Bittensor Revenue Search 47 - Discovering AI start ups with asymmetric upside - Crucible Labs

Hey hey hey folks.

>> Hello.

>> We are the wrong side.

>> You >> better now. Welcome to Revenue Search, one of the most influential podcast live streams in all of Bit Tensor. I'm Mark

Mora, my co-host as usual. Laughing.

Laughing because we have the same intro every time.

>> I kind of getting pretty boring now, but it'll get better in next year. Your vis

what?

>> We also need to change our our intro there, but I still enjoy it.

>> True.

>> But we haven't been on We haven't had a revenue search for for what feels like ages.

>> I know it's been that long. I've been

growing a beard if since we last had a had a revenue search, but >> very impressive.

>> We've been very busy. Lots to announce in the future. But for now, I've been very very excited about this revenue search. And it's probably the a very

search. And it's probably the a very good one to open up on after our little hiatus. Um because we have Crucible

hiatus. Um because we have Crucible Labs. So it's this isn't a typical

Labs. So it's this isn't a typical revenue search where we're looking into a subnet. This is Crucible Labs and we

a subnet. This is Crucible Labs and we have one of the co-founders of Bit Tensor which is Ship or Allah and also David Lowi his business partner. So, um

I think without further ado, you love it when I say that. I lose. Story time.

Let's bring them in. Right. There we go.

Oh, this is the wrong way around. There

we go. You guys need to be on top there.

How's it going, guys?

>> Great.

>> Hey, guys.

Synchronized helloing. Love it. Um

[laughter] yeah. Um again, thank you so much. The

yeah. Um again, thank you so much. The

the honor is is definitely ours. It's uh

very much appreciated. And Allah loved Remy Search so much that he just he left. He's disappearing. [laughter]

left. He's disappearing. [laughter]

Where's he gone?

>> I'm out of here.

>> Yeah. Anyway, well, whilst we're waiting for Allah, David for who are you first for? Because like what what I've seen is

for? Because like what what I've seen is that Crucible Labs works very much behind the scenes, but you do do a lot behind the scenes and like new people

into Bit Bit Tensor probably won't be familiar with um with what what you guys do. So yeah, who who are you David and

do. So yeah, who who are you David and Alla? And then also what what does

Alla? And then also what what does Crucible do?

>> Allah, take it off. Take it off.

>> Sounds good. Yeah, I'd probably get started. Uh so for those of you who

started. Uh so for those of you who don't know me, I'm SH Alla. I've been uh on working with Jacob on the Bit Tensor ecosystem since about 2019ish.

Uh basically, you know, can call me brother contributor as well, part of the founding team. Uh we went through blood,

founding team. Uh we went through blood, sweat, and tears together to get to this point. uh towards 2024

point. uh towards 2024 um effectively decided to it's time to basically start to decentralize the open tensor foundation and decided to sort of start tackling some problems that are

really a little difficult to tackle for the foundation by itself or really just have the bandwidth to do things like building some tooling things like building some uh sort of getting into the nitty-gritty of some of the subnets

incubating accelerating and pushing in that direction and so that was when I decided to partner up with David and let's kind of see if we can tackle this problem together. And this is when uh

problem together. And this is when uh Chris Lab started.

>> Nice.

>> Yeah. I would say our role in the ecosystem has evolved as we've gotten to know it better in a sense like it it evolves, we evolve. So when we started

um the most important thing was we had a billion dollars of emissions going across a number of subnets to have kind of that uh uh allocation done as you

know intelligently as possible and so we became a validator. So that's how we started because validators this was pre uh the the change to detail and so the

validators were allocating and then we were like oh we need to have really full-time you know committed analysts to

uh help explain the ecosystem to people and to better allocate and better make so that we could also be make better decisions and that's when we hired Sam

Mikas and then Seth joined and then you know They spun out unsupervised capital and we backed that because we thought okay bringing people like yourselves like Sammy and Seth into the ecosystem

you know people who were articulated and able to explain it to other people is going to be much higher leverage than anything Alla and I could do. Then we

were like oh this wallet is going to be kind of an important access point for a lot of people. How can you know kind of we make this investing in subnets easier? And so we started building the

easier? And so we started building the wallet. Then we were like, "Oh, there's

wallet. Then we were like, "Oh, there's lots of uh you know, kind of we want to up the quality of the subnets. We think

there are some really good ones, but we would like them to we'd like to there to be dozens of good ones." And uh so >> so it's really evolved as you know kind

of what's the highest leverage thing we can do right now to kind of make the ecosystem better.

>> How many subnets have you in incubated then? We're we haven't well we've helped

then? We're we haven't well we've helped a lot of people along like I think the first subnet that we started working with was Bitmind >> Bit >> that was more of an accelerator role but

I'd say we've worked you know kind of because we have you know some of the um earliest developers from Bitensor on our

team were uniquely situated to help people understand the nuances of uh getting a permissionlessness system working as well.

>> Oh, nice. And so I guess what Ally, you you just had if you've permanently left then >> pretty much. Yes. Uh since about 2024, I think I released a medium article about

it kind of talking about the next steps.

Uh still in connection with Jake, we still talk uh on the regular, but uh I'm not so involved in the current development anymore. I'm more sort of on

development anymore. I'm more sort of on the crisp side of things and sort of, you know, speaking to subnets almost on the daily and kind of getting more involved in that.

>> Nice. Yeah. So I guess you're you're really like the the commercial market arm of >> product side of things.

>> Yeah.

>> So >> was your end vision of Bit Tensor ever different to Jacob's end vision or like h did you both were you both pointing

towards the same place or or or what?

>> That's a very that's a very very good question. Um so to kind of maybe dial it

question. Um so to kind of maybe dial it back a little bit in time. Uh I think Jacob's been working on the idea of just general decentralized AI since like 2016ish give or take. It's been a long

time for him. Uh I kind of around 2019 I arrived at a similar conclusion and we sort of met up um online and he was like you know I've been working on for a while come work with me yada yada started building it together. The

initial idea was just decentralized AI right we weren't kind of thinking that it would become this massive coordination engine to incentivize anything. It was just hey let's see if

anything. It was just hey let's see if we can you know build some models that can learn from each other and grow that way and compound knowledge. And uh

interestingly we sort of you know down the line um you know eventually we created uh the revolutions branch which became the subnets which is sort of what unlocked all of this and allows you to

really start thinking outside the box instead of just thinking of this as an AI incentive layer. it's an anything incentive layer, right? And um we still kind of kept thinking more along the

lines of okay uh how do we create an ecosystem where people can build products easily but at the same time also um achieve the sort of achieve the goal of decentralization in the first

place and um towards sort of 2024 um we decided that it is time to what's the word to say to decentralize the f itself which was a plan since day

one. I think a lot of the uh the early

one. I think a lot of the uh the early listeners uh sorry the early adopters of bit tensor who are listening to podcast probably remember us talking about this for a long time they're centralized to are you centralized to f and I remember I get a lot of DMs on discord are you

going to splitting up it's like no no no it's start centralizing it's not about leaving so um you know because in a lot of ways sort of if you stay in one spot you're sort of a centralizing the team itself and you're also centralizing the

effort so you're not really kind of talking the problem from different directions and that's sort of where the logic came in um so effectively In 2024, we kind of had a very long conversation about what the future should look like.

We ended up sort of at the same conclusion that really decentralization is sort of key here. And um you know, I really wanted to tackle the subnet side of things and kind of the uh the social and economy side of things a bit tensor

and sort of Jacob as we all know is almost like always two three steps ahead of everybody working on the the kernel side of things. So he kind of continued running the F since then and uh sort of

at this point you know it's sort of at a point where I'm able to talk to the product side and he sort of saw on the kernel and um that's where we are.

