Bittensor TAO Bridges: Massive Catalyst for Buying Pressure

This is a huge enormous catalyst for TA because behind the scenes on every bridge in order to buy a subnet token you have to buy Tao first.

It's amazing in the history of capitalism because I would say this is the first time someone's made a piece of code a platform that perfectly represents in the digital world what the free market is.

OpenAI announced a week or so ago that you can no longer ask medical questions and you can no longer ask for financial advice.

AI that can actually be far more disruptive and transformative because it's built by the collective of humanity as opposed to 20 people in PaloAlto or 50 people in San Francisco.

I think there's going to be new power structures in the world at the government level, at individual levels, and I think if you're not educating yourself very seriously and questioning the sort of structure of the existing systems, you're going to be left out.

Well, first off, JJ, so I feel like every time we talk, there are more catalysts than ever for Tao.

At least more catalysts that I'm starting to see and less worries. So, we've conquered my worries a long time ago about the value of alpha tokens and I think that was episodes three through six and we talked about, you know, the happening and things like that. Right. I think a huge catalyst for Tao coming out is that it's very hard right now to buy the subnet tokens.

There's only one way to do it is essentially get a TAW wallet, buy Tao, and then stake into the subnets.

And just for people who are listening for the first time, you could think of the subnets as basically startup companies that are doing amazingly well because they're using the full resources of the Bit Tensor or Tao platform.

But JJ, you were telling me how you're even building bridges between the different exchanges to the subnet token so it becomes easier to buy.

This is a huge enormous catalyst for Tao because behind the scenes on every bridge in order to buy a subnet token you have to buy Tao first.

Even if you're not aware that you're buying Tao because the br how the bridge works you have to buy Tao. The exchange has to buy Tao and that creates enormous buying pressure.

So as companies or subnets like ridges start achieving success or sports tensor or targon or whatever as they start achieving success and investors want to invest in the subnet tokens and the more exchanges that offer it the more they will be bought that means the more tao will get bought.

Yeah, I mean I think that that's exactly correct.

Bit Tensor today is not as widely accessible and easy to get exposure to as a retail investor as say Bitcoin or Ethereum. The options for mainstream consumers today are mainstream exchanges like Binance, Coinbase, Kraen, and others.

And those are sort of like general purpose cryptocurrency exchanges.

Interestingly, I actually noticed the other day I was looking at Kraen. You can actually also buy stocks on Kraen. So, you can buy like Apple and Nvidia and stuff.

And so, I think there's going to be at some point a sort of blurring of the lines between crypto and stocks in terms of mainstream retail platforms and tools.

I imagine you'll be able to buy Bitensor Alpha tokens on Robin Hood someday or through Charles Schwab. Part of what I think enables that like you mentioned are these bridges and sort of integrations.

Leighton Holdings is the company that I started about a year ago with Cameron Fairchild.

Cameron's kind of the core maintainer core contributor of of Bit Tensor has been involved since almost the very beginning. He was one of the first miners in early 2021.

Our team has been building VTOW or virtual Tao which is essentially a wrapping staking solution for Tao itself. You can have access to that from any exchange like Coinbase or Binance that supports EVM based contracts.

And so if you're familiar with Ethereum, the closest equivalent to something like this in the context of Ethereum would be something like Lido or Raph Repstake Ethereum.

And what that allows is just like this fungeible kind of extension point for Tao across all these other chains.

So you'll be able to access and expose your assets from all these other EVM supported chains into Tao. So you can buy Tao from Bass, you can buy it from Arbitum, you can buy it from all the layer 2 from Salana directly.

And then what we're also releasing is virtual Tao that's coming out pretty soon.

And you'll be able to get the same types of capabilities that you could see from VTA for the alpha tokens. And then what we think that will allow is just a lot more liquidity and access into the subnet ecosystems from stable coins and from other chains where there happens to be quite a lot of liquidity. I mean there's a lot of liquidity on Salana.

There's a lot of liquidity on Ethereum and we just want to make it dead simple for people to have trusted access to these technologies.

There's a lot of these bridges that are popping up and something I pointed out on Twitter a few days ago is that most of them are hacked together.

They're not audited.

They're not fully open source and they're not built by people who have been trusted already for many years to ship production quality software that you know our whole our whole ecosystem relies on.

And so I'm not going to name names but I'll just highly highly recommend people be extremely careful.

We don't charge any fees for VAT.

We also do not plan on charging fees for Virtual Tao.

They're all fully open sourced.

They're audited by Zelic.

We're planning on adding many layers of audits, third party audits, and our platform is based on layer zero.

We think it's the most scalable and fee efficient message passing infrastructure for basically supporting all the other ecosystems and chains that are out there.

And so that's that's kind of one thing I think you're right. I think it will be a catalyst for more capital inflows coming into the Bit Tensor ecosystem.

Yeah, because I was looking recently apparently in the past few weeks the market cap of all the subnets added together has gone from about a billion to a billion and a half.

So a 50% increase and that's without any of these bridges without any breakout success from the subnetss. We know there will be breakout successes but they haven't fully captured the public's imagination yet.

And when you and I were last together in Canada, we knew of at least one subnet that was getting calls from VCs but not even returning the calls because the entire incentive mechanism sort of provides everything you need rather than selling your soul to a venture capitalist. And so these subnets are just at the tipping point where they're getting noticed and yet it's still a 50% increase in market cap on them just in the past few weeks.

I think that's why Tao out of the top 100 crypto tokens is the best performing that and Zcash are the best performing two tokens in the P since the flash crash on which happened exactly one month ago today October 10th again if there's a breakout success which there will be and imagine if uh you can now easily you can go to Coinbase or any of the centralized exchanges or unis swap on Ethereum or anything and easily buy the subnet tokens just by looking it up and hitting click The market cap of the subnets is going to quadruple which means Tao could multiply by eight from here and you won't even know people are buying Tao.

They're buying the sub.

>> Exactly. It's just an implementation detail and a liquidity bridge essentially.

You can think of Tao as just the reserve currency as an incentive for all of the subnets.

And from a retail trader perspective, if you're hearing about ridges or you're hearing about Targon or one of these subnetss and you're just like, "Wow, this team built an incentive for people to compete with Cororeweave, Cororeweave is worth $60 billion, and they're aggregating compute more efficiently.

It's secure. It's got confidential compute enabled, right? They've got all the top chip options available on this network.

Man, seems undervalued.

I might want to might want to get some exposure to something like that." ridges.

They're building coding agents.

Coding agents are arguably the most profitable and fastest growing AI products getting built today.

Most estimates put around half of anthropics revenue in the coding agent category.

So Claude and Cursor because cursor is primarily based on anthropics APIs and you know that's something on the order of five or six billion a year in revenue and and and almost two if you add the two companies together almost $200 billion in valuation and ridges actually has a better AI model than both companies.

>> Yeah, this is really just starting to become more widely known.

I think capital efficiency in AI is probably the most underdis discussed sort of like imminent repricing event on the horizon.

I think that even Jensen and and actually Sam Alman was tweeting a few days ago about OpenAI like maybe wanted to compete with Coree. Why would they be competing if they needed all the GPU compute?

They're signing a trillion dollars of contracts. there's a mismatch and there's a there's a sort of a contradiction in that if they needed the tens of gigawatts of compute every possible cycle of those GPU compute resources they're buying why would they be considering building a cloud service and renting out the compute to the ecosystem it's obviously revealing a fundamental mismatch with their capital expenditure plans in terms of resourcing for inference or training when they actually may not need to be doing that and so I also think what potential is revealing is that there's just going to be more efficient ways of building the AI that will end up just leaprogging and being state-of-the-art in so many domains.

