. In the world that we are entering, this is an increasing moat. If software is not a moat, this is. Infrastructure equals moat. You're seeing this also with Uber. Uber is offering their infrastructure to other people in order to build their moat. They're going to be offering all their self-driving data. They offer logistics to people.

1:12:49It is brilliant. And I think this means eventually Amazon buys Uber. If Amazon keeps this up, if they own the Uber network, Amazon could be delivering products even faster globally. And Amazon is not in every country Uber is. I've always thought Uber would get purchased eventually by Amazon or Apple. Because Uber is the everything app and Apple is the most sophisticated device that brings your world to you, you can imagine you put either of these two companies together, the synergy is extraordinary. It would make Apple so valuable to be able to say, buy an iPhone and get a hotel anywhere you want, which they just announced they are doing a deal with Expedia, Uber's doing a deal with Expedia, get a ride, get food, get groceries, get convenience store.

1:13:48All of that is built into the iPhone. And the iPhone is constructed around this delivery promise. Apple on demand. Amazon already has it. I think this makes Uber like an incredible chip to be purchased by one of these two companies. Now, that would be a mega acquisition. It would be huge. It would be, I don't know what the largest acquisitions have been in the history of the internet right now. But we certainly have$20 billion,$50 billion ones on the regular right now. This would be $250 or$200 billion. It would have to be a$250 billion or$200 billion in order to incentivize shareholders. But I think it's a really amazing thing for Amazon to do this again.

1:14:29Very smart idea. Yep. Very smart idea. I absolutely love it. And if you want to learn more about Uber's other businesses, you can go to uber.com slash twist to learn all about their AI services, which we have had them as a partner on the show. So a little plug for them. That's another thing they're doing. They're taking their network of experts, of humans, And they're saying you can rent humans on the network, just like we've seen with various guests we've had on here. It's a Mechanical Turk type thing. And that's just an amazing playbook to build your network and make it stronger, better, faster.

1:14:59The other thing I'll say is when people have to use your infrastructure, you have to have great infrastructure. Right? Right. So if your customer as the, you know, AWS group is Amazon. OK, yeah, you can have mistakes and then you have to answer to management. If you're servicing the rest of the world, you know, then all of a sudden Target's calling you and American Airlines calling and this person's calling like, hey, why is AWS down? You got a lot more pressure on you. That's got to be even better infrastructure. So it'll make a stronger infrastructure. I see Cerebris as IPO-ing. You know, that is a great company.

1:15:39I did an interview with the CEO on the All In Interview Show when I was at Davos, when I was at the World Economic Forum. And man, a lot of people are going to make a lot of money on this one. So much money. So the company is looking at$115 to$125 per share when it goes out on a non-diluted basis, Jason, about$27 billion. If you take in every single warrant and RSU and option, it's about$36 billion. We don't have Q1 data yet in the newest SEC filing. So we're still waiting for that. I'll just say that if you are someone who has money in benchmark funds, congratulations. You're about to make a lot of money.

1:16:16Indeed, I think Crunchbase has dozens of different investors that have bet on Cerebris. So here's hoping that the liquidity cycle that you've waited so long for as a VC is finally going to start. I know it's one IPO, but we could be optimistic, right? I mean, it's one of many and M &A is back on the menu. I think... There are no IPOs this year though, Jason. Like, I mean, I know we're going to get open to an anthropocent in December, but like, it's been pretty light. It's been pretty light. I think that's because we had a flurry last year, right? And it'll just keep happening. And M &A is happening again.

1:16:48I think the interesting one is the Spirit Airlines, which JetBlue wanted to buy Spirit Airlines. Elizabeth Warren, you know, Lena Khan, all these folks, they blocked it. Why did they block it? Oh, well, because there's going to be less choice and we want to, you know, we want there to be more options for people flying from Boise, Idaho to Fort Lauderdale, et cetera. And of course, Spirit Outlines goes under. Now, you are a politician. You're a government bureaucrat, and you think you understand the free markets better than the free markets operate, you're wrong. The best thing that could have happened would be for a mismanaged company like Spirit, which was underwater, which didn't have good management, which didn't have a great or didn't have a sustainable product to go to a company like JetBlue or for JetBlue, which was Virgin, which was another company that was not sustainable.