Amazing. And so so how long until do you do you think Bit Tensor will be like command center decentralized because there's been a a bit of sort of backlash

with the whole net tow influ flow change because that happened like boom and like some of my institution or our institutional investors went wait a minute that that was pretty fast like

what what led up to that that implementation. Um,

implementation. Um, >> I think one of the things to think about just before you kind of get into the details of this implementation is what's

the right balance right now between centralized and decentralized.

Um, like because I think you can get into decentralized very quickly. That's

actually a relatively trivial change in a way, but then the way you make decisions still needs to be coordinated.

And this is where a lot of, you know, chains fall down that kind of move too fast in that direction. And when you need a lot of changes, which is right

now, like after DTA pretty much was exactly as I hoped it would be like, and I think pretty predictable, but then there's a lot of little consequences

and you need to react to those quickly.

So, if we had been kind of going through some period of let's kind of get a voting system and whatever, you know, just to decide on little things, that wouldn't actually be so great. So, I'm

while it's disruptive, I'm actually in favor of moving quickly right now.

>> Yeah.

>> And I don't know how long that lasts, but you know, probably a year, you know, where we can just make these tweaks that are kind of, I think, going to be really important.

what we're trying to do is very difficult like create an incentive mechanism >> that works in a permissionless way. Um

that's a really difficult thing. So we

you know you have to kind of respect that.

>> Yeah.

>> Just to so I was going to say to what David was saying I fully agree. Um I

think speed is extremely important here because everyone's noticed sort of how fast people adapt to different changes.

But one thing that kind of everyone seems to forget as well is you know we we built bit tensor based on speed right the whole point of adopting um you know the polka dot ecosystem was so that we

can create um forkless runtime upgrades very quickly and the reason we didn't become a parach on polka dot is because things are moving way too slowly. So if

you end up with sort of bleeding emissions or you end up with a problem where people taking advantage of the ecosystem and you have to take six months to vote on something that's just that's just not going to work right.

you're going to move way too slowly and everything's going to catch up way too quickly to you and it's just, you know, you're going to end up shooting yourself in the foot.

>> Yeah. Especially with the hing coming up, I think you you do need to have big changes done before then because the harving sort of cements it things a little bit. Mhm.

little bit. Mhm.

>> Um and and like the analogy I always have in my head is that when you look at governance, the best form of governance over the last sort of 250,000 years of

human existence is benevolent dictator.

Um and it's a bit like a parent and a child. When you have a kid, you as a

child. When you have a kid, you as a parent is the benevolent dictator. What

you say happens. But when the kid is like 8 10 years old is they start to sort of become their own person a bit.

And then by the time they're 16, they're like pretty much mini adults. So you

then have to give a lot of like responsibility to the kid and then your role as benevolent dictator becomes the sort of like benevolent observer to to a degree.

>> You're right. You're right. And I think then you have to look at the you know what's motivating the people who are the dictator right now. And I can tell you

no one wants to uh get out of the dictator role more than const.

It's exhausting. It's a really really really hard job where you're taking in like real frustrations from so many people, right? Like because it's their

people, right? Like because it's their lives. They're building their lives on.

lives. They're building their lives on.

And so um it will be a lot better for everyone when this is done. and he's

really asking companies like Crucible to step up.

So, you know, if it could happen right now, it would.

>> So, okay, two quick questions. So, if

Bit Tensor is that child, how old is that child in that analogy? And you said step up. Step up in what way?

step up. Step up in what way?

>> Play like make propositions to the group in terms of what changes you would like to make. do, you know, implement them or

to make. do, you know, implement them or help implement them. So, yeah, that real and I think that's been going kind of from the beginning with this like the

you know, kind of the this, you know, many of the people who kind of were once in the foundation who are now, you know, kind of running companies in the Bit Tensor ecosystem.

>> And as for the child's age, I'd say just learned to walk barely. Not even close.

Wow.

>> Very early. Very, very early. [snorts]

Nice.

>> Mark, you've been rather quiet down there.

>> Always listening. Always listening.

>> So, so guys, so like, so our our number one goal for for me and Simon is is to bring more capital in. And so everything kind of flows on the back of that for

us. Every decision, wherever we're

us. Every decision, wherever we're investing our time and energy, what what's the number one goal for you guys?

And where does that see you investing most of your time and energy?

I mean the number one goal right now is to help people understand the ecosystem that set of investors retail investors

mostly um so that uh they can have more confidence to invest. So, I'd say we're, you know, kind of like more trying to be

your, you know, marketing arm in terms of building materials. Like that's

really what we're trying to do with Sammy and Seth and at Unsupervised Capital. That's what we're trying to do

Capital. That's what we're trying to do with the wallet that we're building. Um,

everything is built towards onboarding investors.

>> And I think in doing that, we sort of discovered there's kind of what's the word I'm looking for? open truth, unsaid secrets about um the Bitens ecosystem.

Really, it's things we all know, you know, things like fragmentation. Um you

know, standards are still emerging in the ecosystem. So, everyone's sort of in

the ecosystem. So, everyone's sort of in a way reinventing the wheel when it comes to wallets, infra tooling, pipelines, etc. Um another thing is something we talk about almost every day

is onboarding complexity. How hard it is to actually um bring developers in to get them started. it's a little too hard for them to get started and it's also too opaque for capital allocators to evaluate the subnets themselves and

that's sort of why we built the allocator in the wallet itself. Mhm.

>> And then finally, it's some subnets are still struggling to I want to say articulate their value capture in the first place. And that's I think a very

first place. And that's I think a very common problem that everyone has. And

that's also because, you know, we're still early and the market is still discovering which primitives are actually durable and which ones aren't.

But, you know, we're getting there and it's sort of part of the things that we're sort of observing and sort of talking to people is that these three problems are still not working, but there are problems that hopefully we'll be able to fix um you know, in the near

future. And do you and David have

future. And do you and David have similar thoughts on alphanomics? Like

what you like and what you don't like?

>> What is that? Alphanomics.

>> Yeah. How a subet how a subnet drives value to their token. So yeah,

>> I think we both start from the exact same spot, which is does this subnet have utility? Like we're not really

have utility? Like we're not really crypto people. That's why I don't even

crypto people. That's why I don't even know that term. um like we're much more AI people and interested in oh if decentralized money is important then

decentralized AI is a lot more important and that's our starting point for being involved at all and then we're like well what are the subnets doing and if we

think they're interesting like if you take ridges for example which is you know harnessing the the tokconomics to

finance open-source development. That's a really interesting

development. That's a really interesting use case and for a coding agent which is a very you know kind of useful AI application like kind of really the most

useful right now in terms of revenues and stuff. So that's a really good idea.

and stuff. So that's a really good idea.

So we're like okay people should know about that and you know that's you know kind of and we're we put our own money in that like we think that's good and uh

so you know and then we should tell people why so that's the starting point that I think Al and I come to it and when you look at ridges um the first

thing you realize relative to a lot of other subnets is miners can make a lot of money right because they're um introducing

knowhow, not just hardware.

And so you can get paid for your knowhow. And so they could be running

knowhow. And so they could be running at, you know, 95% margins um because the costs of their infrastructure is dimminimous relative

to uh their knowhow. And that's

attractive to us because open- source developers, this is one of the things that Ala and I have been really excited about from the beginning is that open-source developers

uh could use uh a financing mechanism which we know bit tensor submets are.