And we've already seen China do this with the open models.

The vast majority of the open models are Chinese models.

And the vast majority of the models being used by startups in the United States are open models.

And so they're being used and sort of fine-tuned heavily because when they're trained in China, they're sort of like politicized, right?

You can't you can't ask about Tianaan Square. You can't ask about, you know, very polarizing political topics in China where they actually censor censor your speech and censor what you what you can say.

And so there was actually a really hilarious someone was asking cursor cursor's coding agent, you know, what it thinks about like some Chinese censored topic and the model was like, "Sorry, I refused to kind of provide information on this topic." And so it was pretty obvious like cursor was using a deep sea fine-tune that they forgot to sort of like um uncensor properly.

But the innovation is coming from open source.

Bit tensor amplifies this and turbocharges it because it creates a sustainable incentive system for rewarding anyone permissionlessly on the planet to contribute to this.

What I think a lot of people don't understand is how do you actually provide a sustainability mechanism for open source?

Deepseek was subsidized by a hedge fund.

They were printing money, speculating and making trades on market changes like hedge funds do, long short hedge funds, and they were generating a lot of profits on those trades.

A big part of the trade, which is from January of this year, which seems like an eternity ago, was we're going to build a disruptive state-of-the-art model for reasoning, release it for free as open- source, and it's going to freak out everyone in the West, all these labs that are claiming that they need another hundred billion dollars to build a better reasoning model. And of course the result was if people forget as a reminder half a trillion dollars was wiped out of the markets in about 6 hours the morning after. I believe they released it over a weekend and then I think the following Monday is when the big market correction happened.

And of course they they they were able to make an asymmetric trade on that and it probably made 5 or 10 billion.

The big difference with bid tensor is you don't need to short a market reaction on open innovation.

you can just directly subsidize the open innovation and the network itself is looking for that.

It's intrinsically looking for more open innovation, more capital efficient innovation.

And it's really for the first time in the history of capitalism that a technology infrastructure exists that can allow anyone on the planet to participate in accelerating open source innovation and actually get paid to do that.

It's really profound that Bit Tensor makes this possible for the same way that Bitcoin incentivizes a more and more secure digital ledger for deflationary ultrasound money, right? To use one of the phrases, incentivizes people to run a cryptographic hash. Bitensor is doing this to incentivize anyone to create open-source permissionless continuously improving intelligence.

So, that's pretty awesome.

>> Yeah. And like you say, it's amazing in the history of capitalism because I would say this is the first time, and we've discussed this before, but I would say this is the first time someone's made a piece of code, a platform that perfectly represents in the digital world what the free market is.

I've we've never seen anything like that where a free market basically has some mechanism to trade and that's kind of the energy of the incentive mechanism and then there are things to trade.

There are people who have you know too much of something and other people need that extra supply in order to create what they are going to create.

So for instance Targon has spare compute so they're able to rent those resources to people who want to build an AI model in another subnet for instance and the tow and send mechanism allows all these things to flow through the platform.

You can think of towel as a currency in that case, but really it's more just a way of it's almost like a barter.

It's more just a way of providing a way for the transaction to happen rather than thinking of it as oh I'm going to use my tow to buy a cup of coffee, but this is really just to keep all the incentives aligned.

It's very fascinating.

And the thing is with this big AI, I don't want to call it a bubble, but there are two things about like the open AIs and the Geminis and the anthropics of the world.

you know, they're catering to the VC need.

People have to understand how venture capital firms work.

On the whole, there are many exceptions, yours included.

But venture capitalists make money by taking a 20% of profits.

Now, what are profits? Well, profit might be when a company sells, but it also might mean when a company has a higher valuation.

So if I'm a venture capitalist and I invest in OpenAI when it's worth 10 billion and then tomorrow it's worth a hundred billion.

It's not sold yet but there's a secondary market for those shares. I can get I can take enormous enormous fees. 20% profit that means 18 billion in fees I'm taking that year.

So it's to everyone's it's a little I don't want to call it like a mafia but you could almost think of it like an organized you know shady activity.

Hey, let's all get together and do the next round at a hundred billion on December 31st because January 1st is when we're going to get paid our 20% fees.

And then what happens?

Okay, all these guys are friends with Nvidia, Oracle, Google.

So, let's orchestrate, oh, you know, Oracle is going to invest so much money in Open AI and then Open AAI is going to buy, you know, data centers, hundred billion dollars worth of data centers from Oracle.

It's kind of this circle jerk that inflates the value of everything. It's very similar what happened but actually to a much larger extent very similar to 1999 when you know Cisco Pets.com all the VCs then everything was kind of trading at huge valuations then it collapsed because there were no users. The difference here there are users but to your point JJ you don't have to bet on a collapse to make money here.

You have to bet on what?

You have to ask yourself what entity is organically building real companies that are undervalued.

Again, ridges with a $200 million market cap versus Anthropic with an almost $200 billion market cap.

So, there's an arbitrage by organically growing companies, never VC funded.

So, you know, the valuation is closer to truth than the VC funded valuations.

And that's how you can make an enormous amount of money.

The problem is the news doesn't several problems. One is there's not enough news about Tao and Bit Tensor. You can't you can't go on Coindesk or the Wall Street Journal and read about Tao. It's one of the reasons why we're doing the Tao pod here and another reason why I did the TAO Daily.

com. So it's just so the news will happen and it'll start to grow.

It's just a little slower than the hundred billion dollar VC funded companies.

Another problem is, as you were saying, it's hard to buy the subnet tokens directly.

Anybody listening to this who's an investor should always think of problems as opportunities because either the problems don't get solved and then everything goes down or the problems do get solved and then everything goes up.

And if you think the latter is more likely, it's a good investment.

And yes, JJ's building a bridge, other people are building bridges, other exchanges are listing the subnets, the subnets are breaking out.

This is a problem that's If I were to bet, which I am betting pretty big, this is a problem that is going to get solved. Yeah, I think there's so many different things to understand about Bit Tensor and the world I feel still has a reluctance to think critically about anything with a crypto association because of the sort of conventional chaos and scamminess and sort of volatility track record of crypto at large. And I would propose a good way of really kind of having a more clear way of reasoning here would be to sort of detach the crypto label from Bit Tensor.

>> And in the same way that you can kind of think about Bitcoin more clearly in the absence of crypto, right?

Like Bitcoin is its own thing.

>> You know, it's obviously part of crypto and kind of created this whole movement.

Crypto wouldn't exist without Bitcoin.

In the same way that open source wouldn't maybe exist in the same way without Linux.

This is sort of the first major successful example of a huge ecosystem that really set the stage for how you do governance and how you do development and how you ship stuff and get people involved and talk about the ideology and the philosophy and all these things.

But in many ways, Bitcoin isn't crypto.

It's its own sort of technology that was built to solve a very specific problem and has actually succeeded and continues to succeed more and more and it does it in its own way.

And I think Bitensor is very similar.

Visav everything that came after Bitcoin and up to BitTensor is sort of like Bit Tensor is its own organism or its own technology and it should almost be detached entirely from crypto and sort of viewed as sort of like fundamental invention or discovery that allows it allows for sort of a reshaping of the way economic incentives are created and utilized by people and it's a super profound thing and we we we talk a lot about the sort of the broader narratives and stuff but like I was asked by someone at the end of a talk I was giving in DC a couple weeks ago at Nvidia's GTC conference and this gentleman was asking you know we're going to be in a world where you know there's all this abundance and there's all this like automation what about the you know two or three billion people that are sort of left out of the equation in terms of like being able to contribute and participate and really, you know, benefit from the economic aspects of things because we're seeing more and more wealth concentration.