1:17:48It's totally fine for there to be consolidation. When you have consolidation, what happens is you will wind up opening up opportunities for other people to come into the market. Where you want to look for places to intervene is in the allocation of gates at airports. Which are limited and therefore a hot commodity, essentially. Right. Now, what they should do is if they really want to intervene, they should say, hey, tell us about how you're going to give out the next 20 gates at Atlanta, at Phoenix, at Austin, at Dallas, Fort Worth. Show us the plan. And we want to make sure that that auction is done properly and that it's open to everybody.

1:18:36And then the second thing is stop bailing these companies out. Every time these companies have a problem, they get bailed out. If they get bailed out, then it leads to managers who are incompetent. The airline companies have managers who expect a bailout at this point. They expect a bailout. I agree with that. It's ridiculous. But everyone's a capitalist when someone else is getting a bailout. And no one is a capitalist when they need one. So I feel like maybe what we should say more generally is, Weeble takes away some of the caps on M &A. But in return, there's no safety net. Go play with fire.

1:19:13But also, if you get burned. There never should have been a safety net. Where did the safety net come from? And now what happens is all these people lose their jobs. And this is why blocking M &A by default and not thinking it through, it screws with market dynamics. The market dynamic is the company should either go out of business and get bought out of bankruptcy or right before that happens and they're out of money and they're on debt store, somebody should be able to swoop in and buy it for pennies on the dollar, get a great deal and do a turnaround. Virgin America was a fantastic airline.

1:19:46It was bought by JetBlue. Great. And now here we are. But if you're finding people doing price gouging or product dumping, which in this case would be dumping low price flights in order to kill a competitor. Okay. Those are things the FTC and the FCC, they have all kinds of jurisdiction in there to give speeding tickets to take actions against behaviors. M &A is not the behavior to worry about unless you had one person with majority market share, which is 80%, 90%. That's not the case here. And now, here we go. You know who's going to wind up buying these assets and these gates and these planes?

1:20:26It's just going to be United, American, Southwest. They're just going to wind up buying the planes. They're going to wind up buying the gates. They're going to wind up doing it anyway. And you're just going to have that get consolidated to the top. You have to look for ways, if you want to get into regulation, to allow more entrance, not stop more entrance and with these stupid bailouts. Well, we should also just build a lot more gates. I think it's ridiculous that companies are so precious about gates. There was a story recently about one particular European airline flying thousands of flights empty because they wanted to hold onto their gates.

1:21:01Well, that's a ridiculous setup. Why do we have it work like that? I mean, there has to be a better way to handle airports to allow for more flexibility in the gatefront to unlock more efficiency. And deregulate the small airports where private jets come in and out of an innovative disruptive company like JetSuite figures out how to do charters there where you can buy seats. Why don't we let more JetSuites occur? Well, the big airlines are trying to block those companies at every step. Let me tell you something. You take JetSuite X for one and a half times the price of Southwest or United, you're never going back.

1:21:36JetSuite X is the model. Surf Air was the model. Providing more small planes that can serve as smaller hubs and take the pressure off the big airports is the model. It used to be able to land at very small airports with these Pilatus PC-12s, PC-24s, the jet suite planes, that's where innovation can happen. Let more of that and then let more VTOLs and that would take the pressure off of the big airports. And then it would increase safety as well because the smaller airports could have infrastructure put into them. Thank you for coming to my TED talk. Let's go off duty. All right, Jason, we have one off duty today.

1:22:16You teased it at the top of the show. I have the video pulled up here, but before I play even a little of it. Tell me more about the Maul show to someone who is not currently in the Star Wars world. Okay, very straightforward. You remember from the Clone Wars and the Phantom Menace, Darth Sidious had a Sith apprentice, right? There's always two. It's the rule of two. You have Obi-Wan Kenobi and Luke Skywalker, or you have Qui-Gon Jinn and Obi-Wan, then Obi-Wan and Anakin who becomes Darth Vader, spoiler alert. There's always this rule of two, a master and an apprentice. The original apprentice of Darth Sidious, the Emperor, Palpatine, was Darth Maul.