And the idea that everybody who works on it is basically becoming an owner is very attractive to me.

>> Incentivize open source.

>> Yeah. and and and how do you think the alphanomics will sort of evolve because we've gone from a very primitive buyback and burn to all sorts of fancy stuff

like the hippiest new alphanomics um like do you see any further evolving like what are your thoughts on it all >> 100% it's going to evolve for sure I think that's part of part of being early

right you get a lot of subnets that are becoming more and more creative in how they do um their I I always call them toconomics versus called alphaomics now.

Yeah. However they do their their alphaomics, it's going to change over time. That's just how it is, right? I

time. That's just how it is, right? I

mean, people are going to discover new ways of incentivizing and that's sort of that's sort of the beauty of Vitensor, right? You're sort of able to ingest any

right? You're sort of able to ingest any technology and sort of any um really any incentive mechanism you want to create, right? I think over time there's

right? I think over time there's probably I foresee opening it up further to help people do their their alphanomics as well. kind of

allow them to come up with more and more creative ways to do it.

>> Things like allowing um sort of the split takes differences things like allowing um sort of for example you have a subnet where minors are more valuable than validators. solution they earn more

than validators. solution they earn more than the validators do right that sort of thing that's 100% I'm telling you as we speak probably constant's thinking about this so these sort of things are are being built right and I think that

these are going to evolve it further and further and further down the line but it remains the problem remains the same which is how do you implement this in as safe a fashion as possible so that people aren't able to you know sort of

take advantage of the ecosystem but 100% I'm telling you right now bit is not going to look the same in a year for sure >> amazing and so I I won't name names, but we Mark

and I have had some pretty interesting zooms over over the last couple of months and a lot of them or some of them were with very influential very deep

pocketed people not within um crypto or AI or bitensor but traditional markets and one of the questions which I've asked some of these who are who have a

little toe in bit tensor is you know you've got x tens of millions in bit tensor already which is like back of a sofa for loose change for you. What's

stopping you from putting, you know, like a billion or two 10 billion in into Bit Tensor? Um, and the answer's always

Bit Tensor? Um, and the answer's always been the same. It's it's the the We lost Mark now.

The answer has been um Bit Tensor is still too young. It changes far too often. So, I can't make a a proper

often. So, I can't make a a proper business case and and decision to then move proper capital into it. So I like have you from have you experienced the

same sort of conversations your end?

>> Yeah, I mean I think that is actually reasonable. Um you know different

reasonable. Um you know different investors have different risk appetite and if people are participating and

watching that's fantastic. Um,

so actually it doesn't bother me at all.

Like I think what I told some investors, you know, prior to the detail launch was detail has a lot of risk and, you know,

kind of if it works out as expected, then, you know, it it on balance is probably better to get in before than

after. And um you know that is true

after. And um you know that is true about every change that you know we're anticipating. We're you know you guys

anticipating. We're you know you guys spend a lot of time talking to new subnet owners. So you have a lot more

subnet owners. So you have a lot more information than they do about what's coming. So,

I'm really excited about some of the new subnets, uh, you know, kind of or people, you know, just in the kind of pre-launching stage. And so, I have more

pre-launching stage. And so, I have more information, but what I'm hoping is that in six months that information is going to prove out and so then on balance,

it's probably better to invest now than later. That's kind of what I try and

later. That's kind of what I try and tell people. If you think these things

tell people. If you think these things are going to happen, but most people are spending, you know, because they have to. If they have 90% of their assets in

to. If they have 90% of their assets in other assets, then they're spending time 90% of their time thinking about other

things. And uh yeah, so that kind of is

things. And uh yeah, so that kind of is logical to me.

>> And sorry, just one last question which I I forgot to ask right at the beginning. So David, you've got an

beginning. So David, you've got an extensive Google background. Is that

correct?

>> Yeah, I spent two years Yeah.

>> Yeah. I spent 19 years at Google. Uh and

then in last year, uh 10 years, I ran a growth fund for them.

>> Holy Okay. So, you're the perfect partner for Alllet. [laughter]

Well, you know, I think ma I've been interested in decentralization, you know, since the 90s and uh you know,

kind of as a consequence was very early into crypto and uh Bitcoin obviously specifically and the ethos of this

really worked for me and so you know, yes, I was very predisposed to spending time on this at that area.

And me bugging him 247 also helped just to add to that.

>> And so so what is the revenue stream for Crucible then? Do you so is it like a a

Crucible then? Do you so is it like a a percentage of owner key emissions from things you sub like incubate or or what?

>> We're we're we're currently doing you know each project we just try and think about what's the best way to finance that for its long-term success. So

that's the way we start. So um when we did the validator we take a take like any validator. So that's relatively

any validator. So that's relatively straightforward. We continue to do that

straightforward. We continue to do that and we, you know, we hope the validator role is going to become even more important over time and we're constantly in conversations, you know, kind of to

try and come up with schemes with subnet owners where we can do more um because we think we can be a really good partner to them um with unsupervised capital like okay, we want to have all this

research, we want to have our money, some you know kind of portion of our money professionally managed, we want to have uh high quality conversations with a team that's doing research. Like

basically the people that would have worked for me at Capital G were very stimulating conversations for me. I want

to have those types of conversations about Bit Tensor. Um and you know they provide great lead genen for me in terms of like understanding you know finding

the smart people faster. And so that the best way to finance that was to back them in starting their own fund. So we

were seed investors in their fund and we're part owners in their fund and that's that's our revenue model. Uh for

the wallet the hope is that that is just a you know kind of helps support the validator in terms of kind of our business model. But ultimately it's

business model. But ultimately it's always good to kind of you know have competition in terms of the you know user interfaces because different users

need different things. like some people want to see all the detail and house stats and some people don't.

So I think it's good that there's kind of many approaches there. And then you know with the subnets we always think about a subnet as a business. Like

fundamentally like it's not just like something you can do like five or five subnets. Not my view. A subnet is a

subnets. Not my view. A subnet is a really hard business to build. And so we become kind of like a why combinator like stake in subets.

So, you don't fancy getting back in the subnet owner owner seat again, Alla or again? But

again? But >> I How about this?

I'd say stay tuned. Best way we could describe it. We're we're not we're not

describe it. We're we're not we're not um sort of not sure if we can get into that quite yet, David, but it's more that we are getting more active in the subnet scene and um there's going to be some changes fairly soon.

>> That's exciting. Will you come back in a share once you you've gone live?

>> Absolutely.

>> Well, it won't be for us for to share because there'll be an entrepreneur in the seat. And that's really, you know,

the seat. And that's really, you know, that you know, that's really our approach there.

>> We're just trying to start as many businesses on Bit Tensor as we can.

>> Nice. What What's the What's What's the one thing that's going to make the difference? You talked about kind of

difference? You talked about kind of bringing retail investors in. Is is the most important thing the wallet is the most important thing controllable

influenceable or is it just going to happen when it happens and nobody can control it?

>> Maybe give me a little bit more context for what you're thinking.