You know, things are pretty obvious to everybody.

Elon's probably going to be multi- trillionaire.

There's probably going to be many trillionaires.

Actually, it's probably not just going to be one trillionaire. And so, like, you know, there's going to be more mega billillionaires, more trillionaires, probably fewer of the broader large universe of people at that level of wealth.

And so my answer to this guy was like, yeah, I think we need to kind of from first principles like reimagine the world of capitalism with different constraints.

And I think Bitcoin inspired Bit Tensor. And so it really should just be viewed as almost its own derivative technology using the same ideas that Bitcoin applies in terms of reinventing money, reinventing capital.

Bit tensor does this for reinventing incentives and reinventing capitalism.

And once you gro this and you start to sort of look deeper, peel the onion back on the progress that Bitensor's made over the last 3 or 4 years, it is immense the rate of progress and there's really nothing like it. There are no other ecosystems that come close to the rate of progress that Bitensor has incentivized permissionlessly in the world.

Something that Jake Steves, the co-founder of Bitensor, goes by const pointing out a week or so ago.

It's actually a really good way of understanding how profound Bitensor is.

It's the only blockchain or cryptocurrency or you want you want to call it permissionless digital ledger based technology that has incentivized non-financial utilities successfully at scale. Right?

So what does that mean? Bitensor is the only technology that has incentivized non-financial utility at scale.

So what is non-financial utility?

Well, it's something like inference compute or prediction models or synthetic data, right?

Or coding models.

These are non-financial products, infrastructure technologies that are basically reshaping the way humans make progress and the way that we learn and the way that we build experiences for other humans and the way that we shape our companies and organize society.

And obviously AI is kind of a makeup of those things.

Inference, compute, data, algorithms, and those are non-financial utilities, non-financial commodities.

And Bitensor is successfully incentivizing extrinsically with the the kind of economic incentive the production of those of those commodities.

And for the first time ever in the history of any application of cryptocurrency has that been done successfully and Bitensor is doing it. All other cryptocurrencies or all other blockchain products uh incentivize financial contractbased utilities.

So if you look at the Ethereum ecosystem or the Salana ecosystem or these kinds of projects also all stable coins, right?

So, stable coins, smart contract blockchains, DeFi, right?

The common denominator across all of those networks is that they are incentivizing financial products and services purely financial, right?

Where Bitensor is actually allowing you to incentivize useful abstract technology improvement in the world.

This is a profoundly different category of incentive.

You >> you're totally right. So it's first off, I like the idea this is going to happen of let's just get rid of the phrase cryptocurrency or particularly the crypto part.

You don't go around saying, "Oh, Amazon and Uber are two TCPIP companies." You don't say that.

You don't say. You don't even say that they're internet companies.

They're just companies. Bitcoin is a store of value. In many cases, it's a currency and blit Bitensor is is like if you go down the top 100 cryptos, I'm looking at them now.

Bitcoin currency, Ethereum, you know, essentially a programming language to make decentralized finance applications.

Tether a currency, XRP a currency, BNB financial exchange, Salana, you know, kind of a currency or kind of like Ethereum, USDC a currency, and on and on.

They're all hyperlquid. It's an ex an exchange to do derivatives trading.

Chain link powers stable coin transactions among other things and on and on.

Bitcoin Cash, Zcash, these are all D5 apps. And then finally down in the list of the top 100, you get to number 30 and Bit Tensor, the very first token in the top 100 that has nothing to do with decentralized finance.

And and what they do is they they're creating right now we're at kind of like the seed stage, but they're creating now some of the most amazing companies and technologies on the planet.

And it should be emphasized again, these are not just AI companies. I think, you know, I can see why the fascination with with AI probably Jake, who you mentioned and the other founder, they probably were fascinated by AI and AI was one of the fastest growing technologies out there at the day tensor was released.

So why not focus on AI? But let's not forget first off at one end of the scale there's there's Bitcast which is an ad agency a decentralized ad agency supplanting the trillion dollar a year centralized ad industry.

There's Better Therapy which is uh uh has nothing to do with AI.

It's it's therapist connecting with customers.

It's like an online therapy program with more incentives than the traditional online therapy companies out there. There's at least three companies that have subnets that are in the cyber security space.

So, here's a platform and the minor rewards are bug bounty. So, the bug bounty business is a very successful industry.

Helps companies from all sizes solve their potential bugs and opportunities to hack into their systems. So, there's at least three bug bounty subnets or cyber security subnets. There's the bank compliance subnet.

Giannes uses AI to create data sets, but it's not really an AI company.

It's more of like a a a company that banks would pay to see if they have the right compliance.

So there's many I would say at the end of the day when the dust settles there's going to be probably at least 50% or more of the subnet are going to be non AI.

Still AI is is probably the biggest growing and fastest growing but you know there's cara which is like a subnet which is like a hyperlquid you know talifi which allows you to sp buy tao using USDC.

So there's some financial subnets as well. So it's really you know think of think of a business this way.

A business has three components.

There's one side which has a lot of supply of something, another side which needs some of that supply and a platform in the middle to help with discovery and transactions and customer help.

And so it's like Uber, there's on the one side there's people who have empty seats in their car there on the other side there's people who would like to be in one of those empty seats and then the Uber is just the platform in the middle that connects the two. And what what Bit Tensor and Tao does is it basically abstracts that platform. So anybody can create a company and have built-in resources to provide incentives to m you don't need employees anymore.

Now you have miners all over the world willing to contribute your resources kind of in a way artificially that Uber you could think of Uber as the drivers are the miners and Uber is both the validator and the platform in the middle that connects customers who need seats.

So so Uber is like what a subnet would look like.

Bit tensor abstracted that away.

So anybody can basically create an Uber or an Airbnb or an open AI which is you know open AI is really on one side you have people who had great AI models and on the other side is people who need great AI models. Oh which is another one more point I want to make on this.

Open AI has seized the public fascination as has you know Grock and Deepseek that you were talking about earlier.

But these are not in the future going to be the most important internet companies just like Cisco which was the highest valued company in 1999 is no longer the most important internet company. In fact it's barely relevant at all. Open AI, yes, it provides a useful resource for consumers, but there are going to be niche AI companies that provide much greater value that are going to, you know, cure cancer and and, you know, stop financial fraud and do all sorts of other things that AI companies are good for.

And I'll just mention the subnet metanova, you know, using AI to find proteins is kind of in this category of niche AI company that could be gamechanging for the entire planet, but it's not, oh, consumers don't chat into don't can chat into it every day, so it's not as well known. But anyway, that was all my points. Well, I think Google, if you think about the way humans discover information, there were 20 or 30 search engines before Google was created. And if you were to say, >> "Look, we're not going to have any more innovation from startups.

Startup innovation is done. No more startups.

Just call it a day. Everybody stop doing your startup.

" Google probably wouldn't have been given the confidence or the success or the capital that they ultimately produced massive amounts of.

If that was the narrative, right?

If that was in the air in the early '9s because we had Metacrawler. I mean, you probably remember this. I mean I think you've started a few internet companies James you know Metacrawler Alta Vista Ask Jeves right like early early internet companies >> there was ei galaxy which was a weird search engine that AOL was using and then of course there was the very first one created by I think it was Tim Berners Lee which was the worldwide web worm.

>> Tell me about the worldwide web worm.

What was that one about?

>> The basic idea was it was an open-source uh like spider. There was this thing called the web and team Burners Lee created a spider that could crawl the web and return all the information on a page.

It was the most basic search engine.