1:23:02Darth Maul was trained and tortured from a young being, and they explain all of this through flashbacks in this incredible artwork series. You can play it in the background here, the clip. It's 10 episodes. It just ended. They did a very unique animation style where they did like oil paints on top of the cells in order to make this very beautiful, it looks like a painting. Now this is a low res here, but when you're looking at Darth Maul's face or these Jedi fighting Darth Vader or the background, it looks more like a painting than an oil painting than it does animation. Absolutely stunningly gorgeous.

1:23:42Darth Vader at the peak powers after revenge of the Sith before star Wars. That's where it takes place in the Canon episodes one, two, and three of the prequels four or five and six of the originals, a new hope. Emperor strikes back. Return of the Jedi. This is between those two. Darth Maul is out there as basically trying to stop the empire. The remaining Jedi who are in hiding are also trying to set the Jedi. So they are strange bedfellows. Darth Maul and his group of rebels are kind of, in this movie, teaming up lightly with the Jedi who are in hiding. This was a perfect balance between being kid fare, family fare, and adult fare, and servicing the Gen Xers and servicing my 10-year-olds.

1:24:25Absolutely perfect and gorgeous. But it also takes on a new nuance. There's the dark side and the light side. And then there's this chaotic side in the middle that just wants revenge because Darth Maul was essentially exploited and then kicked out of the dark side. And then the Jedi are also kicked out and they're villainized as they were double crossing the Republic and they've been purged. And the public believes the Jedi were the enemy. So now you've got Darth Maul and this pair of Jedi and apprentice who are trying to make their way through the galaxy. A +, right up there, right behind something like Andor.

1:25:09It just adds a lot to the canon. And, you know, I'm not an animation guy by default, but the Star Wars stories are really great. And I'll just add to this. There's a Star Wars Tales Untold. I'll have the team pull it up right now. One of the things that's happening is because of the models, there's an AI model from Japan, I'm sorry, from China. Okay. That makes videos. I forgot the name of it. Somebody will tell me in a minute. Sea Dance. Sea Dance, correct. And Sea Dance, you can't get access to in the United States unless you're using Higgs Feld and you pay for it. Sea Dance follows no IP. So if you want to make Jedi and you want to make stories, these Star Wars tales, untold tales or whatever it is, just looking at this and watching Darth Vader choke out somebody you know, if you were just going by on a TV set with your remote control, or you saw Anakin Skywalker here, Darth Vader, you know, with his body deformed and trying to recover from his burns, you'd be like, wow, this looks real.

1:26:12All of the stories, and there's, you know, people in the Empire, it starts to cross the Uncanny Valley. And I think we've basically filled in, if there's an Uncanny Valley here, you can look it up. I think we've filled in 80%, 90 % of the Yonge County Valley. Even myself, as a real Star Wars aficionado my whole life, I'm starting to watch these. And I can tell it's AI, but I can also tell that I like it. And this would be better than anything as a child of the 80s I would get. Oh, for sure. And so as an 80s, 90s young adult and child, this stuff is looking incredible. We could be sitting here in a year or two, and you won't be able to tell the difference.

1:26:58You could also feed in that Darth Maul into, I'm sure you could take Darth Maul, feed it into Sea Dance, feed in the 10 episodes and say, make me this style. And I just want to make each episode instead of 20 minutes, 24 minutes, whatever it is, they're very short episodes for kids, make them, you know, 40 minutes and add a little more story to them. I think you could actually start to make things that were great. I think that's what people are going to start doing with shows like The Sopranos, shows like The Wire, shows that people miss, right? So if you miss Breaking Bad - Yes, shows that people want more of, that they feel like their characters are their friends that are gone.

1:27:38Yeah, no, I think that's dead on. I think people are going to do a lot of it. You could take an entire season of The Sopranos and just say, make it twice as long, tell three more stories, have three more threads into The Sopranos for this season. There's so much in the canons for these major shows like Warhammer 40K and Star Wars in the written word. There's so many novels that are part of those universes, but it's cool to see for people who also would like to be more visual that they can have the same experience. I think it's great. This is what people are actually doing. It's interesting you bring that up.

1:28:07A lot of these Star Wars tales of the untold that we're showing here, they came from the comic books and from the novels. So people are going to the comic books and the novels and just saying, here's the source material. And I guess they put that source material in and they say, just animate these stills. So you could take any comic book. So if you loved Teen Titans and you wanted to make a frame by frame, you could just take the Teen Titans. I wonder why nobody's done this. If somebody wants to do this and come on the show and just show how they built it, I'd love to have somebody come on. I'm not going to give you a bounty because I don't want to get in trouble with IP.