>> Well, so attracting retail investors into Bit Tensor is is the is the wallet the most important thing? Is is the

wallet and some some user manual or is it the the price of tow i.e something

that's relatively uncontrollable that's going to be the most important thing or is it going to be a breakout subnet or are there other things on your list of

hey you know if we can just crack this yeah I think it's the uh subnet like I I think there's no doubt right now that

you know a a subnet going mainstream meaning attracting a lot of uh interest you know as a use case like oh if If

Ridg's coding agent gets used by lots of people, you know, that's really the best proof that businesses can be built on Bit

Tensor. And um then people will be like,

Tensor. And um then people will be like, oh, what's this Bit Tensor about and is it fairly valued? And I think you come to the conclusion that if you can have

some number of businesses that could be as big as cursor then um and maybe not even that big you know kind of maybe

much bigger um then you know bitensor is a good bet and you know it's a good bet in a context of yes I think bitensor could be more valuable than salana

that's kind of what we're shooting for you know and is Is it possible that an AI incubator could be more valuable than Salana? Of course it is. So I think

Salana? Of course it is. So I think that's kind of the path that we hope people go down. Sophisticated investors

go down is that they see something that's really attractive then they spend the time to get curious about Betensor.

They see that there are other emerging things and they're like okay I want to invest in that subnet and I want to invest in Betensor.

Just to add a little bit to what David was saying as well here.

>> Have you guys been to Token 249 this year, last year, year before?

>> Yeah, Dubai.

>> Dubai this way.

>> Okay. So, I've been to the one in Singapore a couple years ago. Uh, and

then I went to Super right after that.

>> Did you see anything cryptoreated that was touching the real world at all or was it all just >> customers for cryp right? And I think that um >> Tensor is sort of positioned to be the

first project to actually bridge that gap which currently exists in the entire ecosystem.

>> Right. And I think that uh to David's point, if somebody's able to build a product that is able to um reach the masses while using Bit Tensor as a platform, it doesn't need to be you

know, you know, the beall endall, right?

If their entire revenue is just from subnets, that's fine. But if they create a you sort of a dual revenue stream where they have a real product working on Bit Tensor and they have sort of web two revenue so to speak that's I think

that's the that's the holy grail here you know um to sort of what's I'm looking for to retail investors I think that's going to be what's most interesting I think sort of in in a more

ideal world eventually it sort of becomes a little bit like building a MongoDB I'm not going to advertise I'm building a MongoDB I'm just using it as a back end that runs my system right and I can I can imagine something that's doing that right they're just running on

Bit Tensor and they're fine, everything's building great. It's just

decentralizing the back end. Retail

investors may not need to care about that as long as the product is appealing.

>> Yeah.

>> What do you guys think? You spend all your time on this.

>> Go for my I have my opinion. I think you simish.

>> Yeah. It's a breakout subnet is is what's going to give give Bit Tense some real traction wi without question. uh

education, you know, coming into the space uh is is is important.

Understanding like what it is and how it works and what subnets are and what that need answers and there is no you sound like R2-D2 right now

>> a little bit. Yeah.

>> Okay, fair enough. I I'll let him Have I R2D2?

You >> a little bit.

>> Try now. Try again. Try again now.

>> Yeah, we just Oh, I thought you said >> is Deina like downloading something on on another computer or something?

>> I don't think so.

>> Are you running a subnet right now?

Because that pot should be a very badly.

Well, sign you take over if I'm all gobbledygook.

>> Um, I think it's a bit of chicken and egg.

And again, this is just my own perspective. Like

perspective. Like >> I think that the subnet that breaks out into mainstream and gets millions of users is definitely going to bring eyeballs and then when those eyeballs

look at the subnet, they'll be like, "How do I how do I buy it?" And they it will just force people to learn about Bit Tensor because it's so freaking complicated to a newbie. Um, I remember

when de the detail white paper came out in Valentine's Day or just before and I I read that thinking I have an IQ of 10 because I didn't understand a word of it. I then spent the next three months

it. I then spent the next three months using chat GBT to try and understand paragraph by paragraph by paragraph. Um,

and the average punter won't bother. So

there is a chicken and egg thing there because I actually think the real tsunami of um bit tent needs a tsunami of proper

capital. Um and that's only going to

capital. Um and that's only going to come from institutions not retail. I

think the re like any retail that's going to be in Bit Tensor is pretty much already here. Uh yes we could sort of

already here. Uh yes we could sort of steal people from Salana and Bitcoin and Ethereum and all the other things but I think I don't think it will move the

needle. Um, we need basically a bucket

needle. Um, we need basically a bucket load of fresh capital and fresh eyeballs. And I think if big

eyeballs. And I think if big institutional money were to pour in, it would then also help um the other stuff.

Like when we saw all of these bridges being made into like Salana and Bass and stuff, I I had a controversial take at the time. I said, "It's great. We do

the time. I said, "It's great. We do

need it, but no one will walk over those bridges."

bridges." And we've sort of seen that the bridge is already there. They're they're

they're operational, but there's minimal volume crossing between ecosystems. And that's because there's no like buzz from other people. Like there's no crypto

other people. Like there's no crypto buzz where and unfortunately we need to go from $300 to three and a halfk then all of those bridges will be like inwards only.

>> Yeah, you're right about I totally agree with you. um this complexity issue uh

with you. um this complexity issue uh you know and the challen how much time it takes to understand how bit tensor

works is a huge negative. Um

and then you kind of think about it from the other side. Building an incentive mechanism that can work in a permissionless way and allocate capital between the best businesses and

decentralized AI is a really complicated thing to do. And so you're like, h is it more complicated than it needs to be?

I'm not sure. But it is really complicated and and it changes um because it's new or growing and um yeah,

so that's a negative. So we need people who are really smart. [laughter]

>> We need really smart people who can kind of digest this quickly and and I find that is kind of generally the people that kind of it resonates with you.

>> And and one of the things that we found is that I when we're trying to tow peel people, we don't even bother talking about bit tensor anymore because of that like what what emissions and all that

sort of stuff. So, the thing that that I I bring out on every towel pilling session is this little picture here. Um,

and I know I've had some criticism saying that the descriptions are not that accurate or please, I'm open for constructive feedback here. Let me if

anyone sees any gross errors here, let me know and I'll just change it. But the

the beauty of Bit Tensor, I think, is that really it's 9 months old. I know

we've been here for three, four years, but DTO was really the birth of Bit Tensor and and subnets. And in just nine months, we've had we've got over a dozen subnets, which are the world's best,

cheapest, or whatever, which I think is incredible. And it goes back to what

incredible. And it goes back to what David said with or all that. There's not

much in cryptoland that actually does anything.

>> Yeah. And I don't even think you need to believe uh like this level of uh you know kind of top you're describing

something top 1% of top 1% like a world's best. You don't even need to

world's best. You don't even need to believe this. The valuations of these

believe this. The valuations of these subnetss is you know kind of like I don't know how you think about valuation and that's a really good thing for us to

discuss. Um, but I think of it as, you

discuss. Um, but I think of it as, you know, much closer to the current market cap than the fully diluted value because I think um, it's going to be years of

development before. So I think like, oh,

development before. So I think like, oh, and what's it going to be worth in the next year or something like that? So

many of these subnets are worth like, you know, $10 million, you know, $50 million. Um, that's a good bet, you

million. Um, that's a good bet, you know. So you're like, "Oh, it could

know. So you're like, "Oh, it could shoots become kind of one of the larger mediating forces between uh demand and supply

kind of of hardware. Yeah, that's a they have they're pretty going pretty well now and they have a pretty good shot. Um

that's how I think of it. You don't even need to say the world's best and it's like is that worth like a hundred million?" Yeah, of course that's worth

million?" Yeah, of course that's worth an answer. So, you know, it depends how

an answer. So, you know, it depends how you come at it. the you know the valuation of Vitensor you know overall three and a half million or something

like that that's debit uh yeah it's a a few percentage points uh Salana could it become Salana yeah I think it has and is it

kind of better than a 2% chance well maybe like a 2 to 10% chance so it's actually to me all you have to do is believe these things are fairly

valued and you like what they're do You don't even need to be hyperbolic.