So if you typed in the word motorcycles, it would simply just scan all the pages that the crawler had gathered that the spider had had gathered and see if the word motorcycles exists on that page. The more times the word appeared, the higher ranking that page would be whereas Google completely changed that approach. Well, I I think you know just to kind of pick up on something you mentioned which a lot of people might interpret this just wishful thinking.

You know JJ and James are just crazy optimistic pie in the sky. They think this bitter is going to work out and be this gigantic thing.

Well, I I would just say look at the success of open source.

How has open source become so dominant and so disruptive and transformative to every single category it's been successfully applied to. Right?

By the way, including AI right now, the open source models are state-of-the-art in almost every domain and they've gotten their orders of magnitude more compute efficiently.

And the dominant approach being adopted by China is to release these open models and they are absolutely dominant.

All the startups in the west in particular the United States are heavily using these Chinese-based open models right and so but let's just look pre-I pre 3 years ago right pre November December 2022 because we're now in November of 2025 open source has also been very dominant rapidly in every area it is applied to and You can just look at the trend line.

The proprietary incumbent approaches to creating technology eventually Wayne and get fully replaced by the open- source approaches.

And yeah, I've been building my career on this on this thesis.

I think it's empirically true if you study the world and technology markets going back many decades.

And infrastructure software tends to be the most obvious place.

You have Linux with operating systems. You have open- source databases and you have many middleware infrastructure tools.

I think this is happening gradually more and more with applications.

But Bit Tensor transforms open source.

It basically says we're doing the same thing.

We're using the power of the internet and the global potential of humanity dispersed across all these different places.

It doesn't matter where you live. You could be in Zimbabwe.

You can be in Canada and South America, anywhere on the planet or maybe other planets when we start colonizing other planets, right? And you can participate in moving humanity forward.

You basically just have to mine this permissionless decentralized network and you're rewarded commensurately for doing so.

And as a result, you are then given the freedom to be your own boss and to not go and work for a company or perhaps even to create your own company. In many ways, being a minor on Bit Tensor successfully means that you actually never even really need to operate within the conventional equity-based capitalist system because you're actually rewarded with direct incentive as opposed to this indirect incentive where it's getting inflated away and you're getting diluted.

And also you have to abide in terms of the world of equity.

You have to abide within the laws of the country that you created the equity in.

A lot of people, you know, quickly confront these things when they choose to become equity-based entrepreneurs and capitalists.

>> Well, if you incorporate your company in America, you're typically going to choose a place like Delaware and you're operating within the Delaware court system, right?

If you just delegate all these things to a lawyer and you don't really care about what you're doing at the implementation detail level, you might not actually realize what you're stepping into.

But your business is literally governed by someone else.

Even if you own 100% of your business, the cool thing about a permissionless network like Bit Tensor, Bitcoin is that your incentive is not governed by someone else.

It is influenced and controlled by algorithms that are mathematically predefined and where the network itself is fully decentralized and not controlled by any single entity.

That's pretty awesome.

And that's what gives Bitcoin the resilience it's earned and developed into over 17 years. And I believe that's the trend line for Bit Tensor.

But, you know, we just have to continue to see more evidence to your point of state-of-the-art, very successful subnets out competing, out innovating the proprietary, centralized, highly permissioned, and in almost all cases heavily venture capital subsidized uh incumbents.

And we're going to continue to see more and more people understand and realize that Bit Tensor is a very legitimate and extremely scalable path to continuing to accelerate the pace of human progress and AI.

And I guess another thing that I think more people will also realize, right, especially as anthropic and open AI become more and more dominant, which I do expect to continue to happen as well.

I announced a week or so ago that you can no longer ask medical questions and you can no longer ask for financial advice.

You're kidding. I'm pretty sure a lot of consumers really want to be able to talk to their model about some medical advice maybe with some disclosures and disclaimers is okay.

And I'm pretty sure they would also like to ask their models for financial advice on different topics as well, right?

And so more and more what I think will be the default position of anthropic and open AI and all the leading proprietary closed models that really want to become what Google is to search over the last 25 years is that they will become more and more politicized.

They'll become more and more woke, more and more politically correct, more and more labbotomized, brainwashed and narrow-minded in topics that are popular for the mainstream that will further entrench those incumbents.

And I think that the reality is as people try the open models, as people become more aware of how do you systematically build AI that can actually be far more disruptive and transformative because it's built by the collective of humanity as opposed to 20 people in PaloAlto or 50 people in San Francisco and you know thousands of unnecessary headcount.

That's really at the end of the day there's like 50 to 100 people that do the vast majority of the core fundamental work at these labs and there's thousands of people there that work that work there that are really just filler and they're just kind of unnecessary in my opinion.

I mean they're probably building data centers.

They're labeling training data and doing things, but they really probably don't need to be full-time employees because you go through these like cycles of training training the model and then you release it.

You do red teaming and then you're running inference and then at that point, you know, you're just running inference at a large scale for serving the prompts the the API queries and the prompts by by consumers on mobile apps or on their desktop or whatever.

And so I think the more people realize that like these models are getting heavily censored, the more people will start to look for alternatives.

And I think that that's going to be a more powerful dislocation for sort of like resetting the fundamental economics of this AI transition that humanity is going through in terms of markets far more so than you know there's a lot of overpriced assets, there's a lot of venture capital exuberance, there's a lot of overspending. I think that's going to be I think that's actually going to be more disruptive than the sort of like overinvestment and glut of overspending going on on the infrastructure if that makes sense.

>> Yeah. And look, you you bring up a really good point about how open source is like the basis for so many of these trillion dollar industries.

Like it's almost like AI right now with the open AIs and I think we're being almost overly critical of these big, you know, gigantic AI valuations, but the reality is we're dead on correct.

But it's almost like AI has done it in reverse.

Let's create the big valuation and then build into it. Whereas, you know, think about the internet started off as something nobody cared about, then it went into, oh, this is useful to send messages back and forth. Oh, here's a web browser.

And it started to be it actually earned its value over the decades.

And and same thing with Linux, you know, and same thing with all the GNU Richard Stallman, you know, projects.

But if you think there's a big difference though between open- source and and bit tensor which I think is really important.

If you look at all the creators of these open-source industries like let's take Lionus Torvalds at at who created Linux. So many trillions of dollars in value have created off Linux.

Both the the Windows operating system and the Mac operating system are based on Linux.

But it's not like Lionus Tolv is worth trillions of dollars. He was b I forget what happened. And he was basically was did he become like a Google you know scholar or or >> well actually no my friend Jim Zlin who was a marketing executive working at a open source company for Apache web server realized in 2004 or maybe 2005 that Lionus should have been compensated more and sort of provided with a structure and you know Linus didn't really care too much about capturing all this value.

he cared more about creating a lot of value. Linus ended up partnering up with Jim to help create the Linux Foundation.

>> And so the Linux Foundation became very successful and it's a great foundation.

They've created a lot of kind of standards and ecosystem investment in open source over the years.

And you know, they generate something like on the order of two or $300 million a year in revenue.

It's a very profitable nonprofit as many nonprofits are in case people aren't aware. And so I believe Lionus has paid 7 to10 million a year from the payroll of that foundation and he's been paid that for something like 15 years or so which is great and now but again you look at the value that Lionus helped contribute to the technology industry by the way not just with Linux with Git which Lionus also built which is a versioning system for how you actually version software because he's a very antisocial person.

So he's like I want to create a better protocol and a mechanism for how we version files for software code and he created git to basically make that more systematic also was motivated to do that because the previous version control system used by the Linux community was proprietary and I believe I believe it was called BitKeeper or um forget the name it was another version control system that was actually closed source.

