Wisdom of the $TAO: the future is decentralized AI

Summary

This episode dives deep into the intersection of cryptocurrency and AI through the lens of Bittensor, an open-source protocol powering a blockchain-based machine learning network. Rather than wasting compute on solving equations, Bittensor miners devote their resources to powering innovative AI applications called subnets, earning TAO tokens in return. Ala Shaabana of Crucible Labs, one of Bittensor’s co-founders, explains the protocol architecture, while Mark Jeffrey of Stillcore Capital showcases his favorite subnet projects and demonstrates how he trained an AI agent to mine crypto.

The episode also features two OpenClaw demos: Michael Ryaboy shows off SetUpClaw, a “set it and forget it” approach to AI agents, and Pete and Maddie Reese demonstrate Bizzby AI’s “startup-in-a-box” for the AI economy. The show explores how decentralized AI creates a free market for compute across storage, GPUs, and AI services, and how AI is even being used to restore old movies and media. The episode closes with an “off-duty” segment where Jason and Lon Harris discuss lighter topics.

Highlights

Bittensor Explained: How TAO Subsidizes AI Products

“Bittensor explained — how TAO subsidizes AI products through decentralized mining.” — Mark Jeffrey, 6:13

yt-dlp --download-sections "*06:13-09:52" "https://www.youtube.com/watch?v=X0AJVcKieyA" --force-keyframes-at-cuts --merge-output-format mp4 -o "bittensor-tao-explained.mp4"

How Subnet Tokens Work

“How subnet tokens work — the economics of decentralized AI applications.” — Mark Jeffrey, 12:20

yt-dlp --download-sections "*12:20-17:34" "https://www.youtube.com/watch?v=X0AJVcKieyA" --force-keyframes-at-cuts --merge-output-format mp4 -o "subnet-tokens-explained.mp4"

AI Agent Mining Crypto

“OpenClaw plus crypto mining demo — training an AI agent to mine cryptocurrency.” — Mark Jeffrey, 28:20

yt-dlp --download-sections "*28:20-31:28" "https://www.youtube.com/watch?v=X0AJVcKieyA" --force-keyframes-at-cuts --merge-output-format mp4 -o "ai-agent-crypto-mining.mp4"

AI Restoring Old Movies

“How AI is restoring old movies and media!” — Jason Calacanis, 33:07

yt-dlp --download-sections "*33:07-40:57" "https://www.youtube.com/watch?v=X0AJVcKieyA" --force-keyframes-at-cuts --merge-output-format mp4 -o "ai-restoring-old-movies.mp4"

SetUpClaw: Set It and Forget It AI Agents

“SetUpClaw — set it and forget it AI agents.” — Michael Ryaboy, 40:57

yt-dlp --download-sections "*40:57-48:08" "https://www.youtube.com/watch?v=X0AJVcKieyA" --force-keyframes-at-cuts --merge-output-format mp4 -o "setupclaw-demo.mp4"

Key Points

• Crypto meets AI agents (4:11) - The booming intersection of cryptocurrency and AI agent technology
• Bittensor protocol (6:13) - How TAO token subsidizes AI products through decentralized mining
• Mark’s top projects (7:37) - Mark Jeffrey showcases his favorite Bittensor subnet projects
• Subnet token economics (12:20) - How tokens work within individual subnets on the Bittensor network
• Crucible Labs wallet (17:34) - How Crucible Labs simplifies subnet token allocation
• AI crypto mining (28:20) - Demonstration of an OpenClaw agent trained to mine cryptocurrency
• Free market for compute (31:28) - Decentralized marketplace across storage, GPUs, and AI services
• AI film restoration (33:07) - How AI is being used to restore and upscale old movies and media
• SetUpClaw demo (40:57) - Michael Ryaboy’s “set it and forget it” approach to AI agents
• Bizzby AI startup-in-a-box (48:08) - Pete and Maddie Reese demonstrate a startup toolkit for the AI economy

Mentions

Companies

• Bittensor (6:13) - Open-source protocol for blockchain-based AI network
• Crucible Labs (17:34) - Ala Shaabana’s company, Bittensor co-founder
• Stillcore Capital (5:49) - Mark Jeffrey’s Bittensor-focused investment fund
• SetUpClaw (40:57) - Michael Ryaboy’s “set it and forget it” AI agent platform
• Bizzby AI (48:08) - Pete and Maddie Reese’s AI startup toolkit

Products & Technologies

• TAO Token (6:13) - Bittensor’s native cryptocurrency
• Subnets (7:37) - Individual AI applications running on the Bittensor network
• OpenClaw (28:20) - Open-source AI agent framework integrated with crypto mining

People

• Mark Jeffrey (5:49) - Co-founder of Stillcore Capital, Bittensor investor
• Ala Shaabana (17:34) - Co-founder of Bittensor, CEO of Crucible Labs
• Michael Ryaboy (40:57) - Founder of SetUpClaw
• Pete Reese (48:08) - Co-founder of Bizzby AI
• Maddie Reese (48:08) - Co-founder of Bizzby AI
• Lon Harris (54:35) - TWiST co-host

Surprising Quotes

“Rather than wasting compute on solving extremely long equations, Bittensor miners devote it to actually powering a range of innovative AI apps.” — Mark Jeffrey, 6:13

“Users mine a crypto token — TAO — but rather than using up their compute on solving equations, they devote it to powering AI applications.” — Ala Shaabana, 17:34

“The future is a free market for compute across storage, GPUs, and AI services.” — Mark Jeffrey, 31:28

Transcript

Jason Calacanis: 0:00 All right everybody, welcome back to Twist. It is Friday. It is March 6th. It is 2026, or in the year of our overlords AO 41, which stands for… Lon Harris?

Lon Harris: 0:12 Oh, it’s after OpenCLA. It’s 41 days since we first discussed OpenCLA on the show. AO 41. We’ve said AOC for a while, but that’s already somebody’s initials, so…

Jason Calacanis: 0:23 Yeah, we can’t go with AOC. Although that would be… I like AOC personally as the designation, but we’ll go AO 41. It rolls off the tongue. And you know, there’s so much going on in my world right now, Lon.

Lon Harris: 0:40 Yeah?

Jason Calacanis: 3:00 We’ve got a full show for us—

Lon Harris: 3:02 My God, too much show.

Jason Calacanis: 3:03 My friend, my partner Mark Jeffrey is here from Stillcore Capital. How are you doing, Mark?

Mark Jeffrey: 3:08 I’m doing great, Jason. Thanks for having me on as always.

Jason Calacanis: 3:10 Yeah, good to have you on because the intersection between AI, open-clause agents, and crypto, you know, I famously a couple years ago said, ‘Hey, listen, if you’re in crypto, pivot to AI.’ I’m not prepared to say that was good advice, by the way. People have given me—every time Bitcoin goes back down into the 60s, somebody retweets it, ‘Oh, Calacanis nailed it.’

Lon Harris: 3:32 Do you remember we were office shopping here in Austin and we went to this office and they had that tweet of yours printed out and framed? They didn’t know—it wasn’t for our benefit. They didn’t know we were coming. That was just part of the decoration of their office. ‘If you’re into crypto, pivot to AI.’ That was the wisdom.

Jason Calacanis: 3:48 Yeah, so the crypto people will dunk on me for that, but here’s the truth. There is some great innovation and I just did an interview with the new head of the SEC on All-In, that’ll be coming out on the All-In interview show next week, Chairman Atkins. And then I also had Selig, the CFTC head, so Chamath and I just did a great interview with them. We taped it recently and it’ll be coming out next week. And they are really laying the groundwork for making crypto legit and they’re going to take it back from all these offshore locations and bring it back to America. So this is a great thing. So I am not saying pivot from AI—

Lon Harris: 4:21 Back to crypto.