>> I think um I'm going to talk this from the other end of things more historical.

Uh I talked David Zer about this so many times at this point. He's so tired of hearing it. So feel free to mute me if

hearing it. So feel free to mute me if you want, David. Um basically when the revolutions branch first came out and the first subnets eaked out, the amount

of garbage that came out of that was just staggering. Like I was just sitting

just staggering. Like I was just sitting there going like what the hell is I'm not even going to name names here but what the hell are these people doing?

You know what I mean? Um the interesting thing though is towards the end towards the beginning of da right towards that when that paper first came out the the if you took a snapshot of bit tensor

when it when revolutions first came out like within the first week versus that last week right kind of like it's kind of like a few seconds after the big bang versus a few million years after the big bang. It's just a completely different

bang. It's just a completely different shift, right? The quality gone up, the

shift, right? The quality gone up, the team quality has gone up, uh the amount of serious people on the ecosystem has gone up, and it's sort of like we ended up with a much more interesting ecosystem at that point, right? And then

DTO hit and we ended up with a similar sort of a similar reset, right? Um had a lot of subnets kind of call me like, hey, we don't know how to do this.

What's going on? Um you know, obviously a lot of miners left. A lot of people kind of decided this is like too complicated for us, yada yada. We ended

up with a similar drop where the quality went down and sort of people didn't really know what was happening and then it has climbed since then. Bit tensor

earlier this year is not comparable to Bit Tensor today, right? Because the

quality has just changed so much. You

know, we still have a par distribution of quality subnets, but it's still extremely um obvious how much it has evolved since then. And I think that your paper has been kind of describing what I've been saying for a long time,

but effectively it's the same thing.

Maybe not necessarily to David's point being the best, but it is still a very workable product and it's still something that you can build on top of and sort of stand on the sort of shores of giants to to say the least.

Yeah. And I think all of the things that you say like kind of describing, you know, I think Bitmine has like a

reasonable chance like of becoming, you know, the world's best deep fake detection. I think deep fake detection

detection. I think deep fake detection is one of those things that would be done, you know, if it's done in a decentralized way would give me a lot more comfort. So I think the idea of it

more comfort. So I think the idea of it works for me and the approach that the team is taking works for me and I think it's like super early days super early

days like you know Bit Tensor four years ago early days and but those are that's you know that's the bet you're making you're not betting that this is Nvidia

today you're betting that it could be that and so you know reasonably you're like you look at the team you do what any venture investor would

Nice. And so Alan earlier you said that

Nice. And so Alan earlier you said that you know you don't see Bit Tensor as an AI platform. It's a a permissionless

AI platform. It's a a permissionless incentive mechanism platform or words to that that effect. And I completely agree. Do you

agree. Do you I think we're on the we are on the cusp now of seeing nonI businesses that already exist in the real world start to

reverse into Bit Tensor because they're like oo this look this looks interesting and I and I see you know sort of instinctive negative knee-jerk reactions

by some people in the community um like what what is your take on that existing businesses reversing into Bit Tensor?

You mean non-AI businesses?

>> Non-AI businesses, right? I mean, it's it's it kind of makes sense intuitively if you really think about it from a business perspective and from an engineering perspective. Um, for a lot

engineering perspective. Um, for a lot of businesses, you know, to give you the simplest simplest solution, right? For a

lot of businesses, sort of decentralizing their compute and having people cry it for them is much cheaper than going to AWS and paying an arm and a leg for like an EC2 instance, right?

That's sort of the the economic sort of cheat code in a way that people will try to use to kind of get into Bit Tensor.

But at the same time, the the businesses that are that are kind of moving into Bit Tensor today that are non AI um to be honest with you, I'm not entirely sure what their incentive mechanisms would look like, but if they

had built it correctly, then they've not only sort of outsourced their back end, but at the same time, they've also created a back end that is a resilient generalist and not a narrow specialist, right? you have a lot of people

right? you have a lot of people contributing to your open-source back end that is, you know, powering your entire product and it's continuously being audited, right? You're

continuously keeping an eye on it.

People are continuously contributing to it and is growing over time. Whereas,

you know, sort of with a static business, you just have a back end that runs and that's pretty much it. There's

not much else you can do about it. And

if you have a decent enough team, it's going to grow with that team. But, you

know, it sort of caps your growth in a lot of ways. So I think sort of from a technological perspective and from a business perspective it moves makes a lot of sense to move into bit tensor but having said that not everything needs to

be decentralized right there's some there's some seriously critical systems that should you know a lot of sort of the centralization of that system is not really a good idea >> so I think it's good to take that with a

sort of a teaspoon of salt with this um sort of take it as a huge positive people are realizing the benefits but also don't take it as a negative somebody if somebody doesn't do because

they think their system can't take it.

>> I mean, I'm fine personally with any business that wants to use Bit Tensor as their uh chain. That I I think is kind

of my starting point. It so happens we live in a time where AI companies are kind of the center of

entrepreneurship like obviously and um you know we're about a few years before kind of AI agents are using

decentralized rails for payments and other things. So, um, it kind of makes

other things. So, um, it kind of makes sense to me that there would be a lot of AI, you know, decentralized AI business opportunities. And that's kind of more

opportunities. And that's kind of more why I'm I'm not like religious about how Vitensor should be used. I just think it's, you know, given the environment,

given the once ina-lifetime, even better than once in a lifetime experience that we're going through right now with AI and AI agents and the

future, you know, kind of that that's where the attention would be.

>> And recently we obviously we we've had a bit of a price tumble because US government shut down. Everything's sort

of sort of been dragged down with it.

Um, but also during that Polychain dumped a shitload of uh of of towel. I

think like to your knowledge, are there any other massive whales like Polychain which have basically been accumulating towel from the early $10 sort of range

and then are set for another profit taking? Like is that the big sort of

taking? Like is that the big sort of whale out of the room now or >> I think it's kind of hard to tell to be honest with you. Um we don't really follow wallets that way and kind of check who's doing what. It, you know,

it's kind of honestly it's a bit of a waste of time to kind of see who's doing what. If they want to profit, they can

what. If they want to profit, they can profit take. Um in a lot of ways sort of

profit take. Um in a lot of ways sort of the best thing we did was raise through an OTC sale and not through sort of giving equity out just because it allowed us to build how we wanted to build and not have sort of to add people

to our cap table or anything like that.

And that's sort of the benefit of having open tensor being a non for profofit.

But I think that um honestly I am the publishing stuff I've of course heard through the great finan Twitter but not aware of really anybody else as far as I know.

>> Mark it how's your internet?

>> Who knows?

>> I've got absolutely no idea. No, I don't know what's going on. Um but I want to just touch briefly on on the wallet.

What what are the big challenges with the wallet? Um, and what are you doing

the wallet? Um, and what are you doing to encourage retail investors to use your wallet and in the wallet? What are

you doing to encourage them to get involved in subnets?

>> You want me to take it on? You want take on dude?

>> Yeah. Why start? Yeah.