He was like, "We need to build an open source one, of course, because why is Linux, you know, the dominant open source project using a proprietary version control system." But Lionus really, you know, if you look at even if he had 1% of the equity of GitHub, GitLab, Red Hat, all the Linux companies, you know, he'd be he'd probably be a billionaire just with 1% of the ownership of those companies.

>> And and this is my point, though, that that okay, he would have had to work that out for himself. By the way, Mosaic, the first web browser that allowed graphic displays created by a young college student named Mark Andre, he also had to create for himself a separate entity, a separate company by which he started making his millions.

That was Netscape. And of course, now he runs Andre Horowitz, the big VC firm.

And similarly Tim Berners Lee he didn't make a company about the web or else he'd be worth like $30 trillion or whatever but he was rewarded in some artificial way you know somebody com I think this one might have been Google gave him some reward or whatever but with Bit Tensor with Tao it's open source but there is financial incentives there is reasons why a subnet wants to would want to go on Tao and Bit Tensor there is a reason why miners would want to mine the subnets because they get rewarded either in Tao or in tokens denominated in Tao and Tao has intrinsic value.

Yes, it's open source like all these quote unquote crypto tokens but you know Tao has value because you need you need to buy Tao in order to register a subnet.

You need in some cases to buy Tao I think to be a validator or a minor I forget which one. you need to to buy, as we spoke about earlier, you need to buy Tao to buy the subnet tokens of the subnetss and that is sort of a replacement as we've discussed in other episodes, that's a replacement for the e buying equity in these subnets.

So there's there's huge enormous reasons why the platform gets denominated in tokens where there's legitimate buying pressure.

There are reasons to buy Tao.

It's not a memecoin. There are people who need to buy towel to perform certain tasks.

And so that as the value of tow goes up, it's it's this flywheel where as the value of the towel goes up, it becomes more attractive to register a subnet and more it becomes more attractive become a minor becomes more attractive to buy subnets which in turn creates more buying pressure on Tao.

So here's this amazing thing where it's this open-source platform for creating it's an open source platform for general entrepreneurship that many companies are launching off of and the ways in which to transact in any way on this network is through a token that is rising in value which further creates and escalates the incentives to participate in it.

That's why it's an amazing thing.

It's open source 2.0 No, because it's this evolution from the last three decades of >> I fully agree, but something that you mentioned makes me want to add a bit to it, which is it's an opt-in system.

You have to decide to participate.

You're not born with tow. Right.

>> Right.

>> You need to wake up one day, listen to the tow pod by James Altter and Joseph Jax.

And you need to learn and educate yourself.

And you need to have an open mind and have an intellectually curious mind to really kind of like unlearn and reset the conventional wisdom of going through business school, learning most basic concepts of capitalism from an early age and sort of never really questioning those things.

I think education is the most fundamental and important thing that is really representing the barrier between mainstream understanding and adoption of what Bitensor represents in terms of this disruptive potential, its transformative potential and the people that are going to stay stuck in the right in the in the conventional capitalist system.

Such an important point.

Such an important point because when I talk to people not only about Bit Tensor about Bitcoin, they're stuck not what quite everybody wants to think that the system oh it's been around for 100 years.

It's got to be correct.

It's got to be perfect. People need to understand there are a every industry on the planet evolves including the financial system including entrepreneurship and so on.

But people need to understand there are as good as banks are for us we keep our money there.

We can use credit cards.

we get interest but you need to understand structurally and culturally the flaws of banks and the banking system in order to really appreciate what an innovation Bitcoin is for instance and with entrepreneurship you need to understand you know we discussed in the last episode what the word permissionless means this is just part of it but you need to understand what the flaws are in starting a company in the US or in any country and the VC funded model and the public markets model they're not really public markets they're just called that it's really for the general public.

VCs don't really take venture style risk.

A lot of it is is baked in and very scammy.

So, you need to understand what the flaws are instead of just buying into the system. And you're right, it does take it does take education as well as these financial systems. I think that's a very important point.

Yeah, people when they first encounter Bitensor and they start hearing, wow, it's kind of reinventing equity, company equity.

Wow, it's reinventing the way we do AI, you know, training and compute and building out all these all these technologies, incentivizing them.

What are incentives? We have intrinsic incentives, we have exttrinsic incentives.

There's all these concepts.

And what I would just say is we're living in a moment where everything is changing, where the way humans are going to coordinate between each other is going to change in massive ways.

the way value is going to be represented is changing rapidly and the way and something Ray Dalio has been talking about about the reorganization of power in the world and sort of the new world order.

I'm a big fan of what he's been talking about on those lines for many years now.

I think there's going to be new power structures in the world at the government level at individual levels and I think if you're not educating yourself very seriously and questioning the sort of structure of the existing systems you're going to be left out and I think that that's a really critical thing.

You know something I proposed which I I really hope we can kind of make progress on at some point and and you know maybe hoping is not the right way to do it. We just have to be systematic and figure out how to make this happen, James. But you know, if every human who had some ability to dollar cost average buy tow at a level of5 or $10 a month, something nominal, we would be able to include all of humanity in this system. And I I I would wager there's about a billion people on the planet that could afford to do that.

And so, you know, if we did that at a level of, you know, call it five or 10 billion dollars a month, that's a billion people times five bucks is 5 billion a month. Let's say it's 10 billion a month. You know, it's 120 billion a year. It's really actually not enough.

We need more than that.

But I think that that would be probably a quarter or maybe 20% of what we really need to accelerate at the level of capital investment by Google, Amazon, Apple, Tesla, etc., Microsoft combined, which is, you know, they're investing almost a trillion dollars a year combined and next year will be a trillion dollars a year, right?

and the sum total of humanity coming together and doing that, at least the people that can afford to do it, which is probably around a billion knowledge workers, right?

The people who make, you know, enough to like sort of say you have spare income of probably a hundred or $200 a month at least, maybe even more than that.

Call it 10% of your spare income.

Dollar cost average buy tow.

It'll be a great store of value and a way in which you can actually include yourself and your future generations in permissionless open continuously improving intelligence and uncensorable permission.

P permissionless meaning borderless meaning it's not controlled by a single government government or entity or corporation.

And it's probably the best highest value long-term investment you can possibly make.

like what what's the ceiling on the value you want to put on continuously improving open permissionless knowledge intelligence for the future of humanity there's no price you can't put a price on that I have I have the be I was just thinking as you were saying this I have the beginnings of a solution to how you I don't want to say extract this money but again everything boils down to incentives what's the incentive by which someone should do this so imagine I have a kid I have several kids but imagine I have one kid and I'm given the choice you want, the goal is you want your kid to have the best education possible.

And so I have two choices. Send the kid to Harvard or he spends four years um and he can and during those four years he has to be a minor for one or two subnets.

He has to be a validator and he has to build a subnet. That I I would always choose the latter because you're going to get such an education in entrepreneurship, in incentives, in negotiation, in pricing, not all in economics because there's many economics issues when when >> gameory game theory >> game theory completely a game theory.

By the way, many of the subnet owners 50% of their time is thinking about how to deal with people trying to exploit their incentive mechanisms. So you have to think about game theory all the time and by that's the real world.

They never teach you about exploits and corrupt businessmen in college. It's all an idealistic world in college.

So building a subnet being a minor for let's say one to three subnets being a validator for subnets.

You're going to learn coding, you're going to learn technology, you're going to learn economics, you're going to learn incentives, you're going to learn management game theory.

You're going to get such a great real world education.