1:28:39Yeah, I was going to say. But I would love to get a how-to series. just take um here's a here's a deep pull dazzler dazzler is this crazy comic book character that came out in the 80s of one of the x-men mutants if somebody does this um because everybody was fascinated by dazzler in the 80s kind of a cult character that appears for like five seconds in one of the um x-men movies you can pull up the cover of the first dazzler she can turn sound into light. So she's, there she is. That's Dazzler, but just pull up the cover of the Dazzler cartoon. So anyway, she's on roller skates. It was like a very of the seventies moment where they made a character based on like the vibes.

1:29:24And if you pull up the covers of Dazzler, you'll get, you'll get a real flavor of it. So somebody take the first episode of Dazzler, feed it into Sea Dance and come on the program and show us how you did it and make it into a couple of minutes of dazzer. So there she is. I think that's episode one. And she's there with Scarlet Witch, Beast, Iron Man, and Spider-Man. Gambit was also very popular. I think Gambit would be cooler. And just use CD-ANCE and then show it to us and how you built it. All right, everybody, we have a bounty for annotated.com. I bought the domain name annotated.com. I wanted to do a clipping service for like delicious back in the day, shout out Josh.

1:30:07And I want somebody to build it. You can win$5 ,000 if you build it. I get to own the IP. You write the code. I give you 5 ,000. We're going to do this over a month, I think. We have a couple of dates on the website. Just go to annotated.com and we'll put you to this lovable website we made. It's basically a sidebar Chrome extension. We made some images of it, but you pick how you want to build it. Alex, imagine you're a fact checker or you feel like you want to fact check a TechCrunch story, right? Or you want to add your feelings to it. You go to the TechCrunch story where you were editor at large or editor-in-chief, whatever you were, managing editor.

1:30:43Somebody highlights it, highlights a section, and then writes a commentary on it. And then it makes a landing page from that. Would that be fair use in your mind? I think it's absolutely fair use. It also drives attention back to the original article. But I think the point that it creates a landing page is what I like, because what I don't want is to see if I'm just going to a website to see all the people's scribblings, but I would love to have a way to say, show me the top five citations or notes from other people, a bit like in a medium story in a sense. Correct. So now you got it. The address is the URL.

1:31:17So once you have that URL from that TechCrunch story about, I don't know, the launch of Uber, and it's an old story and people want to put comments on it. So let's say I take it, I'm like, I highlight a paragraph about the Open Angel forum where they raised money. And I say, here's some commentary about that. And here's a picture from it. Then you, Travis, highlights a part of it, different part. And he says, oh yeah, and this was actually, here's the backstory to that. So two different paragraphs. But as you're pointing out, you could say this URL has seven annotations. And then you put the seven together in a very long page.

1:31:53You click on any one of them, it opens that page. All of them link back. Nobody can take the whole thing. So there's no stealing. It's just, and it's fair use because it's commentary on the original piece. And I want to be able to do it for an audio. So you can take a clip from this podcast. We'll limit the clips to like maybe 90 seconds or 60 seconds. And we're going to downgrade them to like postage size video. So, and link directly back. If you click on that video, you click on that story, you go right back. And it says, this is from this story. This is from this video. It just has so many uses.

1:32:26And so the only thing we're going to do as part of this is we own, or we own a license to all the comments, just like Reddit owns, you know, a license to all of the comments pasted on its website. And then that's the business model eventually is all that commentary can power a large language model. One other thing you could do with this idea, Jason, is let people who are the original people behind the article, behind the story, behind the poem, whatever it is, comment. So here's an example that I really love. This is Eminem's song Criminal over on Genius, which gets lyrics and puts them online.

1:32:58And he says, and you can't stop me from dropping each March because his plan at the time was to release an album every year. But then as he says here, I became a drug addict. but without that context, I would have thought, well, like dropping every March, like a syllabus or like, I wouldn't know, but you can just hear directly from the people. And I think it's very powerful and also very democratic. Interestingly, when the genius guys came to one of my conferences and I had them on the podcast back in the day, I talked to them offline. I was like, Hey, you know, I own annotated.com. Do you guys want to do a DLO?