>> Okay. So, uh, the wild really was born out of a bit of frustration that there's no real tow native way to really do much with Tao in the first place, especially when it comes down to detail. And uh I

think David and I were getting a little flustered with sort of having follow every single subnet and what it was doing kind of trying to speculate on that as opposed to effectively became day traders doing it. And uh it made a lot of sense to just have something that

would allo auto allocate for us and not have to worry about it. And that's when the idea for the allocator kind of came into being. Um and then once that was

into being. Um and then once that was sort of built out and it started running I think we had a very quiet release back in June July something like that in summer and uh effectively we decided

that you know we need to add a little bit more bit more layers of security to this to this system because you know it's easy to just take and forget but stake lock and forget is even more valuable and that's sort of when the

work for the ledger started and we kind of been building it since. um has been out while it's working fairly well. Uh

we've had a very healthy increase of users since and um the nice thing about it is it literally you can just lock away your town and not worry about it and then the allocator will take care of taking the rewards you get from root and

allocating that to all the rest of the subnets. Sort of you can get um exposure

subnets. Sort of you can get um exposure in that way whereas your principal is untouched but your rewards are being used more efficiently than just staying in root. Um, in terms of institutional

in root. Um, in terms of institutional investors, I've been speaking to a few and um, you know, sort of a lot of them will work with sort of their own validator and they wanted to sort of lease the allocator for themselves and

use it that way as opposed to using our validator and our allocator for that.

Um, and so far we've kind of just been ongoing conversations, but um, we're trying to figure out a way around um, allowing for institutional investors actually jump in. you know, the massive

um the massive exchanges like Binance, Kraken, Coinbase, they're really not taking advantage of the subnets at all.

They're really just staying in root and that's it. I think that's bit of a shame

that's it. I think that's bit of a shame honestly. So, we think the wallet's

honestly. So, we think the wallet's actually positioned best to take those guys on.

>> Yeah. And then long term, I would think, you know, if you just kind of I think our view is it would be good if it was a Robin Hood like product in terms of all the information you can get, just what

people are used to. Um, and I think just, you know, directionally that's where we're heading.

>> Nice.

>> So, I know a couple of institutional entities that have TA sat on route and obviously I'm constantly proddding them buy alpha, buy alpha and the thing that

they always say is we don't have a wallet that can store alpha. So they may be with like BitGo or or or with Kraken, you know, the c for their custodial

ones, but for them to go out and buy alpha through the liquidity pools and then bring the alpha back to their wallet. Do you know of any like can can

wallet. Do you know of any like can can crucible have a a custodial institutional wallet?

>> Yeah, we could do that. I think there's other people who are going to build that. Obviously, everybody needs we we

that. Obviously, everybody needs we we need kind multiple competitors um or players, you know, to accommodate different users to be able to do that.

And it's just one of the many examples of where Bit Tensor needs to mature. H

>> how far away are you from that? Because

that's the only thing stopping Oblong from converting a big chunk of their tower into alpha.

>> Yeah, well, we could talk to them and we could probably do it pretty easily to be honest. Like our we're working on our

honest. Like our we're working on our road map. um are you know kind of I

road map. um are you know kind of I think rightly or wrongly we're more focused on building the Robin Hood type product than you know kind of more

institutional type product. Um

and so that's kind of our roadmap but we're you know we we you know if we had something that was strategic to do we would do it.

>> Sweet. We need to talk then >> we have really good we have really good team. We have a really good team. So,

team. We have a really good team. So,

uh, you know, I feel very confident, uh, that, you know, we could do that.

>> Has anything surprised you building the wallet? Any any any kind of user

wallet? Any any any kind of user activity that that you've thought, gosh, I did not think that would be the case.

>> I mean, I think the main thing is we need to abstract away the complexity.

And I don't think we've done that yet.

Like, I think, you know, we're kind of early on. we've only been, you know, I'd

early on. we've only been, you know, I'd call it more of a beta product right now. Um, and you know, but over the next

now. Um, and you know, but over the next six months, I think we even over the next, you know, six days, I think we'll do a much better job of doing that. And,

um, I think that's kind of what you users would expect. I was there an early investor in Robin Hood. Robin Hood

wasn't Robin Hood what you see today kind of at the beginning either. Um, so

it's kind of more just a trajectory thing.

>> I think for me it was two things. First,

that there was not yet a town native wallet at all whatsoever by the time we built this. Rather surprising. Um, that

built this. Rather surprising. Um, that

kind of did all the things that we're trying to do. Uh, so, you know, proudly saying that we are the only time vlog doing this right now. Uh, the other one's actually more funny. I got an

email literally yesterday from somebody saying they're trying to stake their ship tokens to my wallet thinking that I am, you know, the owner of some wall with a book. [laughter]

And I'm just sitting there going, "All right, well, you're kind of close, but not quite there." Yeah.

[laughter] >> Might need a name change soon.

>> That That's funny. What are you are do you know already? I know you're only kind of in beta there, but are you attracting new users in for the first time or do you think or are they are

they moving over from other wallets?

>> Yeah. I we think the latter.

>> Mhm.

>> Because that's you know the only people that really we haven't done any broadscale marketing or anything like that. So the only people that know about

that. So the only people that know about it are bit tensor you know kind of community because we're you know primarily talking through X and Discord and whatnot.

What do you think the the killer feature is? Do do you think it is the the

is? Do do you think it is the the allocator is >> I mean I think you know I think you know Robin Hoodl like transparency is what

people need you know kind of ease of doing research and so I think all of that is kind of I need to do my taxes I need there's just a lot of stuff that is

just kind of basic and when I was talking about bringing people into Bidensor the allocator is what I told them they should

like, oh, you know, kind of put your put your allocation into the top 20 subnets, you're going to probably do better than having it in root.

That was kind of, you know, and that's kind of what I wanted to do. And um, you know, short of kind of making, you know, decisions like on specific subnets,

that's a reasonable strategy to me. So

it I don't think we're thinking of it more as just build a first class experience of Bit Tensor.

>> Yeah, Max.

>> Yeah, I think that that ledger integration is crucial because one thing I've said for the best part of 10 years with crypto is never ever ever ever touch a browser wallet because it's like

the most unsafe form of a of a wallet.

But now that you've you've got Ledger backing up, it's it's no different than securing your metam mask with a treor or ledger, etc. So, >> we're also moving to other platforms as well. It's not just going to be a

well. It's not just going to be a browser either. So, we're the browser

browser either. So, we're the browser was sort of at the start. We're moving

into other platforms as well. It's not

just going to be browser wallet.

>> Sweet.

Nice.

The Any other questions?

What questions should we have asked that we haven't asked?

What um I mean it's super interesting you know kind of what are the challenges for people coming in and you know kind

of because you're seeing those firsthand and what can we do to address them like when you're talking to all of these investors. So one of the things you just

investors. So one of the things you just talked about was oh would be great if you know we could bring on this institutional investor you know kind of and they can hold the tower they want.

Are there any other examples of things that we could do or you think just need to be done um that will kind of advance this?

>> Have an answer at the moment?

>> Well, you can follow up. You know where to find me.

>> Yeah, but that's basically the type of thing that we're asking ourselves every day is just how can we make the ecosystem stronger? And yeah, you kind

ecosystem stronger? And yeah, you kind of see it's not about crucible bill crucible. We don't, you know, we're

crucible. We don't, you know, we're we're going to talk about the products that we have in market like the wallet, but we are much more in about just making Bit Tensor better like we think

if we make Bit Tensor better, we'll do fine. Um, so that's kind of our

fine. Um, so that's kind of our attitude. And uh, so if you kind of meet

attitude. And uh, so if you kind of meet people that you think are really good, um, and I guess this is good for all the listeners too. If you meet people that

listeners too. If you meet people that are you think are really good and need help onboarding or they're a business that you think should you know kind of

be brought to Bridge and we're, you know, we're a good person to reach out to and we we're we're not trying to own everything like we're trying to help

people keep our business afloat more or less um and you know kind of have a small stake in a lot of businesses.