Jason Calacanis: 4:22 Back to crypto. That would be foolish. But I would say keep an eye on crypto because there are things that crypto is, you know, just I think at the end of this year going to be able to do legally and already starting to do legally that could be quite material in AI. And that’s why I wanted to bring you on. Maybe you could tell everybody what you’re working on today. And when I see you pushing your chips, because you’re a great investor and you’ve always been ahead of the curve. People can look up Mark’s Wikipedia or Grokapedia page and see all the innovations he’s seen. I don’t know, he’s typically 18 months ahead of everybody else and he’s probably nine months ahead of me. So I always go to Mark, ‘Hey Mark, what are you thinking about?’ Tell everybody what you learned over the last two years and why it’s important and where you pushed your chips.

Mark Jeffrey: 5:11 Yeah, so my chips are pushed on a thing called Bittensor and the coin for that is Tao. And it’s a very interesting new phenomenon. It’s basically taking Bitcoin’s mining side and making that programmable on a new chain, so it’s not on Bitcoin, it’s on its own chain, the Tao chain. And by making the mining side of it programmable, it incentivizes 128 subnets which are sort of like companies to build new AI products. So there’s 100 million dollars a year that are handed out basically as a subsidy, just like how Bitcoin hands out new Bitcoins to miners, right, to subsidize its own growth, the Bittensor chain hands out Tao to subsidize the growth of new AI products. And it’s been doing some very incredible things. We’re only about four years into it. The last year in particular has been—

Lon Harris: 6:00 been pretty incredible.

Jason Calacanis: 6:01 So Lon, this is important because, you know, we talked about the project known as Bitcoin, the tip of the spear that got everybody engaged in this. But the math problems and all the CPU was going towards busy work.

Mark Jeffrey: 6:15 Right.

Jason Calacanis: 6:16 It was not actually serving a purpose other than, you know, establishing the Bitcoin network, which was an important thing to do, but it also was using a lot of energy.

Lon Harris: 6:27 Right. And this was a big knock against Bitcoin in the early days, if you’ll recall, oh, it’s so wasteful, we’re using all this power so that they can crunch these numbers and then we don’t even need the numbers to be crunched and it’s just a mine—you know, like, that was, this was a huge knock against Bitcoin in its early days was that we were wasting all this power just solving these problems so we could create this artificial scarcity.

Jason Calacanis: 6:49 So Mark, take us through two or three of the most important inspiring projects and then for full disclosure, you were starting a fund to take advantage of this called Stillcore and I joined you as a partner in that, I put some of my chips in. I think I’m going to be ready to put some more chips in, I’ll be honest, this year I’m feeling a little frisky. Take us through maybe two or three of the projects, if we could show them on the screen, even better, but what’s inspiring and interesting?

Mark Jeffrey: 7:17 Yeah, let me, I’ll just, I don’t have anything ready for the screen, but let me just respond to your request. So Jason, the biggest news in BitTensor in the last 24 hours has been the launch of Ridge’s Subnet 62’s product. Now it’s a vibe coding platform, it competes with Claude and Cursor, and basically it’s scoring 73 to 88 percent on the SWE benchmark test, which is sort of like the SAT that measures how effective a vibe coder is. So it’s comparable or better than Claude, and it’s scoring 96.3 percent on the polyglot test, which is a different version, a different benchmark, but measuring the same thing. So the price that they’re launching at is 29 dollars a month. So it’s five to seven times cheaper than Claude Code and Codex at a comparable benchmark, so it’s the same or better in terms of quality. And it was built on BitTensor for roughly 10 million dollars in emissions from the chain, while Cursor had to raise against a 2.9 billion dollar valuation to accomplish the same thing. So pretty interesting.

Jason Calacanis: 8:26 Okay. So this is a co-pilot for developers running on the BitTensor network. It’s one of the subnets. So you can go buy Tau like you buy Bitcoin. And I don’t give financial advice, but if anybody’s listening and you had a hundy that you were going to spend on some margaritas, I’m going to say, ah, maybe drink some water tonight. Get some filtered water, tap water, whatever, have a nice tea, put the other 99 dollars into just buying a little Tau for yourself. Why not? But so this Ridge project, now is this an o— Open source project is it a fork? Where is the code base coming from?

Ala Shaabana: 9:04 The codebase was coming from the miners. So basically the miners compete with each other to improve the efficacy of the ridges product. So just like, you know, like Bitcoin does what Bitcoin does for stranded energy, BitTensor does for stranded talent. So anywhere anyone anywhere in the world can compete and the miners that are improving the subnet 62 product, um they’re earning something like 50k in subnet tokens per day. So, you know, if you if you if you win yeah if you win the contest, uh you you you know you could you could make a million dollars very fast if you’re the best in the world right and you’re some guy in Turkey and you you can’t get to Silicon Valley, this is a way for you to earn and own part of the AI revolution. And that’s one thing that’s great about crypto, right? So you can’t really do this in any other way. So it’s pretty interesting.

Jason Calacanis: 10:58 So unlike if you were making commits to open claw, you get, um you know, some kudos and you get to feel good, fantastic you’re helping humanity, open source is a gift to humanity, fantastic. It’s democratizing. But here there’s an incentive layer which is the subnets tokens. So I make this coder, this copilot better, I make some tokens and then I can put those tokens and I don’t just get a high five, I won a contest to to fix something in the product and I get to put that in my wallet instead of going to work for cursor. Is is that a way to think about it, I could go work for cursor as a developer or I could work on this open source open subnet?

Ala Shaabana: 11:43 That is exactly correct. That is exactly correct. And when you earn the 50k worth of subnet 62 ridges tokens, um you can hold onto them or you can cash them right out, right? So you can get the 50k and put it in your bank if you want. Um or you may choose to hold onto them because uh the market cap of ridges…

Michael Ryaboy: 12:00 is right now is very low. It’s like around $30 million compared to Cursor’s 30 billion with a b. So right, so if…

Jason Calacanis: 12:08 Now if I’m not a developer and I wanted to buy that subnet, that’s different than buying dollar sign TAO, correct?

Michael Ryaboy: 12:16 Correct.

Jason Calacanis: 12:17 How does one buy that? Because TAO is listed on Robinhood and Coinbase now or not yet?

Michael Ryaboy: 12:23 TAO itself is on Coinbase, I’m not sure if it’s on Robinhood.

Lon Harris: 12:27 It’s not. I don’t think it is. I’ve looked for it before.

Michael Ryaboy: 12:30 But TAO you can get on Coinbase. I’ve actually gotten it on Coinbase so I know it’s on there.

Jason Calacanis: 12:34 So you can get it on Coinbase if you want to have the top level, but explain to people in plain English how the subnet versus the $177 TAO works.

Michael Ryaboy: 12:42 Yeah, so you have to stake the TAO to get the subnet tokens. And one and you basically need to get a BitTensor wallet. And a great BitTensor wallet is the Crucible wallet which Ala will talk about down the road here on the show. What a segue. Yes. But that is his product, right? To make it easy for people to get into these subnet tokens, right? It used to be hard over the last couple months, it’s become a lot easier. And Crucible is leading the way in making it easier for people to do that.

Jason Calacanis: 13:06 Alright, so let’s introduce Ala on the line and keep this train moving.

Lon Harris: 13:09 Yeah, so let’s talk to our next guest, Ala Shaabana from Crucible Labs. Now not only the co-founder of Crucible Labs, that great company that makes the wallet we were just talking about, but he’s the co-founder of BitTensor itself and the Open Tensor Foundation, leading the development and maintenance of BitTensor since 2019. And Crucible Labs, of course, invests and builds within the BitTensor system with a focus on decentralized AI. Ala, thank you so much for being here.

Ala Shaabana: 13:38 Thanks so much for having me guys. I appreciate it.

Jason Calacanis: 13:39 You’ve heard sort of the setup here for BitTensor and you’ve got a commercial company that makes a wallet. This is a venture-backed company, a company that you just self-funded?

Ala Shaabana: 13:55 It’s a company we self-funded, yes.

Jason Calacanis: 13:57 God. So this is like being in the early days, Ala, of Bitcoin where I don’t know, Mt. Gox came out, some wallets came out. There were lots of different small projects. You found a blocker. It was very hard for people to know how to get the subnet’s tokens into their wallet and that’s what your wallet allows people to do and they could just sign up and it works I guess similar to the early days of Bitcoin, yeah?

