Transcript
the tensor provides a similar ownership layer
for these AI products.
Now, each subnet also has its own coin
and to get that coin, so there's 128 subnet token coins
kind of inside and below the main tile coin,
to get these subnet coins, you must stake tower.
So it's very, very different from any other chain.
And so subnet tokens are in a certain sense
re-hypothicated tower.
There is a direct economic relationship
between the subnet tokens and the main tile coin.
Now, this is very unlike the gas token joins,
sorry, gas token chains, Ethereum, Solon or base,
where there's absolutely no economic relationship
between the gas token and the protocol tokens
of the protocols that are built on top of those chains, right?
So there's no economic relationship
between Ethereum and Uniswap, right?
There's just different things that perform
different functions.
But if it's answered, there's a very direct relationship
between the tile token and the subnet tokens.
They are in effect the same thing,
different forms of the same thing.
Now, there are other things that you can use the tile token
for to get a subnet, if you want to be a subnet owner,
you have to buy it using tile.
If you want a mine of subnet, you have to register
to mine using tile.
And many of the subnets will sell you
their services for tile.
So you have this sort of subnet stacking phenomenon
where subnets are buying services from other subnets
using tile.
And that happens quite frequently in our universe.
Now, subnets have been around for about a year
and they're only just now starting to get product
market traction, which is about right, right?
This is sort of when early startups get traction.
So you're still early, it looks, you know,
just my opinion, it's not financial advice,
but I think you're still early.
But the pot is starting to boil.
We're starting to see like the little bubbling.
It's not like raging, boiling and steaming yet.
But the little bubbles are starting to churn
in a way that we've not seen until this present moment.
Again, we're about a year and change
into subnets and subnet tokens.
Also, in about one and a half weeks is BitTensors exploit
conference.
And we expect that many subnets will make announcements
about real AI product and revenue traction.
You know, conferences are where these announcements tend
to be made.
We have great expectations, as they say.
Now, subnet owners will plow their revenue into buybacks.
Now, why is this?
And that's typical.
It's not always true, but for the most part, it's true.
Why is this?
Well, because for them, it costs them about one tenth
of what it would cost them.
Otherwise, like they're alternative, you know,
they can, the subnet is their product department,
if you think about it that way, right?
So if you are a company that wants to make an AI product,
your choices are raise money, hire a bunch of people,
or create a subnet.
And if you create a subnet, that, again, there's 256 miners
producing your products, you got to keep those people happy.
How do you keep them happy?
Will you buy back Pratokin with the revenue that you earn
from selling the product that's made by your subnet?
And typically, that costs about one tenth
what it would have cost you to hire a bunch of people.
And plus, your product is better.
It's made faster.
And why is it better?
Because you've got 256 people competing to make it.
And they're fighting like hell.
There is no trouble getting people to mind a subnet.
People from all over the earth will jump right in,
and they will compete like hell.
I know, because I'm mind.
It's almost impossible if you don't know what you're doing.
So the ferocious form of capitalism
produces better product cheaper,
and it's a better option for a lot of these companies.
OK, whatever night covered here?
Yes. And there's also multiple vectors converging
to produce how demand.
First of all, we have the buybacks, like I said.
Recently, a lot of subnet tokens and Tau
has become available on Robinhood, base in Salana.
We also have the root reborn effect.
And there's some rumors about upgrading 256 subnets.
If that happens, theoretically, and again,
it's not financial advice, but you just
sort of look at how things work here.
Because you must buy your subnet with Tau,
if we double the number of subnets from 128 to 256,
that would theoretically create a whole lot of new demand for Tau.
Plus, we have a Tau ETF coming up
from Barry Suburge Digital Currency Group.
They've been following the recent run-up of ZCash.
That was one of the big elements is that that ETF was approved
by the SEC.
And Barry Suburge was behind that.
Barry Suburge is very famously a big ZCash person.
But his other coin that he's very big on equally
is big on his bit-tensor Tau.
He has also filed for an ETF for that.
And that could be approved any day now.
We have no idea when that will happen.
But it's been a process for a while.
So we could wake up any morning now in the SEC.
We'll open now so they've approved the ETF.
And also, just in general, the open and decentralized idea
is now something that's taken seriously, where maybe two
or three months ago, that was just sort of crazy talk.
Now it's like front of conversation.
So it's very, very different thing.
So that sort of ends my soloiquille.
Again, if you want to see the sort of textual version
of this, pill that, TauBuffles.net microsite.
And I've got a description of the 18,
most interesting in terms of this discussion subnet.
It's not necessarily the best ones,
but the best ones for illustrating the wide variety
of things people can do.
So without further ado, I'm going
to bring Shizu up while this happens.
But hey, Zeus, why don't you say hello to the people
and add as you see fit to the explanation that they just gave.
And I'll get Shizu up here as well.
Yeah, pleasure to be here, as always, Mark.
I did want to talk about maybe in the unlock
I had while developing my bit tensor thesis.
If you know anything about why combinator,
you know, they do startup accelerator programs.
You know, they take a piece of equity
and they help kickstart a startup.