That's kind of our aspiration.

um built on Bit Tensor. So I think a and just help people because if Bit Tensor does well, we do well. So that's kind of the net of it.

>> The ecosystem wins, we win.

>> Yeah. It's there's nothing uh uh altruistic about it. Uh

it's just a change we want to see brought into the world. We want to see this product brought into the world.

That's it. Simple.

>> Okay. So, uh, one question that that has popped up, if you, um, if you were to look at all of the top 100 wallets in Bit Tensor right now, I would say probably 90% of them are just sat there

on route doing bugger all. What do you think is required to start proddding these, let's say, the top 500 wallets to get their ass out of root and into

alpha?

Well, I think we're going back to uh what Mark said, which is, you know, a subnets need to break out and the, you know, those top 500 wallets are the best

situated people to, you know, kind of figure that out. So, I think there's the uh like why do I invest in subnetss is because I could make more money, not

because I'm trying to help people, right? And I've made a lot of, you know,

right? And I've made a lot of, you know, I've done well, you know, I'd say, uh, you know, I haven't made a lot of bets, but the bets have all worked out. So, I

think it I kind of am relatively confident that it can be done.

I'm extremely confident that it could that you can make money by investing in Subnets and you know, you guys are doing it every day. I I don't invest as frequently as you, but I think if it's

as relatively straightforward, you find a team that you like and you back them and hopefully it becomes something. So,

um, and I think there's good opportunities for that.

>> Um, there's two questions that have popped up on the thing. So, if anyone's watching, if you want a question, just put it in the feed. You don't it pops up on here. So, first one is from Bat Tech

on here. So, first one is from Bat Tech says, "How can we contact the Crucible Labs team about incubation?"

Yeah, I think X is the kind of main channel that we're using to talk with the community right now.

>> Okay, cool. And yes, you give them a follow. And Bat Tech followed up with

follow. And Bat Tech followed up with another question, which is how do we reach out to you guys? I guess Twitter again. Yeah.

again. Yeah.

>> Yeah, it's just easy.

>> Sweet.

Right, that is all of my questions. I'm

I'm questioned out. What about you, Mark?

>> Yeah, that's been a really interesting session. So, I'm so so questioned out.

session. So, I'm so so questioned out.

He's off and out of the stream there.

Where's he gone? Uh

>> yeah, he's he's off the um Yeah, I mean with the way that we think is very very similar to the way that you guys think.

You know, what's good for Bit Tensor is good for us. We spend a a great deal of our time, as I'm sure you do, talking to people where we are adding as much value

as we can. Lots of people that have not got a subnet, have got an idea, want to find out if they can use Bit Tensor, want to invest in something. Um, and we try and plug people in wherever we can

because a rising tide floats all boats and it it will for for every alpha holder, every subnet, every fund, and all of us. So, that we're completely on

board with that. Um, and I'm excited about the wallet. I I think that the ledger compatibility is is absolutely

key and trying to encourage people to dip a toe into alpha uh is a is a is a big win and it is it is difficult because you do feel like you're day

trading. You feel you feel sometimes

trading. You feel you feel sometimes that you're you're going up and you're going down. It's it's all over the shop.

going down. It's it's all over the shop.

uh and and it's it's very difficult for people to pull the emotion out of it and take a step back and go well I've invested in that because I think that in two to three years time that's going to

be absolutely fantastic because if the markets moved against you violently in a week you start to have doubts.

>> Yeah. Yeah. I think one of the challenges for uh you know kind of these early you know kind of early participants is there was some price discovery you know which took place in

the first you know six months of detail where there was just a lot of exuberance around the subnet prices and um yeah so that's worked itself out now obviously

so people are coming in now have kind of are coming in at really yeah kind of attracted valuations Go for it, Mark. Sorry.

>> And and it's helpful that that an increasing number of subnets are moving from uh creating a a a product to to generating revenue and and you see them

flicking their their activity and their energy over from let's try and develop something to let's try and sell this.

>> That's really interesting. So

>> yeah, and that takes time. You know,

most uh entrepreneurial businesses, they get their seed round, preede round uh and then you don't hear about them six to nine months um you know until

they get their next round. And here you kind of need to build in the open which puts a lot of pressure on business owners. Um and you know can lead to some

owners. Um and you know can lead to some short- termism I think which is a negative um and so you kind of need to resist that and keep your head down and

just keep building and that's not always what crypto investors want. They want to kind of have something number go up every day and yeah I don't think that's

you know kind of realistic. And so what you're kind of seeing is subnets go from like, you know, in the top 40 subnets to the top five subnets and then they stay at that price and you're like, what? I

want it to keep going up. Um but you know kind of realistically you're kind of you've seen kind of something go from preede to seat

and what do you think is the like it's very easy to get so bullish with bit tensor like we're we're in it we're in the thick of it and in our minds it's like a it's a foregone

conclusion that tow's going to go to the moon and subnet's going to the moon all that sort of stuff but the question again that we get a lot when we're talking to these big investors is like What what's the downside? Where where's

the risk? What's the what's the the bullet that's heading your way which you're so >> there are so many risks that it's ridiculous like you can't even start to enumerate them all. Uh you get the nervous

>> let me give you kind of you know risk number one permissionless versus permission. I would say that's the most

permission. I would say that's the most interesting bet that Bit Tensor takes.

Let's be permissionless from the beginning. So if you kind of have an

beginning. So if you kind of have an analogy to autonomous, Google took the approach with autonomous driving that it's better to just go for full autonomy

from the beginning and Tesla took the approach of we'll just you know kind of have partial autonomy. Now today uh you know kind of Whimo looks pretty good,

right? You know you see them all over

right? You know you see them all over San Francisco. The carts are you know

San Francisco. The carts are you know extremely safe. Um it seems like

extremely safe. Um it seems like building in a permissionless environment was a really good approach. Um but you know Tesla could end up with a much

lower quality product in a year a lower cost higher quality product in a year.

Um and so you could kind of come to the conclusion that going mass you know so similarly in the bit tensor sorry for dragging you through that metaphor but that's how I think about it. Um, you

have people trying decentralized training of projects like that in a permissioned way, you know, where they have a small number of miners doing exactly what they're you're telling them

to do and you have people trying to build on Bit Tensor in a permissionless way. And we think building in a

way. And we think building in a permissionless environment is the better way to get to the end state. like if you don't figure out how to deal with, you know, kind of permission unpermissioned

actors, you'll never figure it out. Like

people will always dam you. And so we always tell people when they're building a subnet, yeah, the first thing you need to deal with in Bit Tensor is dealing with, you know, kind of rogue attacks,

people not doing what you want them to do. And yeah, focus on that.

do. And yeah, focus on that.

What people on other chains are saying is that we'll deal with that later.

let's just figure out how to get the application to work in a decentral and that's a valid approach as far as I'm concerned. So I take the attitude

I'm concerned. So I take the attitude like yeah on balance bit tensor is kind of what I think is the best approach to be built decentralized applic

>> I don't think people can predict what all the use cases are for decentralized AI and so I like that bit tensor doesn't take a strong opinion there but more

creates you know uh incubator for other people to you know kind of explore their ideas. But I

could see a really smart team on some other using you know you know you know I working on another project that will ultimately come out with the best decentraliz.

So, okay, Alla's pro, you'll probably have a better take on this, um, Alla, because it it tends to is sort of on two flywheels at the moment.

>> Well, sorry, someone's haird dryer is on. Don't know [clears throat] who Mike

on. Don't know [clears throat] who Mike is.