Ala Shaabana: 14:22 Similarly, yes. So to kind of put it into context, when we first launched the subnets, when effectively people were able to speculate and kind of look into each project and try to basically invest into it, I found myself effectively day trading because I was just looking at each project and evaluating each one and seeing which one works and which one doesn’t. I found that sort of it took up most of my time. And the better idea was to actually create an allocation mechanism to really kind of take my rewards that I’m getting on BitTensor and effectively distribute them towards the top performance subnets. And that’s sort of where the idea for Crucible Labs’ wallet was started in that it becomes a very hands-off… thing, you literally stake and forget. You place your TAO within the Crucible Labs’ uh within the Crucible Labs allocator and it will basically allocate the rewards you get from your principal to the subnets.

Jason Calacanis: 15:14 Got it. Okay. So, maybe you could give us your thoughts on why you’re so obsessed with this project and if this succeeds, will the subnet tokens make their way onto Robinhood and Coinbase? And then, what— why is that a blocker? How does that work? Do I have to call Vlad and say, ‘Hey, open this up’? Because I can talk to him about it, or I can email Brian Armstrong. I’m assuming they’re aware of this, but usually they’re fast. What’s— is there a blocker here that we don’t know about, Ala?

Ala Shaabana: 15:51 TAO itself, T-A-O, is acting as a medium towards access to all of these subnets. You’re unable to really, I want to say, buy and sell these without going through TAO in the first place. And as a result, effectively these exchanges will have to really go through that process first before they’re able to even allow users to buy and sell. So this is not quite a blocker, but it is more a mechanism to ensure that a, TAO itself represents all the subnets and at the same time the value accrual of the subnets is reflected back into TAO in the first place. So as a result, um, there are some exchanges that are looking into listing it in this way, but it is a bit of a process because there’s also things like legal KYC on the subnets themselves that they have to go through. Each subnet has to go through effectively the same scrutiny that TAO went through before it was listed in the first place. And there’s hundreds of these, so it takes a while to do that.

Mark Jeffrey: 16:41 Actually, Kraken is ahead of everyone else. Kraken is a top validator on the BitTensor blockchain. So they are very active in our space. They do not yet list subnet tokens, but being a validator would be the first step to doing that. So my guess is they’re coming.

Jason Calacanis: 16:57 And as an investor and partner in Stillcore Capital, I assume, you know, this was part of your original premise that you’re dipping into the subnets and making investments across those, yeah?

Mark Jeffrey: 17:07 We are, yes. That’s in fact our charter, to invest in subnets. Yeah. So.

Jason Calacanis: 17:11 And so that’s what you do. You have this premise at Stillcore, hey, we know that TAO has the potential to be the next Bitcoin, Ethereum, whatever it is. It’s not guaranteed, nothing is. This is a bit of the Wild West, this is, you know, again, back to tip of the spear. But after you make that bet, you’re making the second bet, and the second investment is which one of these will break out? And so we just started talking about this new revolutionary concept of, hey, what if you could replicate Cursor? So then my question is if it’s only worth 30 million, that is the lowest valuation of people in this space, how do you evaluate that as an investment?

Mark Jeffrey: 17:56 Yeah, I mean, we look at these, we look at these subnets like you would look at startups, right?

Michael Ryaboy: 18:00 Right, you know, and you know I come out of that world, right? So it’s the same lens. What’s the total addressable market? Is this product actually competitive in that market? Does it have a significant advantage? What does the team look like? How good are they at marketing, right? Because it’s not enough to make the best product. You have to market it the best also. So if you kind of can come out and check all those boxes, and each one of the companies or subnets that we’ve invested in to date has checked all of those boxes. So we’ve announced three investments: Ridges is one of them, Targon is another one, subnet four, and they provide very, very cheap inference.

Jason Calacanis: 18:39 Oh, great. Inference based on—tell me that the name of that one again.

Michael Ryaboy: 18:43 Targon, subnet four.

Jason Calacanis: 19:50 So that subnet number four, Targon, is doing inference, which means instead of me sending something to Claude Opus and, you know, paying for it, I can go to Targon and have my open Claude do inference there? Is that what’s possible?

Michael Ryaboy: 20:05 Yeah, so it’s—you can do that. So Targon supplies its industrial-grade inference, so it’s private, right? Right, you’ve heard about this, I know you’ve been talking about this lately, right? So private inference: your prompts are not consumed and used to retrain, you know, some big company’s model.

Jason Calacanis: 20:26 Yeah, we just had Eric Voorhees from Venice AI on the other day talking about this same thing. He’s also doing private AI.

Michael Ryaboy: 20:32 Yes, but this is provably encrypted end-to-end. They have a trusted execution environment. You do have to pay for the inference, right? So it’s not free. They’re a company, right? So and they actually do have investors. So they are a traditionally venture-backed company as well. Their investors include Tobi Lütke, who founded Shopify, and Ram Shriram, who was the first investor in Google. So yeah, so they’ve got some blue-chip investors as well.

Jason Calacanis: 20:57 Okay, so that subnet is a— for-profit corporation, and in addition to that, there is a subnet token to incentivize people, I’m going to guess here, to put up infrastructure hardware, to provide the hardware, or are they standing up their own hardware as a company, Targon?

Mark Jeffrey: 21:19 Nope. Targon, I mean, Targon probably has their own hardware, but the bulk of it is served out of their mining network, and people are supplying the inference to the Targon mining network. Now, there’s some rules. They have to run the trusted execution environment and the setup that Targon dictates to them. But if they do that, you know, whoever supplies the most inference to Targon to resell gets the most Targon Subnet 4 tokens as a miner.

Jason Calacanis: 21:44 Got it. So there’s two ways, as you can see, like Bitcoin, there’s so many pips and shovels and services and application layer here, there’s all different places to invest. Targon obviously has some all-star investors in their corporation, but another way to make an investment in Targon’s vision is to buy the subnet token.

Mark Jeffrey: 22:05 Correct.

Jason Calacanis: 22:06 Okay, and we’ve done that from Stillmark. Fantastic.

Mark Jeffrey: 22:07 We have. Yep. And the third investment that we’ve made is in Hippias, which is decentralized storage. So kind of similar to Filecoin, but, you know, done in the Bittensor way.

Jason Calacanis: 22:17 This one is less interesting to me because storage is such a commodity. Why would you consider this one if storage is a commodity?

Mark Jeffrey: 22:26 First of all, the price for Hippias is quite a bit lower than everyone else, including Filecoin. So it’s 400x to 4,000x less expensive. So they’ve been able to very, very significantly shave off the price. And they also are very good at the Hippias token, the tokenomics for Hippias are very interesting and very novel. They have deterministically tied the value of the Hippias token to the increased usage and increased revenue of the Hippias product. So you need the Hippias token to mine. So you must stake Hippias to mine. So it’s a stake-to-earn, right? So basically they’ve got all these things that are continuously driving demand for the Hippias token. There is no scenario in which Hippias the product and the revenue does very well, but the token does not. So that was a very interesting new innovation.

Jason Calacanis: 23:07 Got it. So this is one of the challenges in crypto is where does the value accrue to, and they have solved for that. And that was one of the reasons in your investment thesis of why to engage with this project. Again, Hippias, private corporation as well with investors in the corporation, I assume. Where does the hard drive space come from?

Mark Jeffrey: 23:35 The miners supply the hard drive space. So storage is provided by the miners. The miners must be extremely performant, so they’re very snappy about, you know, serving up the file. They have to be online all the time, so they can’t just, you know, go offline randomly. So the quality of service that they provide and the amount of storage that they provide is the criteria that is used to award subnet tokens to them with each mining cycle.

Lon Harris: 24:00 So we keep talking about miners. Who are the miners in these systems typically? What is the profile of that individual corporation, hacker, zealot, early believer?