1:33:29And they're like, no, we like genius branding better. So I was going to sell them the domain. I had, I think I bought the domain for 30 K 20 years. ago and I offered it to them and I was like, Hey guys, uh, give me, you know, uh, 500 K in equity and I'll, I'll give up the demand. Okay. Last thing for me. And then I know we got to go, but like, do you remember the genius guys? They were nuts. They were bonkers. One of them was nuts. They all had like a shtick. They wore like deep sunglasses. Somebody pull up the clip of this genius.com co-founders. One of them who passed away from, I believe a brain tumor, I was told or he actually confessed that he was suffering from mental illness and that was explaining his yeah Magabad I think Magabad if I'm pronouncing it correctly but yeah there they look at that with me look at the observer there's me and then a picture of the guy I don't know what that observer story is but I know I got into him look there's a young jcal at the DLD conference if you click on the story that that's from uber investor rap cheese cover and have the greatest twitter rivalry um you had beef with the genius boys i didn't know that mogadabim um was saying crazy stuff to me and i was like hey dude are you mentally ill i mean this is my younger years this is in 2015 what did i say let's scroll down here um uh now now we're just aping in the new mts i don't even remember this one uh check your privilege oh yeah look at this well no scroll down a little bit i don't have that tweet anymore i got rid of all my old tweets back in the day but he was coming at me something hard uh no one physically threatened anyone whatever you're still the biggest douche alive i mean he was crazy um and he would start fights with everybody i wasn't trying to say that people should be a little bit unstable i was trying to say that like they were fun like snap chat era spectacles uh vending machines fun uh do you remember meerkat which that like go live immediately thing from your phone.

1:35:28Yes. Fun. We need to have more fun. Everything's so fricking serious. I think SAS literally just drained the joy out of technology. Yes. I agree. If you are going to be in the era of C-dance and generative AI, we should be having a blast. So someone make me happy. Now, you didn't put anything on off-duty. Alex, the idea of off-duty is you bring something. Is there something amazing like a gadget, a piece of food, a trend? a dress that you love what are you into i am currently obsessed uh really hook line sinker got me by the collar with a game called captain of industry now if you are a person who plays video games you've heard of factory automation games the the the og here the dream theater of this genre if you will is factorio which i have over a thousand hours in um and i've done my 100x and 1000x science runs so don't don't come at me with factorio knowledge captain of industry is different jason it's a little bit less scientific a little bit more you're on an island turned into an enormous factory and i have just a little clip here i thought i'd share but i will say if you really want to get worse at your job and ruin your sleep go ahead and start playing captive industry because this game will eat you this is a windows game i'm playing probably plays on anything i just pulled up this video what is this i'm looking at is this like a technology tree or something That is a technology tree, yeah.

1:36:49Oh, look at this. It is shockingly complex because you can change the entire terrain. You have to deal with dumping, so you build piers and there's shipping and there's a global economy and trains and trucks. Is it multiplayer too or no? I don't have any friends, so I don't know. But you can find that out. But this has kept me sane during the very small child hours, if that makes sense, Jason. So a huge shout out to the guys behind Captain of Industry. I love your game. You're the best. Is this like a big title game or an indie? Pretty darn indie. I think it's like 20 or 30 bucks. It's still in early access.

1:37:24It's made by a small team. And it's just fucking great. Why don't we have them on the pod? Have them on the pod. I will bring them on the pod. It is not multiplayer. Yeah, I would love to have some of those small game folks and understand how that startup, part of the startup ecosystem works. Okay, I have a couple of ideas for companies we can bring on. We'll get that done. No, this one specifically, the one you're playing. Oh, no, we're going to have them on first. But I have them on first. Indie Studios. Yeah. Indie Studios. I like the idea of just understanding how that works, because now you have all these people who are willing to pay in advance for the game and Patreon.

1:37:55And I don't know how they do it, but I think they do raise funding early and they have enough support because the gaming community is used to paying a little bit. And yeah. Yeah, yeah, absolutely. Anyways, we're back on Wednesday, noon, Texas time, 1 p.m. Eastern. We'll see you all then. Go Knicks.