>> I don't know. Um, anyway, I I'll continue. So, there's there's two

continue. So, there's there's two firewalls with Bit Tor as I see things from sort of like a top down approach.

>> Mhm. If TA starts mooning and it goes, let's say it's all of a sudden $3,000 towel, it is it's a really positive flywheel because all the miners that

have fixed OPEX, they have to sell or they then will sell less towel to cover their OPEX, which means there's less sell side, which means the price goes up and there's other knock-on effects like

that. But also, it's the same on the

that. But also, it's the same on the downward. So if if if tow price goes

downward. So if if if tow price goes down and it's like $200 now for that same minor or validator or business to cover their fixed real world opex they

got to sell more towel units to to cover that. So in theory at some point there

that. So in theory at some point there is a a low price where bit tensor just doesn't work or we have a mass exodus of

talent and and and people like workers.

Where do you think that prices are right from how you know? It's a very good question and I think sort of the nuance behind it is subnet specific and that's

actually I think where the beauty of detail comes in right because some subnets require like it come goes on to actually more specific goes into margins right and miners the margins in the

first place and really if the miners have sort of high margins then this problem is sort of minimized as much as possible you won't need even if even if the tow price goes down significantly

the margins are well enough to cover most costs and if the top price goes up, you sell even less. So, I think really the key there um really is on everybody, but if you want to focus on one specific

entity, I think the one that everyone's talking about is the miners. And I think if their margins are as low, sorry, are as high as possible, then you'll mitigate this problem as much as you can. And that's exactly why sort of uh

can. And that's exactly why sort of uh Dave and I have been talking to every subnet we talked to and it's always been please reduce your margins as much as possible because of course for some of the rich subnets, you know, you they can't help it, right? That's just how it

is. But for a lot of these product

is. But for a lot of these product subnets out there, they don't need all that power for a lot of the time. Unless

>> you know, if unless they had coded their way in a sorry code their mech incentive mechanism in a way that um puts less onus on the miners in a way, you're also putting less onus on the price because

you' make it a little bit more resilient to uh price drops like this.

>> So what about um like minor intensive subnets? to compute subnets. They have

subnets? to compute subnets. They have

so those miners have the thinnest of margins. So like

margins. So like let me though like the yeah the the subnets will be uh you know there'll be less demand and then there'll be less

supply you know kind of the same way that Bitcoin miners come in and out of the market.

>> Yeah.

>> Okay.

>> Yeah. I think it's it's a combination of the two. I think what we you know kind

the two. I think what we you know kind of what you're highlighting is yes if this if everybody thinks bit tensor is going to be a good mechanism then it

becomes a better mechanism and if everybody thinks the product is then then >> right yeah okay yes I think all of that is very intuitive

>> it doesn't matter what dress you put on a turd it's still a turd >> well I mean if we don't if If we don't achieve any great sub if we don't have any great subnets on bit tensor bitensor

will not succeed and if there are good subnets on bitensor and you know that you know we've created the market market mechanism but we need people so what you

guys think a lot about investors coming in we don't really spend you know we are building tools for investors because we think that's kind of a necessary and we do believe you know kind of a a

transparent market is kind of essential to kind of a well functioning market, but we spend a lot more time talking to developers because that's the community

that we need to really bring on, whether it's kind of going to open source AI conferences, you know, broadly or other people in the crypto AI ecosystem. And

that's the metric that I use to get excited to be honest. Like, do I feel like the quality of the teams coming in is 10 times better than it was two years ago? Well, of course I do.

ago? Well, of course I do.

And I think the other thing here is um to David's point, I think developers are much more important right now than anything else. I think worrying about

anything else. I think worrying about price this stage, this early stage of Bit Tensor is kind of like buying your socks before your wedding suit. Does

that make sense yet? You know,

>> give it some time. And I think it's uh it's sort of the innovation really lies there specifically. not and that's not

there specifically. not and that's not so appealing to the average crypto investor who's kind of thinking more daily in terms of how they're um but I

think it's more realistic you know it's it's been a development in crypto has not been that speedy like if you just kind of step

back it's like okay how many Ethereum applications do you use every day you know and that's been around for like a decade um so it takes time uh to build

these applications. But you know, I

these applications. But you know, I think like I think of Bitcoin as just getting going. I think stable coins is

getting going. I think stable coins is getting going. Like I think these things

getting going. Like I think these things are 1% of what they're going to be, you know, like stable coin usage for sure.

Um and that everybody's going to have crypto rails built in. Like every

company like Google is going to use crypto rails to move their own money around. You said the infrastructure gets

around. You said the infrastructure gets built out and then you have the potential to do a lot more and that's a lot like you know oh we saw the

you know kind of browser came out in 1994 but people had narrowband and then you know kind of in you know let's say 1997 the peak of the kind of or the

beginning of the kind of bull market for the internet there was like 75 million people on the internet like it looked like nothing uh because nobody had

access. Um so I think that's kind of

access. Um so I think that's kind of where we are overall in crypto.

>> Yeah, >> it just takes time for everybody to get, you know, kind of and I think agents, AI agents are really like I don't want to think about addresses and it's so

anxietyprovoking everything that you deal with on crypto, right? It's like a kind of a hard

right? It's like a kind of a hard product, but it's going to be so easy for agents to use this infrastructure.

So, I also think it's just like, yeah, it's kind of >> I I was thinking the other day like like I don't watch trash TV, but you like you

you see things like the Kardashians and all these like fly on the wall type documentaries where they're following someone who's like a Zedlist celebrity and whatever. And I I was thinking

and whatever. And I I was thinking that's actually a very powerful tool if you were to get some form of influencer who is about to about to or wanting to set up let's say a tech business and they go through all the thought

processes every single day and then they set up the business but the business they're setting up is a subnet and that this subnet is a bit like a fine where it uses hippius uses you know video

compression it uses shoots and gradients and all these different things to then launch products but then uses some of the the the agents that some of the subnets are creating where but

everything to be on chain like have the most transparent company on the planet where the the the stripe or Quickbooks or the online accounting is all like

like a graphana dashboard of some sort so people can see live revenue coming in where does revenue go out to. Um I think that would be pretty cool. Um

>> yeah, I agree with you. I think uh that's why I think a lot of people who are kind of early to crypto kind of are

kind of still so passionate about it is yes we can have uh countries coordinate activity we can have companies coordinate activity and wouldn't it be

great to have some other mechanism for coordinating human activities just kind of seems obvious >> I mean I will watch the hell out of a keeping up with the Jacobs's TV show where just let Jacob do random things in

Peru. Very down for that.

Peru. Very down for that.

>> Yes. And then the the the Telegram uh escapade.

That' be amazing. Right. I'm I'm aware, David, you you got a tight schedule. Um

I think you got another meeting to head off to. Thank you so much for your time,

off to. Thank you so much for your time, Alla and David. It's it is an honor. Um

if anyone wants to follow them, follow them on Twitter. Follow them on Crucible Labs or at Crucible Labs. Are there any last passing comments that you'd like to make folks?

>> Just thanks for having us.

>> Yep. Had a great time.

>> Amazing. Mark, the quiet man in the corner.

>> Well, I've been let down by uh by the satellites today, but uh yeah, it's been a been a great conversation. Looking

forward to getting you back on when some more of these uh subnet rumors have been unearthed.

>> Looking forward to it. You'll see the the people running it. It's going to be great. Fantastic. Thanks guys. Thanks

great. Fantastic. Thanks guys. Thanks

guys.