Ala Shaabana: 24:14 We don’t know. It’s like Bitcoin miners. It’s exactly the same thing. They’re people from all over the world. There is no onboarding KYC process. It’s permissionless mining, just like a lot of crypto.

Lon Harris: 24:25 So I have… it could be, I’m just taking a guess here, I’m a company like a maybe a Digital Ocean or something. And I’ve got massive storage that I haven’t sold yet, just making this up. I don’t know that Digital Ocean has done this, but you could pick another hosting provider. Hey, I’ve got, you know, these hundreds of terabytes that I’m not using, I’m putting online every day waiting for a customer. Hey, why don’t I just throw them in here, get those hard drives warmed up and monetize them while I’m waiting. If I can make some amount of money off them while I’m standing them up, hey great, and then maybe they will perform better and the token reward will accelerate and that would be better than me even selling into a customer directly.

Ala Shaabana: 24:54 Correct. It’s exactly like Uber, right? Like if you have all these unused cars that aren’t, you know, doing anything, you might as well put them on the Uber network and drive around to make some money, right? Same thing, just with storage.

Lon Harris: 25:03 Ala, when you look at the Marks bets there, what do you… what do you think of his bets? I’m looking for a little outside validation here. Walk us through those three projects and tell us if they succeed, how could they impact the infrastructure and the future of the inter-webs?

Michael Ryaboy: 25:17 First off, they’re very solid bets, well done Mark. Um, the one other thing I wanted to mention is there is… all these projects have a very interesting entry point into the real world, which I think is a problem that a lot of crypto suffers from. Um, you know, if you go to really any crypto conference you always see these projects that are marketed from crypto people to crypto people, but there’s very few that actually hit everyday users. And I think that’s what the entire industry’s missing. And I think the first projects to solve this are the ones who are going to be the ones really reaping the benefits. And I think whether it’s Hippias, Targon, or really any other subnet on the BitTensor ecosystem, whoever hits that first is a bit of a bit of an arms race happening right now within the ecosystem, which is rather exciting. But the ones who actually get there, and I think BitTensor does have the potential to get there first, those are the ones that are going to benefit the most. Whether as the teams themselves, which is another point that I like to evaluate when I’m looking into subnets, each team behind those subnets has I’m going to say a solid pedigree within the ecosystem. For example, Targon actually… that’s a good shout out to the CEO, but he actually is one of the first people that mined on BitTensor when it first started back in 2021. So he has a very in-depth knowledge and understanding of how the system works and so… this is the kind of person I’d be betting on. So I think all three of those… all three of those bets are very solid.

Jason Calacanis: 26:53 Mark has actually built an OpenClust skill that I think we have to take a look at called Vibe Miner. It sustainably

Lon Harris: 27:00 …and profitably mines a token SN 85 Vid-AI-O? Is that how you’re… am I saying it right, Mark? Vid-AI-O?

Mark Jeffrey: 27:07 Vid-AI-O? Yeah, V-I-D-A-I-O. So, yeah.

Lon Harris: 27:09 Vid-AI-O. I don’t know how to pronounce it. I’ve just seen it written, but yeah. Good guess.

Jason Calacanis: 27:15 Can you… Can you show us a little bit of how this works? I have a screenshot here in the doc, if that’s the best we could do.

Mark Jeffrey: 27:21 Yeah, just do… just show the screenshot. Sorry. But I can describe it. I can describe it.

Jason Calacanis: 27:26 All right. So we’ll… we’ll pull up the screenshot for the doc. So tell… tell us about why Open Claw and crypto mining are… are a good fit. I was fascinated by this.

Mark Jeffrey: 27:34 Yeah, so I… you know, I watched you guys, right? When Claw came out, you guys loaded it up right away on your machines. I waited until there was a hosted Claw before sort of diving in. And then I thought, like, how can I apply this to Bittensor? And I’ve always wanted to be a miner on Bittensor, but it’s… it’s usually a lot of… a lot of sort of sysadmin work and, you know, just stuff I don’t really know how to do, right? It’s not my… it’s not my forte. So when Claw came around, I was like, oh, I’ll just have my Claw become a miner on a Bittensor subnet. So I had the Claw actually choose which subnet to mine. So it’s mining subnet 85, which is video compression. It’s using AI to both upscale and compress video. And it does it 50 to 90 percent cheaper than anyone because it’s using Bittensor contestonomics to do it. And I am now one of the people upscaling and compressing. And I have to do it within 60 seconds of getting a request from the validators. And I’m using Bittensor infrastructure to do this. So I’m using Hippiest for the storage, right? So I’m getting 400 to 4000 percent off, right, on storage. And I’m using Targon for the… for the machine that I’m using to actually do the video crunching. So again, I’m getting… it’s like one tenth the cost.

Ala Shaabana: 30:04 So, and I’ve been able to make it profitable but not sustainable. So it costs me about 10 bucks a day. There’s about $3,000 being handed out by the Vid-IO network per day to miners. I am making about 30 bucks per day and it costs me 10 bucks, so it is profitable.

Jason Calacanis: 30:19 Wow.

Ala Shaabana: 30:20 However, after the immunity period where new miners basically get for like 24 hours, I usually last another 24 hours until more excellent miners kick me out. As Jason likes to talk about the age of excellence, BitTensor is the most ferocious form of capitalism and age of excellence that I’ve ever seen. So I get booted out frequently and I’ve not figured out how to stay in the rankings, but I’m working on that still. So, but yeah.

Jason Calacanis: 30:46 Okay, I gotta repeat this back to you so I make sure I understand what’s happening here. There are a bunch of distributed services being empowered by BitTensor that lower the costs of doing common jobs. A common job is, you know, or a large job, I don’t know, programming a large language model or processing large video jobs, yeah?

Ala Shaabana: 31:05 Yes.

Jason Calacanis: 31:06 These are large jobs in the world. So you’re doing arbitrage saying, I’m going to use the lowest cost BitTensor network for storage, I’m going to use the lowest cost BitTensor for inference—

Ala Shaabana: 31:25 Yeah, well, I’m actually just using the machine, so it’s not really inference in this case.

Jason Calacanis: 31:29 Just machine using GPU or CPU, whatever it happens to be, processors. So then you’re sending the jobs through this machine now that is grinding down to the lowest cost possible. And what this does is, it creates a free market incentive to just lower the cost of compute across storage, across programming copilots, across video processing. Is the core job here video processing?

Ala Shaabana: 32:07 Yeah, this is using AI to compress video and upscale it.

Jason Calacanis: 32:09 Right.

Mark Jeffrey: 32:10 Who’s doing those jobs? I’m just curious in the world.

Ala Shaabana: 32:11 I don’t know, I couldn’t tell you.

Lon Harris: 32:12 Oh, well there’s all those apps too that’s like take your old photos, take your old grainy videos, upscale them, make them look high def. Like that’s a big thing and especially Hollywood does it all the time. Like they’re using AI to like make Blu-rays or make 4K restorations of older films and TV shows. So yeah, there’s a lot of upscaling with AI being done right now.

Jason Calacanis: 32:29 You know what’s so interesting about that Lon, we had this whole thing where they were trying to preserve these final prints of you know amazing movies and like they would have some number of you know minutes in a print that were you know decomposing—

Lon Harris: 32:46 Exactly, for right, yeah, because the old film stock had started to decompose, yeah.

Jason Calacanis: 32:50 And in all likelihood, AI could not only restore it to its previous glory, it could restore it to— …a glory that the film was not capable of at the time of shooting.

Lon Harris: 33:03 Well, it’s a little bit like what we saw at the Wizard of Oz at the Sphere. Like they can use AI to not just restore the original image to look like new, they can add stuff in the background, they can blow it up to a different aspect ratio, they can they can put stuff in the background that wasn’t there. I mean, you can the sky’s the limit with what you can do with this technology to old films.