Tao University. Someone just set up Tao University to basically in, you know, explain to people the benefits and maybe we could do this, but explain to people the benefits of doing these activities on Tao.

My kid will be so much more successful if I put him through Tao University than if I send him to Harvard, where okay, he'll get good connections, but that lasts for so long.

Without skills, you're nothing.

These are the skills you'll learn from the ages of 18 to 22 when the mind is still molding.

you're going to be so far ahead of everyone who didn't learn those skills between 18 to 22. Like that's how and then okay if if you want to be go through the program and be a certified TA University graduate then that's $5 a month, $10 a month and you get a better education than spending, you know, $80,000 a year at Harvard. I think it's already there.

You register some tow, you lock it into a subnet and then you're if you're successfully mining, your ROI on that is immense. But you do have to put up some investment.

I think Bit Tensor is continuously improving curriculum that requires you to learn a wide range of topics like you said technology, game theory, community development communication navigating the scammy things from the things that are actually creating and producing value and real utility.

And those skills, you're getting paid and rewarded to do that. It's sort of like a paid university degree that never ends.

is that it's a lifelong thing as opposed to going and getting massively in debt and getting politicized regurgitated memorization of tests and then getting a score that is for all intents and purposes a very poor representation of your life potential. A very very lacking indicator of your success. a lot of people that have degrees from these elite universities and many of them are some of the least impressive people I've ever met in my entire life. Some of them are really amazing for sure, you know, I'm not going to say all of them.

A lot of them are really awesome people, but many of them, you know, they they got the same degrees and they studied the same topics and I'm like, wow, something's probably not working there.

>> Yeah, 100% agree. And I think I first wrote about colleges being, you know, dead in 2005 or the Financial Times.

And back then it wasn't a conversation like people were very upset at me.

I think now it's a legit conversation to be having with your children and it's more widely accepted and I think platforms like TA where it's so clear you get this worldview education.

Yeah.

>> From business models to economic all the things we discussed. I think this will create a more educated society the more people participate in this kind of ecosystem.

It's very exciting.

>> I would say many of the youngest people ju just to top this off because I think what you said is so important.

Many of the youngest people in the Bit Tensor project they could be minors.

They could be submit owners. They are really brilliant people.

They're super uber geniuses.

A lot of the same people I know.

But they're teenagers.

Some of them are not adults even. Some of them are 15, 16, 17 years old.

and they're incredibly brilliant because they get it.

They came from a video gaming background or they were really into programming competitions or something and they're just sort of like autistic and really into all of the economic angles of Bit Tensor sort of incentivizing things and they get it so fast, much faster than I did when I learned about Bit Tensor in early 23 and then I, you know, I I I didn't really get it.

It took me a year to fully comprehend everything as it was still evolving.

And many of the much younger people that I'm now fortunate to know and work with and consider many of them friends, you know, they for whatever reason they got it so much faster than I did.

And I think it just requires that kind of fluidity of understanding things and not being encumbered by all this conventional wisdom.

>> Yeah. No, I agree. I mean, I remember it was just in June. We met at I think it was like June 3rd, the beginning of proof of talk. the conference in Paris and you and I met for breakfast and you were the first person really in the ecosystem I was meeting face to face that meeting and then just the rest of the conference and then all the experiences I've had since this is like a really smart this is a high IQ group of people like all the submits all the people involved I feel just grateful like I've had such a great education talk you talking to all these people is geniuses you know and and I'm gonna take take advantage of that genius again which I want to ask you a more in the weeds question.

Now we discussed this briefly on text this week but this new uh and this is really in the weeds so people could pay attention or not but this new idea of allocating Tao to subnets through the net flows I understand conceptually so every day or every block tow is allocated to subnets based on how good they are and they would measure how good they are previously by what the price of the subnet was now it's being measured and I think this is correct now it's being measured by how many people are interested in staking in that subnet because that shows longerterm beliefs in that subnet.

So that subnet is potentially worthy of more tow being allocated to it to pay miners and have less volatility and so on. But what I want to understand better is will this penalize the subnetss where the miners are using more expensive resources so are inclined to sell their rewards, their subnet tokens a little bit more quickly versus the subnets where the miners are more incentivized to stake their rewards rather than sell the rewards to pay their bills.

I would argue that you have to ask the question slightly differently because I think there's ways in which the answer could be yes in both cases depending on which subnet you asked the question about.

So, I think what you were texting me about, and I I kind of gave you my off-the- cuff answer, which maybe needs a little bit of revision, but for pure commodity subnets where you're just incentivizing inference tokens, or maybe you're incentivizing data, training data of some kind that's like fairly trivial to acquire.

Maybe you're scraping Twitter, maybe you're scraping Reddit or something, or you're you're you're incentivizing compute resources, maybe like very abundant H100s or AA100s or something.

Maybe not black wells, but like more more widely available compute.

I think you could make the case that TOFlow requires those subnets to have slightly different incentive mechanism designs in terms of how the validation works versus subnets that have less commoditized outputs where you're really incentivizing something that is already inherently very scarce and you and you need a less rigorous incentive mechanism designed because it's just like look just give me this thing and you know the world has very little of it and so it's pretty easy to measure that you provided that thing right depending on how it's actually produced so the abstract way I've been sort of thinking about tflow which I'm a fan of I think it's a good protocol upgrade and I think it does basically acknowledge the scarcess of tao there's 21 million tow now it's getting emitted as of basically a month from now it's going to be emitted at half the amount it has been for the last four plus years which is 7,000 200 tow per day.

So we're going to be going down to half of that, which means everyone in the ecosystem is going to need to sort of understand that their incentive design has to be that much more precise at a minimum, right? Like you need to get far more clear on how you verify and validate that the work and the production of the work was done in a way that cannot be exploited or or hacked or manipulated.

This is a big problem.

exploitation and what I would call value leakage across almost all the subnets is definitely an issue. I would make a couple of comments about how to mitigate this.

Towflow is definitely one step in the right direction. Another thing I think that we need which I think is going to be a big next phase of refinement for the tensor's protocol is the degree to which you can basically be more precise and expressive about how to allocate the emission and the tow the people who participate in creating your commodity your subnet output and so by the way primarily very very obvious who those people are it starts with an N it's the miners right no surprise the people who are doing the work should be rewarded with most of the incentive Right.

Today there's a pretty common standard across all the subnets where the subnet owner gets 18%.

The validators get 41% the miners get 41%.

So 18 4141 right a lot of reasons why validators might not actually get paid 41.

There might be different reductions in that or there might be burning or redistribution of that back to the miners.

And so but in a lot of cases the miners are not getting rewarded more than let's say 50 60% of the total emission going into a subnet right and so as of today right before the first having next month let's say you're getting 1% of the the tow emission which would be around 70 tow per day 70 tow per day if your miners are really only getting about 60% of that per day right like that means that the protocol has a sort of upper bound on what the miners can directly be paid based on the emission percentage right that stakers are actually increasing.

So if you stake more into a subnet in an available liquid supply of tow the higher up the rankings on emissions that subnet will go they're going to earn more emission.

But the big thing that I've been trying to kind of unpack recently is as subnets become more and more clear on how to measure the value of the commodity of the output from the subnet.

What they should really start to internalize I I believe this is a journey. I don't think it's a binary thing. I think this is very important for people to understand.

It's not like we're going to figure this out tomorrow.

We're going to figure this out next week or next year. I think this is going to take some time, but they need to understand that they should aspire to be like Bitcoin.

They should aspire to have the same degree of measurement precision that Bitcoin has in rewarding the miners on Bitcoin.