Jason Calacanis: 33:23 And when you think about that, Mark, the idea of losing art in the world to physics, you know, the Mona Lisa, you know, some great film, all these Grateful Dead shows that were bootleg. Not only could they be restored to their previous high water mark, they could be restored to a higher mark. You could—

Lon Harris: 33:48 Yeah, I mean you’re already seeing it, like if you go buy like The Duelists, that Ridley Scott movie on Blu-ray, you’re seeing it crisper with more detail than people saw it when they went to the theater in the ’70s.

Jason Calacanis: 33:58 If we were to, Ala, think about what could happen with infrastructure in the world, AI agents now could be examining the Bitensor network. I’m sure somebody has it. You came up with this clever idea, Mark, but your agent could go find 10 similar opportunities, yeah, Ala? And we might be in a situation where we’ve just got a, I wouldn’t call it a death spiral, but we would have a compression of costs that would just be relentlessly being pursued. Not in a human sense of being relentlessly pursued, but in a 24/7, 365, billions of tokens pursuit of lowering costs.

Ala Shaabana: 34:29 Yeah, I think that’s that’s quite possible that it could happen, but I think at the same time it will it’ll actually push competition, in my honest opinion. I think that’s actually going to be great because cost is one thing, but there’s also quality that comes out of that as well. And I think that’s what a lot of a lot of these projects that are fairly new tends to overlook a little bit, in that they’ll provide something at lower cost than anybody else but they’ll also be cutting quality. So I think that eventually these agents are going to be balancing between how well these tokens are that they’re pulling out or how well these services that they’re pulling out compared to how much that cost is going to be. And I think eventually it’s going to balance out to a specific ecosystem. And I think that Bitensor kind of has the advantage there because we have 102 projects and growing that could each compete within itself.

Jason Calacanis: 35:30 And there’s a foundation, Ala, that you’re part of that sets the agenda for subnets, yeah?

Ala Shaabana: 35:39 There was there is an OpenTensor Foundation that’s effectively what was started with the ecosystem. There was myself and Jacob Steeves, my co-founder. At this point we’re actually working on decentralizing even further. So I’m no longer part of the foundation, I’ve started working on Crucible Labs full-time. Jacob Steeves is also no longer part of the foundation, he’s working on his own subnet. And so now the foundation itself is sort of in a way— It’s handing over the reins to the community and now it will sort of following the exact same pattern as Bitcoin where now it’s being the project is being maintained by really the entire community and itself because it is open source and so anyone can contribute.

Jason Calacanis: 36:14 How does a new subnet emerge? Is this like doing a new top-level domain where there’s politics and payments and, you know, or can anybody who’s an entrepreneur say, ‘I have an entrepreneurial company, I want to be part of this. I want to launch a subnet’?

Ala Shaabana: 36:30 Yep, absolutely. Anybody can launch a subnet. There’s no—it’s permissionless, which is kind of the great thing about it. If you have a great idea, you can come into the ecosystem and launch a subnet.

Jason Calacanis: 36:40 Is there a certain amount of Tau tokens you need to launch one? And then what is the entry price now then?

Ala Shaabana: 36:44 There is a registration cost to it, and that cost is variable according to demand. So according to how many people want to enter the ecosystem, the cost will go up. If there’s less people trying to enter the ecosystem, the cost will go down. It’s a bit of a moving average sort of.

Jason Calacanis: 36:56 Got it. And how many subnets are allowed to be made, Mark? Is there like a cap on that? And what’s the cost now?

Mark Jeffrey: 37:05 At the moment there’s a cap. It’s 128 subnets, but down the road, not too much further—we don’t know exactly when, but soon-ish—that cap will be removed or grown. It’s—the cap is there to create quality, right? So we don’t want sort of a free-for-all like we’ve seen on Solana with pump.fun. You know, we don’t want pump.fun subnets, right? So the cap is there to sort of, you know, make sure that the proper mood is set on the chain. So it feels like the library, not like the party, right? So that’s what we want.

Jason Calacanis: 37:29 Got it.

Mark Jeffrey: 37:31 And, yeah, the cost, I haven’t—I haven’t checked recently, but it’s—it’s been as high as a million dollars to buy a subnet. So it just depends on when you go look. And I’ve seen it as low as like 100k. So it depends on the price of Tau, it depends on what’s available.

Jason Calacanis: 37:52 So this is not going to be chaos. It’s going to be orderly. Let’s, Mark and Ala, if I could have you guys come back in 30 days and give us an update, that would be really great for the audience because I think we’ve incepted in people’s minds here a bit. And one of the great ways I think to create continuity here is just check in every 30 days. What happened in the last 30 days? What bets have you made? What have you added? So I’ll see you guys on April 6th-ish.

Mark Jeffrey: 38:07 Love to do that. Put it on the calendar now.

Jason Calacanis: 38:10 I’ve made a note—I’ve made a note for my agent, he’s on it.

Lon Harris: 38:13 Okay, alright. And then nobody has to remember anything. We just sit here and tell our agents to do it. Just drop those guys off. That’s pretty mind-blowing, huh?

Jason Calacanis: 38:29 Oh my gosh, I want—I’m going to put in—I’m going to put in 5 or 10k. I’m going to get some stash together. I’m gonna do it.

Lon Harris: 38:40 I love it. I’m here for you making some bets and for your future. I mean, you’re going to retire. What are you, 45 now? How old are you?

Jason Calacanis: 38:46 Ugh, I’m 47. I’m 47.

Lon Harris: 38:49 47, okay. So you’re like, yeah, you’re six years, eight years behind me. It’s time to start thinking about your nest egg for sure.

Jason Calacanis: 38:59 Exactly. It’s time. It’s time. So, yeah, I just think a little bit… I mean, you need to have like, I would say at your age, 20% in things that can go to zero, but that could also 100x would be pretty nice, right?

Lon Harris: 39:10 No, I mean, my Solana is in that… it’s in that seat. It’s down… it’s down a lot, but I’m not… I’m not giving up. I’m not giving up yet.

Jason Calacanis: 39:18 I mean, it’s… I think it’s like a wise bet until you find a better bet. Let’s say, you know, your Solana goes back to where you bought it or it doubles. You can look at it and re-evaluate and say, “You know what? I’m going to pair 20% of this and put it into the next big thing. I’m going to pair 50% of it, or I’m going to take some wins and put it into QQQ or something,” you know?

Lon Harris: 39:31 BitTensor, yeah, here we go.

Jason Calacanis: 39:32 Or BitTensor, yeah. But 20% to that is, I think, a really good wager, a bet. You know, I like young people, especially since you don’t have kids yet… or may never…

Lon Harris: 39:52 Thank you for adding yet. Thank you for adding yet. I appreciate that.

Jason Calacanis: 39:54 Well, I mean… let me tell you something. This is going to be great for your retirement. Uh, you know, once you have kids, there’s no retiring. There’s just an, you know, a never-ending list of bills and complications coming.

Lon Harris: 40:05 We’ve got Michael Ryaboy. He’s here to show us Setup Claw. Michael, good… good to have you on the show. Uh, I love this, Jason. I found this on social media. It is a sort of bespoke, white-glove, open-claw concierge service for founders and executive teams. Uh, if you need, you know, if a CEO, a CFO, a head of sales needs leverage without creating new security risks, they get personalized OpenClaw service with Setup Claw.

Jason Calacanis: 40:42 Okay, so this is the Uber Black. This is the VIP level. setupclaw.com. Not going to Amazon or Cloudflare or popping up your own instance on your Mac mini. This is, “Hey, I want to make something for my company, but I’m a CEO, I’m a head of sales. I need to make sure it’s secure, and I need to make sure it’s working.”

Michael Ryaboy: 40:58 Basically, the story is… about a month ago, I go in and I go on Twitter and I say, “Hey, if you’re anywhere in San Francisco, then you can… I’ll go to your house on a Lime scooter and I’ll install OpenClaw for you on a Mac mini.” Now, this blew up. Suddenly, I have way more clients than I can handle. My Stripe is blowing up. And I’m thinking, “Okay, this is ridiculous. I install OpenClaw for some people,” and I’m thinking, “I’m having a terrible time.” And then every time I install OpenClaw for someone, they come to me a week later saying, “Hey, like, this broke. That broke.” You know? So, what I’m basically doing now is I’m going in, I’m building out custom agents for people. I’m still helping people with their Mac mini setups. I think that’s important. It’s a lot of fun, keeps… you know, keeps me fresh. But it very quickly became clear that, hey, like, this is not the right solution for most people. People need a SaaS. People need to just set their agents and forget it. Uh, people are busy. They are, you know, six, seven, eight, even nine-figure founders. founders who are like just like they don’t have time to configure open claw and deal with the bullshit associated. So what we what I did is quickly I built this SAS and this SAS is I consider the easiest way to run open claw right now. You deploy an agent and then you set your crons, you set how it works, you connect it has a full computer running so here’s like launch.co which you can just like use and once you log in you can uh you know save your authentication but it also connects to you know a bunch of tools. The goal is to eventually give every agent its own email, its own wallet, its own stealth browser so to give make it just like the most complete general horizontal agent platform available.

Lon Harris: 42:49 Okay, so open claw in a box well done. Are you still charging people for the VIP service and how do you charge for that? I know there’s a lot of CEOs who want to pay you a significant fee to do this, yeah?

Michael Ryaboy: 43:03 I hop on a 15-minute call with you. I figure out okay what are the things that you need most? Like what is what would your agent look like? What are you trying to automate? And then we hop on a second one-hour call where we actually go and we automate it. If you need to automate automatically update your CRM and Salesforce or your niche CRM, we go in and we set everything up during this one-hour call and you end up having one agent. And then we give you Hypercare, which is this idea of hey if you need additional agents or you need help with your existing agent or you need to make sure your agents are performing in a specific way, we help you with that or we help you set up automations to you know automate your business. So we have a bunch of automations running you know for a bunch of clients.

Lon Harris: 43:45 And what are you going to do next? I mean the open claw in a box has become a great startup. It’s kind of like building websites in the web 1.0 era, everybody needs a website, everybody’s going there. But it’s quickly being commoditized, I think you would agree Michael. So what is your eventual vision for this? In a year or two, you know we just saw Amazon release their open claw in a box service. What happens when there’s you know open claw in a box on every major platform? What where are you going to go with this next as an entrepreneur?

Michael Ryaboy: 44:10 Open claw itself is much less important than the innovations open claw brought which is proactive agents with uh the idea of heartbeats and crons. Now these agents are only as powerful as their tools and integration. So what we need to do is we need to level up the level of tools that they have to an extreme scale. Every single agent should have a wallet, it should have an email, it should like if you allow it to. It should uh be able to have access to a bunch of APIs by default. It should be able to make websites, it should be able to deploy them. You could start stacking these tools that these agents have and that’s when you have a big…

Lon Harris: 44:50 So the most empowered agent is the goal uh for you?

Michael Ryaboy: 44:54 The most empowered exactly. So we want to empower this agent to an extreme extent.

Lon Harris: 44:57 All right, great job Michael. Before we get to our next demo…

Jason Calacanis: 47:18 Who do we have next?

Lon Harris: 47:20 Father-daughter entrepreneurial team for you to meet here, Jason. This is Pete and Maddie Reese. They’ve created Bizby, which is, it’s sort of you tell Bizby what company you want to start and Bizby takes care of the rest. Maddie and Pete, thanks for joining us.

Jason Calacanis: 47:37 Wow, look at this. Father and daughter. I love this.

Pete Reese: 47:40 Yes, great to be here.

Maddie Reese: 47:41 Yes.

Pete Reese: 47:42 Yeah. Yeah, so this is my daughter. Yeah, amazing that it’s been, I don’t know, 25 years ago that she was born. But yes, so we’ve been working together for a number of years and this is our latest project and we’re super excited about it. We’re claw-pilled like you are.

Jason Calacanis: 47:55 You’re claw-pilled. Okay, show us what you built. Yeah.

Pete Reese: 47:58 So this is Bizby. As Lon said, you pretty much the concept is you type the idea for your business in the box and you hit start. And then you go through a procedure, a chat with Bizby. And this is an agent that will kind of walk you through…

Ala Shaabana: 48:00 walks you through you know the process of trying to decide uh you know if this is a good business trying to come up with a naming idea for it checking if the domains are available uh even goes through a logo iterations for you

Jason Calacanis: 48:13 Ah

Ala Shaabana: 48:14 and then you can go back and forth with it and kind of um you know iterate on that and then eventually it’ll when you sign off on the logo then it’ll you know show you branding package

Jason Calacanis: 48:27 at least

Ala Shaabana: 48:28 and then it will you know pretty much uh you know whole brand this this happens within minutes by the way this is uh this is

Jason Calacanis: 48:36 So this is a startup consultant in a box. This is an incubator in a box. This is a way to think about it Maddie?

Maddie Reese: 48:43 Yeah, yeah. So kind of our target audience is people who will become small medium business owners. So we are targeting people who do not currently have an existing business and our thesis really is that AI job loss or AI related job loss is going to be a massive deal even even this year but especially in 2027. And we want to provide a solution to that. We want to provide a way for these people who don’t have a way back into the traditional workforce to start something for themselves and be able to make money and live their lives.

Jason Calacanis: 49:21 This is a great idea Maddie because uh we have a concept which is surplus and as entrepreneurs when you see a surplus in the world that can be an opportunity. Now it obviously sucks to lose your job or get laid off but then a cognitive surplus appears. Smart person, they were working at Block, they were working at Microsoft, they were working at Amazon they get hit by layoffs they got two or three friends they worked with they need to figure out hey where am I going to get started here or maybe somebody’s a stay at home parent and they get back into the workforce they need a place to start and so uh this is the place for them to start that journey. What is your business then? Because obviously the criticism or the challenge you’ll get from the venture community if you decide to go that way and that’s but one way to do it is how is it different than just being a wrapper? How is it better than just using Claude directly? So have you given that some thought Maddie?

Maddie Reese: 50:22 Oh yeah definitely. We’ve built a lot on top of it. Uh so obviously the basis is kind of OpenClaude. Uh but there have been we’ve integrated a ton of different things we’ve built a lot of custom skills we support over 350 plus businesses right now um or different types of businesses. Uh and everything is customized for that individual business type with the amount of things we’ve built on top of it we we’re feeling pretty good.

Jason Calacanis: 50:49 Yeah, I mean and if you’re… go ahead Lon.

Lon Harris: 50:51 I was just going to say Jason one thing we should add after they’ve sort of I think the funnel on this is very clever I think after they’ve kind of shown you all the graphics of the logo and like hey this could be a real… Business, here’s what it would look like. Then they hit you with you want a team of AI agents to help you run the business? Do we just need your credit card? Like, that’s the next, that’s the next step. You sort of get open claw assistance to help you run the business now that you’ve launched it.

Jason Calacanis: 51:14 Exactly. That’s the real product.

Pete Reese: 51:16 Yeah, so it’s the team of agents that actually do the business for you. Um, as a new entrepreneur, you use your phone to run your whole business like, like many of us are doing with our open causes. The team takes care of all the messy business stuff on the back end that many of these people are not accustomed to and have no interest in getting involved with. Uh, they simply get to do what they do, maybe deliver the service, maybe it’s an in-person thing or maybe it’s something they do over the computer. But the team at Bizby, their agent team takes care of everything in the background and, uh, they just, they just collect the checks.

Jason Calacanis: 51:52 Fantastic, we wish you great success with this. Where can people learn more Maddie? Where can people learn more and, uh, engage this new product or service?

Maddie Reese: 52:00 Yes, so we are at Bizby.ai and that is B-I-Z-Z-B-Y dot A-I.

Jason Calacanis: 52:07 All right, Bizby.ai give it a try. What have you got to lose? Start a business. Uh, yeah, you gotta be resilient. What’s it like to work with your dad Maddie? Is he, uh…

Maddie Reese: 52:19 It’s incredible. I mean, we’ve worked together a lot, so I just, I’m very lucky to be able to have, you know, both of my parents are entrepreneurs. I grew up in an entrepreneur household. Um, it’s all I’ve known growing up. Um, and I just feel so lucky to have grown up in that kind of environment.