And so let's look at Bitcoin just as a case study and example because Bit Tensor was modeled and designed 100% after Bitcoin in terms of the fundamental architecture, right?

No fair launch, token supply constraint, admission schedule, h havingings, lots of attributes.

Bit tensor extended on that and embraced proof of stake where you basically have stake weighted allocation and rankings of the way subnets are actually organized and the way emissions flow, right?

And so there's a proof ofstake proof of work hybrid. But if you really just look at Bitcoin and the one incentive mechanism that it has, which is rewarding people to compute a SHA 256 hash, not people, but that that the machines that people can organize, these AS6, right, that that compute the SHA 256.

When you show the proof of the hash to the network and it propagates, there is no ambiguity that the work was done.

It is 100% cryptographically certain that minor who might own one ASIC or 50,000 AS6 doesn't matter organized those machines off of the network to compute that amount of SHA 256 hashing and they demonstrated by showing the hash proof to the network and if it computes successfully in Nakamoto consensus which is a deterministic measure of did it actually reach below the difficulty threshold of that block height, you then get rewarded in Bitcoin, right?

It is a deterministic cryptographically airtight incentive mechanism with no exploitation potential.

You cannot exploit that.

The only way you'd be able to do an exploit is if you did what's called a civil attack on Bitcoin's network, which is where if you controlled more than 51% of the hash rate and then you could actually, you know, manipulate the Coinbase.

You could print more Bitcoin. You could delete tokens.

You could distribute more.

You can change supply. You can do all kinds of things.

And that would be obviously catastrophic.

What Satoshi had predicted in the Bitcoin white paper was that there would be a disincentive to actually manipulate the network if a 51% attack were ever possible by the attacker because they would have a a higher incentive to mine the network fairly than they would to actually take over the network because they would be rewarded and paid more to just mine the network as opposed to undermining all of the other tokens. The price would plummet if everybody thought that it was attacked.

>> Yes. And so you can coordinate you can also coordinate civil attacks with several entities.

And so but Bit Tensor's network has a subjective or fuzzy verification design which is how do you create continuously improving intelligence?

Well, it's kind of hard to measure that with a cryptographic hash that's perfectly precise. But I would actually argue we should attempt to go more in this direction. I think a lot of the subnets are computational in nature where you're producing some inference tokens or you're producing some training data or you're training models from scratch and these are computational processes and I believe that actually is possible to reach a point in time where we can really more closely mirror the symmetry of economics on Bitcoin which is by the way there are no separate validators that are paid to validate Bitcoin blocks.

Miners validate through the process of mining on Bitcoin, right?

There are full nodes that validate Bitcoin transactions.

They are paid zero.

They get 0% of the Bitcoin emission.

100% of the Bitcoin emission goes to miners that are presenting proof of the SHA 256 hashing that happened on some machines that those miners organized.

100%. And by the way, Satoshi, again, there is no premine.

There's no 18% allocated to Satoshi.

So, what I think my big grand prediction, I'll just put it here, posted a tweet about this a few days ago, is I think in the next year or two, we're going to start to see early signs of ranking the most long-term durable and disruptive subnets on Bit Tensor is when they start to look like Bitcoin. when the owner percentage goes to zero, when the burning and the recycling of tokens stops to occur because you don't need to do that.

Your incentive mechanism is perfectly precisely measuring something where the percentage of emission you're getting from tow itself, the overall TOA network starts to become the real price reflection of the value of your subnet and the commodity that your subnet's producing.

And validation is part of the mining process.

Why do we have separate validators?

Why would we need to have separate validators?

If in the process of mining, you have so precisely measured that check and proof of the miner's contribution also verifies that the work meets the standard of quality in some increasing difficulty threshold right function.

You don't need a third party or a separate persona to do the validation.

You can do it in the same process.

Therefore, the miners would be paid 100% of the admission, 100% of the incentive.

That's where I think things will eventually go. Again, it's not going to be, to reiterate, it's not going to be an overnight thing.

It's not going to happen in a week or two.

It's not going to maybe even happen in a year or two.

I think it's a gradual process where again the indicators what I would argue would be the indicators would be margin compression on the subnet take rate significant margin compression on the validator take rate and margin expansion and increasing saturation to the point of being fully saturated of minor take rate of the emission going into the subnet. But the question is like while that's happening while that pro process is happening which figures out how to reward miners more substantially.

Yes. Yes.

>> What's going to happen when we're fully in this new net flows model where like take the subnets that provide compute?

>> Yeah.

>> Those those miners have electricity bills the next day that they have to pay.

>> Sure.

>> And so they have to sell just like with Bitcoin miners, they have to sell some portion of their in tokens that they burn through incentives to pay their utility bills.

But there are some subnets in the tow ecosystem that don't have electricity bills that don't have to pay for their resources.

Their mining is maybe more intellectual.

Are those subnets going to benefit more from this net flows model and the compute subnets or inference subnets? Are they going to be hurt right away? potentially, but I think it's a temporary transition based on how wellpriced that commodity is and importantly how well priced the derivative prerequisites for producing that commodity are. Right?

So, for example, to produce the current block height difficulty of SHA 256 hashing on Bitcoin, you need quite a lot of electricity to do that. You need a lot of power and cooling for the machines that are going to be running that hashing and you need a [ __ ] ton of electricity to be able to support that.

What's interesting about Bitcoin is it's actually a mechanistic pricing market for the cost of both designing the chips to do the hashing and to run them and to pay for the electricity to run them and sort of like vertically integrate a single product that's perfectly priced on that.

Interestingly, you can run the math on this in terms of the cost per terahash relative to the cost per jewel, the electricity cost per jewel.

And there's lots of statistics for Bitcoin mining out there that cover this. And it I think it's almost exactly perfectly pricing.

like you basically sell the Bitcoin and you'll pay for exactly what it would cost to produce enough wattage, electricity, and hardware to earn a block of Bitcoin.

Now, what's missing in that though is the degree to which Bitcoin is representing capital, right?

If that happens faster, if the velocity of the re-representation of capital eclipses or out outweighs the velocity of pricing electricity and pricing AS6 to secure and represent that capital, which is SH 256 hashing block height, you'll then have an asymmetric differential of the value of Bitcoin per, you know, sort of unit of electricity needed to produce a block, right?

And and and that by the way that that happens quite regularly which is why a lot of the big Bitcoin miners go and say okay look we think that that's actually going to be a trend that the basically the absorption potential of Bitcoin of all other capital assets will happen at a faster rate than Bitcoin can perfectly precise precisely price the cost of doing the mining that that what what the big Bitcoin miners have done as a consequence of that conclusion which is kind of like a speculation is they get a lot of debt and they say, "We're going to get all this tons of mountains of debt and never sell the Bitcoin because we think the Bitcoin is going to be worth so much more in the future.

" And unfortunately, because of the degree of volatility of the Bitcoin price, the timing has been off on that. And many of those big miners have had to sell.

Many of them have not gone out of business.

Some of them have gone out of business, but a lot of them have had to sell some amount of the Bitcoin to pay for the facilities for the debt or to pay for, you know, margins that they needed to reach or whatever. And so I think to your point, it is true for certain more expensive commodities on Bit Tensor.

It's just going to be it's going to be more difficult for those for those subnets to really scale. But I would actually argue right and this is kind of the way to really look at it.

If you think from first principles like Elon does with his businesses about how to design something using the absolute highest potential, most fundamentally well engineered design constraints uh and organizations of atoms and materials, right?

You can make almost anything 10 times cheaper, 100 times cheaper, maybe even a thousand times cheaper.

If you work and you say the only boundary and the only playbook and rulebook we're looking at is the laws of physics.