Jason Calacanis: 52:41 Do you have siblings?

Maddie Reese: 52:42 I do, I have two siblings, two little sisters. One of them is actually our co-founder. Um, she’s kind of in charge of our growth and content. Um, and then the other one is, is 16 and she’s being a kid.

Jason Calacanis: 52:54 Got it, and your dad doesn’t have favorites…

Pete Reese: 52:57 They’re all my favorite.

Jason Calacanis: 52:58 But I can see right now Maddie, I think you might be winning right now. Because every dad’s dream, I can tell you as a girl dad as well Pete, I have three daughters. I’m always like, who’s gonna come be Khaleesi and run, launch, and this week in startups? And like, one of them’s like, no interest. The other one’s like, very interested. The other one’s like, I want to work with sharks. I’m like, that’s fantastic, you go work with sharks, you could be an influencer or work with fashion. But I got one of them who wants to do business with dad and that is…

Pete Reese: 53:26 It really is a dream come true.

Jason Calacanis: 53:28 It is kind of a dream come true, isn’t it? Yeah, to get to work with your daughter. Yeah. All right, well done Pete and Maddie. Wish you great success with the company.

Pete Reese: 53:35 Yeah, thanks for having us. Thanks for doing it.

Jason Calacanis: 53:37 Hey, thanks for having us.

Lon Harris: 53:37 First of all Jason, you sent me this clip on TikTok, I don’t know if you remember this. I do want to take a look at this Bob Dylan clip that you sent because I think it’s really fascinating. Uh, this was Bob Dylan from a 1986 interview in Hamilton, Ontario for a BBC documentary. He was making the film Hearts of Fire, but some pretty prescient comments I think about…

Jason Calacanis: 54:00 computing, AI, and collaborating on art with machines. Here’s a look. You could have your own little band, like a one, a one-man band with these machines.

Lon Harris: 54:09 What year is this?

Jason Calacanis: 54:10 This was 1986, basically.

Mark Jeffrey: 54:13 That type of music doesn’t have any roots and it doesn’t have any foundation to it.

Lon Harris: 54:18 There’s the downside.

Jason Calacanis: 54:19 Yeah. I think that’s the sort of key insight that Bob makes here is like… some of that, I haven’t even messed around… 40 years. Okay, we can stop here. This is incredible.

Lon Harris: 54:29 You can mimic music, but it doesn’t… it lacks… it’s not grounded.

Jason Calacanis: 54:34 It needs to have the soul.

Lon Harris: 54:36 So this was I think really important. He’s always ahead of the curve, speaking of people ahead of the curve, he’s always been ahead of the curve, but he’s always, um, you know, and had these great insights. And he was always very obsessed with technically understanding the nature of songs. And he had a great show on Sirius XM for a while that I used to listen to. Um, and I forget the name of it, but on the show he would talk about songs and what made them great. If you are…

Jason Calacanis: 55:06 It was Theme Time Radio Hour, it was called.

Lon Harris: 55:09 Theme Time Radio Hour. And you know, I don’t think it’s on the air anymore. And people who were involved in this, who made clips for it, etc., they never actually were in the room with Dylan, like the radio show. They just, he kind of asked them questions and they, you know, they cut around him or whatever, but he did an incredible job of picking specific songs. So I think that’s going to be one of my recommendations for you. If you want to go down the Bob Dylan rabbit hole, like my pal Timothy Chalamet. My pal, we’re both Knicks fans, I met him.

Jason Calacanis: 55:40 You’re good buddy, yeah.

Lon Harris: 55:41 Good buddy. Well no, I met him twice, actually. But really, you know, great actor and really a true fan because when I talked to one of the two times I talked to Timothy Chalamet, who obviously did the Dylan movie and was fantastic, I asked him about the 80s, and if he ever got into that, and he was literally telling me specific songs from Infidels and other albums and Jokerman, and we had like a really interesting discussion.

Jason Calacanis: 56:00 Yeah, he goes deep. He got really good at ping pong too for Marty Supreme. He goes deep. He gets serious.

Lon Harris: 56:06 So I’ll add two here. I’m going to add two. There’s a very famous Pennebaker documentary, Don’t Look Back. That’s table stakes. But then there’s two by Scorsese who, as a great auteur, appreciated Dylan. No Direction Home, 2005.

Jason Calacanis: 56:21 Man, No Direction Home is incredible.

Lon Harris: 56:22 Incredible documentary. And so, if you were looking for something to do this weekend, I would go watch Don’t Look Back, No Direction Home, and then my personal favorite was, you know, when Dylan came out of the 60s and really wanted to kind of do something that he enjoyed. He did something called the Rolling Thunder Revue.

Jason Calacanis: 56:52 Yeah.

Lon Harris: 56:53 And this was a tour where he was in like a van and like driving around. This is Dylan coming out of the height of people’s obsession with him.

Jason Calacanis: 57:00 Right.

Lon Harris: 57:00 You’ll know this because he was wearing white- is that Kabuki makeup?

Jason Calacanis: 57:03 Pancake makeup on his face, yeah.

Lon Harris: 57:05 White, which he found out about because his violinist, who he had had, I don’t know if it was romantic or not, but he had a bit of an obsession with, was somehow involved with Kiss.

Jason Calacanis: 57:19 Right.

Lon Harris: 57:20 Who had also put the white makeup on.

Jason Calacanis: 57:21 Sure.

Lon Harris: 57:22 And my favorite version of all time of ‘Tangled Up in Blue’ and a number of other songs come from that Rolling Thunder Review period. And that Scorsese bookend documentary 2019 is extraordinary.

Jason Calacanis: 57:36 It’s sort of a documentary. It’s- it’s fictionalized, it’s very Dylan. It’s like a fake documentary, it’s about 60-70% of the information is accurate, but then they add in a bunch of like not accurate stuff for fun. So like all the Sharon Stone stuff, like that’s all made up, but like that’s a very Dylan thing to do. Like obscure, combine truth and lies so you’re not sure what you can really believe.

Lon Harris: 57:59 Just creating mythology, right? I guess would be the way to say about it. So this is great. We went down a little rabbit hole here, but a great set of Dylan documentaries for you.

Jason Calacanis: 58:06 And what else you got for us heading into the weekend?

Lon Harris: 58:13 So I definitely have three quick things I want to mention. First, I’m really enjoying this, it’s a little cheesy, it’s a Civil War scripted miniseries on Prime Video called ‘The Gray House’. Kevin Costner and Morgan Freeman both executive producers on it, written in part by John Sayles, the legendary filmmaker John Sayles, and then Roland Joffé, who directed that De Niro movie ‘The Mission’, bunch of other period stuff.

Jason Calacanis: 58:43 ‘The Mission’, my favorite film. One of my top five films of all time.

Lon Harris: 58:46 So that director made this- he directed every episode. It’s an eight-episode miniseries, it’s about Union spies, it’s about these southern women living in the Confederacy who were getting messages to the North to the Union Army based on what they were- they were living in Richmond, they were part of society, so they knew Jefferson Davis’s wife, they were in touch with the servants who worked in Jefferson Davis’s house, which was the Gray House, instead of the White House, they had the Gray House. And so it’s- it’s sort of loosely based on a true story, but it’s this very kind of old-school- you like it- I’m enjoying it. It reminds me a lot of like 70s and 80s TV miniseries, like from the ‘Lonesome Dove’ or the ‘Roots’ era, when those were like these big events and they’d have battle scenes, even like ‘Gettysburg’, that one, it feels like that where it’s a little hokey, it’s a little throwback, but very, very well done, very well acted, big epic battle scenes and stuff, I’m really enjoying it. So that’s on Amazon, all eight episodes are up on Prime Video right now.

Jason Calacanis: 59:40 And by the way, that was Barry Diller, the media executive, he’s the one who created ‘Roots’ and these movies of the week and this incredible genre.

Lon Harris: 59:51 Yes.

Jason Calacanis: 59:51 Um, which I think ‘Shogun’ was part of.

Lon Harris: 59:57 Sure, exactly. It- it feels to me that’s what this feels like, where it’s very heart-on-its-sleeve, it’s very sincere. It’s a little- it’s a little- little throwback it’s a little cheesy I guess you would say for for I think modern viewers would be like what is like what is this a little little you know old school but but if you grew up on those kinds of TV mini series I think you feel right at home it’s very assured storytelling very confident and it looks great there’s a lot of like big recreations of battle scenes and it looks really cool they did they did a good job with it.

Jason Calacanis: 1:00:22 He actually did a memoir too, who knew?

Lon Harris: 1:00:25 Oh right, Barry Diller. I may have talked about it on social media before, but this is my favorite biography of the last year or so. Who knew? And he talks about his personal life, his marriage to Diane Von Furstenberg, and you know being quietly closeted and and his struggles with that but then doing QVC and you know again USA Network all the way to IAC.

Jason Calacanis: 1:00:52 One of the real legendary executives of that whole the cable TV sort of era.

Lon Harris: 1:00:54 The last, kind of the last of a generation there and still in the game which—

Jason Calacanis: 1:00:58 —and Dara Khosrowshahi worked for him, the CEO of Uber. So there’s a whole generation of CEOs who cut their teeth working for Barry Diller who cut his teeth working for you know the series of executives before that in this 70s and 80s. Who knew? Great Barry Diller biography, there’s one for you for the weekend. All right. What else you got, Lon?

Lon Harris: 1:01:26 Up next, uh, this is on Disney Plus right now, I think Hulu probably as well. It’s a National Geographic docu-series, Incas: The Rise and Fall. I went to Peru many years ago, visited some of the sites. I’ve been really enjoying it. It’s just a really well-done sort of overview. There’s me at Machu Picchu from my Peruvian trip. So I couldn’t resist checking out a four-part Incan documentary and it does a really good job, like it digs into—I feel like we hear about these amazing empires in like South America or the Caribbean or North America, but you don’t really dig into their history, like it was this king and then they expanded this way. And this does a really good job of using what we know about Incan culture because they didn’t have a written language, so we’re sort of still piecing it together archaeologically and anthropologically. But uh, it’s really interesting.

Jason Calacanis: 1:02:12 Well and people don’t know you say Inca, you think of Machu Picchu, but they created you know a series of networks of roads, 25,000 of them including bridges over crazy gorges.

Lon Harris: 1:02:26 The Inca Trail.

Jason Calacanis: 1:02:28 The Inca Trail and this empire went I don’t know I think thousands of miles kind of the spine of South America and I’ve never I mean I just have a cursory knowledge of this so I’m going to watch this with my kids and obviously that all ended in the—

Lon Harris: 1:02:44 Conquistadors.

Jason Calacanis: 1:02:45 15th, 16th century. Yeah.

Lon Harris: 1:02:48 Yes it was Pizarro, Pizarro and his crew landed in Peru, uh, they set up shop in Lima on the coast and then they proceeded to conquer the Incan empire. We have a second photo of me in Ollantaytambo. this was the one Incan site, if we can pull it up. It’s a fortress. That’s me in the distance there. This is the only place where the Incans actually won a battle against the Spanish, because they were—

Jason Calacanis: 1:03:11 Is that? The Spanish had cannons.

Lon Harris: 1:03:12 The Spanish had a lot of advanced weaponry. They had sailed there and so they conquered. And you can still, in Cusco today, which was the Incan capital that the Spanish conquered, you can see they built their cathedrals right on top of the Incan temple. So you can see that it’s Incan in the foundation, and then it’s like a cathedral above. An amazing place if you ever get a chance to visit Peru, folks.

Jason Calacanis: 1:03:33 All right, there you have it folks.

Lon Harris: 1:03:35 One last thing I want to tip people off and you as well, Jason. On Sunday on Disney+, another National Geographic documentary is arriving. It’s called Ghost Elephants, but here’s the interesting thing about it. Directed and narrated by Werner Herzog. We’ve got a trailer. It’s about a South African conservationist and botanist, Dr. Steve Boyes. He’s in Angola tracking a mysterious, previously undiscovered species of elephant. They have not really done a good job of promoting this. I feel like people do not know this is happening on Sunday. But Sunday on Disney+, you can go watch a new Werner Herzog nature documentary. We got a bit of the trailer.

Jason Calacanis: 1:04:10 I cannot wait. He is incredible. His voiceover and his performance, vocal performance is always extraordinary. You know, everybody loves when I tell them about whatever gadget I’m buying. You know me, I got a very heavy Amazon finger. My trigger for Amazon, very high. Was dealing with a bit of an issue around the house. Two of the daughters get into audiobooks, the 10-year-olds. And then another daughter, she’s in her music phase. So I got three kids running around with over-the-air headphones. Bit of a problem because you say, ‘Hey, it’s dinner time,’ there’s no response. And then everybody’s kind of zoned out. That’s okay with me. I love the fact that people, you know, experience audiobooks. I’m really excited that my younger girls are getting into that early. And I’m really excited that my daughter and I share music, and we talk about Queen and we talk about Bob Dylan and we talk about Dire Straits—

Lon Harris: 1:05:02 The classics.

Jason Calacanis: 1:05:04 She loves Bowie. I feel like I’m doing my job. So I had to find a solution for this problem. And there is a company called Shokz, S-H-O-K-Z. And I think I may have seen Bill Gurley wearing these years ago. These are bone-conducting headphones. And they go over your ear. And they have one called OpenRun Pro 2. And you can get like a small size. You put this over your kid’s ears and people use this for sports mainly, because if you’re in the gym you don’t have to wear something bulky. But the sound is incredible. It doesn’t, you don’t have to use the bone conducting, you probably wouldn’t use the bone conducting now, you’ll just listen to it through the speakers, but you can. And now when somebody’s on the swing listening to some music, or somebody’s in their room listening to an audiobook and I say, ‘Hey, it’s dinner time,’ or I come in and start talking to them, they, you know, can hear me and they can pause the music, etc. And I much prefer the open-air headphones. If we’re traveling, obviously on an airplane, I’m not sure how these will perform, uh, compared to the noise-canceling ones, but this has been a nice for me when I’m hiking and working out as well. Going around ranch, I have to be aware of my surroundings at all times.

Lon Harris: 1:06:13 Yeah, you don’t always want it to be completely immersive.

Jason Calacanis: 1:06:16 When I’m at my desk, I will wear my Focal $5,000 Utopia headphones and I’ll get into— I’ve been listening to some Coltrane. Holy cow. Man, I mean, I guess I’m going into my jazz phase now, but I listen to some Miles Davis, some Coltrane, and it is amazing on those headphones. But for a podcast, I’m going to say these are the best you can do. Another amazing twist: Monday, Wednesday, Friday, we’ll be back next week, and on Wednesday, we get a double drop. You get your double drop. You get This Week in AI and you get This Week in Startups on a Wednesday. On a Friday, you also get a double drop. You get your This Week in Startups and you get your All-In for the weekend. So you got the double drop days Wednesday and Friday for J-Cal. And Monday, we just catch up on what’s going on. Alex will be here Monday, Lon with me on Wednesday and Fridays, the All-In crew on Thursday, and obviously, I do the roundtable for This Week in AI.

Lon Harris: 1:07:06 Mm-hm.

Jason Calacanis: 1:07:07 We had a spectacular one. If you can do me a favor: thisweekinai.ai. We’re just waking up the channel now. Thank you to our partner PayPal, who underwrote the show for the first year. That was really gracious of them and a really savvy move because the show’s off to an amazing start. Three— think of it as like a roundtable like All-In— three amazing CEOs and me, the world’s greatest moderator, just going deep, deep, deep on the week’s AI news. AI, AI only.

Lon Harris: 1:07:34 There it is.

Jason Calacanis: 1:07:35 There it is. We’ll see you next time. Bye-bye.

Lon Harris: 1:07:36 Bye.