If the laws of physics allow for it, can we re-engineer and reorganize the atoms, the atomic organization of things that create a product more costefficiently than that product is currently made today.

That's really what Bitensor is all about.

It's saying we are incentivizing permissionlessly economic incentive for people to create the same products and services used by humanity at large within the laws of physics.

The reason Elon's businesses are so successful, even though he's using fiat and equities to represent his businesses, is the guy thinks from first principles for breakfast.

He wakes up every day and he's like, "Yep, that's the way we're that's we're going to do everything that way.

" Most people don't think this way.

It's really hard for most people to think purely from first principles in terms of how to engineer things.

But Bit Tensor kind of basically kind of says like well you're not going to succeed as a minor unless you're actually doing something from as close to first principles as possible today.

Otherwise another minor is going to come along out compete you or another subnet that's better designed whose incentive mechanism is better measuring the value of something at some level will out compete you and then it'll just continue and you know that that's actually a really good thing. So once again, you've totally changed my mind on on one of these issues because this is this has been the latest issue I've been worrying about.

But what you're basically saying is that yes, things are changing. And by the way, I also I agree with this new model.

I think it's going to more correctly represent what the true value is of each subnet.

And but what you're saying is it also then is going to provide an incentive for miners who are competing with each other to more efficiently produce their commodity from first principles thinking.

And by the way, all the subnets are asking the miners to do hard things.

So it's not like it's easy for some and harder for others.

It's just now there's new challenges for certain categories of miners.

They have to produce more efficiently uh you know per subnet token that they're getting.

Correct.

>> They have to produce more efficiently the commodity that they're producing whether it's computer inference or whatever.

Yes.

>> So it's it's interesting. So so it shouldn't really >> in the long term or even the medium term have an effect.

Maybe >> I'm very I'm very positive on it long term.

I think it's situational.

I think it's always for protocol level changes.

You always have to ask which subnet are you assessing, you know, the question against because there's a lot of subnets that are doing different things and you have to look at how they're measuring their commodity.

You have to look at what the miners are doing, how they're verifying things.

But Bit Tensor I is a sort of a permanent infinite game of capitalism. And in the world of equity, entrepreneurs have to choose whether they're going to play a finite game or an infinite game, right?

Like the the best entrepreneurs choose to play infinite games and they have to think about generational transfer.

If they don't live hundreds of years, many of the best companies have to instill these principles and then future generations take over.

But Bitensor doesn't have a board of directors.

It doesn't have a shareholder base that's sort of taking votes on major decisions.

It doesn't have this kind of bureaucracy.

It says there's an incentive.

You focus on the design of the incentive.

And if the world values that incentive in a way that is more than the cost of producing it, that subnet's going to do extremely well because Bit Tensor Tal can basically subsidize it using extrinsic permissionless rewards.

It's not a B system.

It's not donations.

It's not some subsidy. It is pure directed incentives for continuously improving innovation.

And that is extremely profound.

And it's kind of like unlimited and unstoppable.

Companies have to constantly grapple with this.

Well, [ __ ] who's the CEO, right?

The CEO is going to be there. If they're not running the company, you know, maybe the company dies.

Governments deal with the same thing.

Even in China, right?

like you have a big monarch or someone who was elected maybe for whatever reason and then they're there for 50 years.

Well, if they die there's going to be a big change in control and power because there's new policy changes.

United States has it and even in a more brittle way every four years we have this split political system where there's this bifurcation of ideology and on one hand you have people that want one thing and the other thing and now the government's been shut down for almost 2 months and everyone's missing their flights and the whole country is probably going to plummet to 1% growth from what it was a couple months ago which was probably 2 or 3%.

So it's really important that Bit Tensor continues to refine these kinds of like minor changes that allow for the incentive distribution of Tao to be allocated in the most waste free manner and to reduce the amount of waste because initially there's quite a lot of waste to get people to come in and understand how the system works and now there's less and less waste because obviously potentially becoming more valuable, right?

is becoming worth tens of billions of dollars and then we'll get to hundreds of billions and trillions.

You know, think about how the United States developed at the very beginning is like the dollar was just created and you had like Alexander Hamilton who was like a merchant commodities merchant.

He actually had the Bank of America, first Bank of America and he was a banker and then they were like, well, look, we've got to print money for the society and we've got to create commodities and financial derivatives and products in an exchange.

And New York Stock Exchange has created an enormous amount of waste.

And then the government has to start grappling with being this new powerful entity on the planet.

And so now we're helping other nations and we're sort of printing our currency to sort of get out of all these situations.

that our currency while the nation's economic prosperity because of private capitalism and free markets has grown from 100 billion to 50 trillion in the last 100 years.

The purchasing power of our fundamental currency has dropped by 95% in the same period.

It's dropped by 95%.

The purchasing power of the dollar has dropped by 95% in the last 100 years. I think it's like 70 or 80% in the last 50 years.

>> This is such a sad thing. We cannot have this happen for the future of free market incentives to create permissionless technology and permissionless innovation.

We need more durable fundamentals.

And I think as a new emergence in the history and the timeline of money, Bitcoin has served as a really great example. It's working.

It's actually working. It's 15 years old, very young. Actually, 15 years is not long, right? It's actually quite a young system.

It's worth a couple trillion dollars.

But we have to take ideas from this to say, is it possible to create an incentive for continuously improving digital technology and AI, but also all technology that maybe can last longer than the quarter of a millennia the United States has lasted. 250 years.

Next year will be the 250th anniversary of the United States of America since 1776.

And it's really not that long.

If you think about it, the most powerful and influential people in the world of financial markets, including our current administration, are quite pessimistic and negative about the situation with the dollar if it fundamentally doesn't change in terms of how it's represented and traded in the world.

And everyone says growth. We need growth.

We need growth. The solution is growth.

We're going to get out of it by growing by 5 to 10%.

AI is going to save everything.

We're just going to go to the moon and you know, we're going to grow, you know, 5 to 10% in the next 5 years or like, you know, 3 4% per year compounding forever.

And I have to say, James, I'm not I'm not convinced that we will be able to grow at that rate with the current structure that the nation has.

I'm really not convinced. I think this goes back to that first principles thinking which is okay so let's say instead of just thinking purely in terms of okay this is what the American economy was doing this year how does it need to imp what numbers need to shift so that we keep the stock market keeps growing you have to reinvent what is the economy what is the way equities are changed and valued and this is what the crypto world has done starting with Bitcoin but even more so I think with BitSense we started off this podcast saying well what are we doing here are we just being overly optimistic about Tao or Bit Tensor and so on.

And the reality is investors have to be professional optimists that even if you're say okay I'm down on the dollar I'm down on the current system so that might not be the best >> way to get the returns I need to make myself happy or make my investors happy.

So what is where can I be professionally optimistic because that is where I will get the best returns and what what we've concluded and that's how we met each other is that this tow ecosystem is the place to be optimistic and it's and it's exciting and I haven't been this optimistic and this excited about investment almost in my entire career which has spanned many decades now.

>> Yeah. I mean, I think uh to be optimistic is extremely important, but to place your optimism correctly is more important.

>> Yeah. Right. You can't be blindly optimistic.

Oh, the dollar is going to do this and that and we don't know.

We don't know that. So, we have to find where what what we do know is where to be >> we have to like you said, find where to place our where to stake our optimism and that's in staking towel.

>> Correct. Exactly. Well, this is a great show.

Thanks again, James.

I'm glad we've made it 11 shows so far.

>> 11 episodes. Sounds good.

So, thanks again, JJ. And I'll talk to you next week.

>> Yeah, we'll see you soon.