Now, when it comes to bit tensor,
they're doing the same thing with subnets.
And I had a conversation with you.
I think it was like a month ago in regards to the manner
in which they do it.
I'm actually a really butt hurt person
when it comes to the 2021 cycle
because there was an exorbitant amount of waste inside
of a lot of these quote unquote governance protocols.
They would give these grants out, I mean,
in the tunes of millions of dollars
and they would provide nothing.
I mean, these guys would take the money
and practically run off to the sunset.
There was almost no reason to actually provide
the work that the grant program gave you
because you got the money up front.
You got the money up front and you didn't actually
have to do anything.
Meanwhile, when it comes to bit tensor,
the only way you can actually get money
as a subnet owner is by having a productive asset
because if your asset isn't productive,
it's gonna go straight to zero.
If there's no reason to hold it, it's gonna be a net negative.
There's gonna be no productivity there
and it'll be nil-nil.
And miners, the people that are providing talent,
the people that we can think of as these like
startup accelerators,
well, if they don't have anything good to provide,
if their work sucks, then well,
they don't get anything for it.
So the best, and I really like when you say
ferocious form of capitalism,
the best workers end up being rewarded
and the way I see it, even if they expand
to 256 different subnets,
we're not gonna have that crazy amount of waste
because only the successful subnets
that are providing value to the people
holding their subnet tokens are going to maintain
that ranking and continue to get bit tensor emissions.
And another thing is, I mean, bit tensor as a whole,
if you think about it, if they're running
the same tokenomics as Bitcoin,
the higher the price goes, the more the emissions
actually mean something.
If we're getting to the same amount of tokens
at $200 per towel, if this thing goes to $400,
you're getting double the emissions for the same work.
So naturally, you're going to attract better talent.
And so that's something that I think is great to add
and something that maybe explains it further
than just it's another crypto product
and there's a ton of money waste being had here.
Yeah, that's actually a really interesting point.
The only way it's compared to the grant programs
is actually really great.
So thank you for that.
Nobody gets nothing unless you produce something
and that something would be better be better
than everybody else because that's the only way
anybody gets anything.
You don't eat.
There's no free lunch.
Yeah.
Now there was sort of a free lunch for a while.
Like you can produce a shitty subnet, right?
So you can buy a subnet and you can do a poor job
of running it and collect fees.
But to sort of check and balance against that is,
basically your market cap determines how much emissions you get.
So if people determine,
basically the people out there decide that your subnet sucks,
they'll sell your token or we'll never buy it.
And your emissions will dump and bit tensor
has sort of an internal cleaning mechanism,
a garbage collection mechanism called deregulation
where the weakest link gets pruned regularly
every once in a while.
So if you suck, your subnet will be deregistered.
You lose the money that you spent getting it.
So but that's what should happen, right?
Because there should be garbage collection.
So everything inside of bit tensor is contests
within contest within contest,
kind of at every level.
And so there are safeguards against that.
And I would say the quality of the subnet's over the last year
has gotten orders of magnitude better
than what we were seeing even six months ago,
but definitely from a year ago, would you shizzy
and and hazers, would you guys agree?
In a room.
Yeah, I mean, I've got to be candid.
I was not a big fan of bit tensor before probably March.
I didn't really talk about it much
because the things provided in detail weren't that great.
I mean, the subnet's where they were kickstarting,
they were trying to figure things out.
There were a lot of exploits, a lot of messes.
But I still remember very vividly
during the all-in podcast,
Chimath Paliapatia chatted alongside Jensen Huang,
which is a CEO of NVIDIA,
and they broke the news that subnet three
did this big technical accomplishment.
They managed to go ahead and train a decentralized model
with different computers all around the world,
a contributing to building it.
I thought it was cool.
Yes, the model was a little bit AI-slop,
but they proved that it could be done.
And I was like, oh, well, finally, something is being done here.
Maybe I should look into this more
and then quickly you learn, oh, wait, well,
shoots was dominating OpenRouter.
Targon, co-authored a paper with Intel,
something about encrypting data.
Whoa, there's a lot of things happening here in BitTensor.
And so that flywheel,
I feel like it's become a lot more aggressive
in recent times, more than anything.
And also, I talked with,
I remember it was a CM kid over its DSV fund.
And he mentioned that early in detail days,
the way the network kick started,
all these subnets was pretty much by root,
which is just staking your towel regularly.
And that caused a hyperinflationary scenario,
it was very hyperinflationary.
A lot of these subnets were very, very hyperinflationary
because we were seeding the liquidity of the network
through root towel.
So a lot of selling was happening there.
However, I think around March,
it's pretty safe to say,
a lot of those problems kind of ended
and there was enough liquidity seeded in the network.
So it's become a lot easier, I think,
to justify getting into subnets,
especially now that version 440 came out,
which is the one you were talking about,
where the top 32 subnets are now getting
the overwhelming majority of rewards.
I think that's a productive thing.
That's a great thing for the network.
And yeah, that's what I have to add.
Yeah, actually, let me comment on something you said,
because it actually dovetails
with your critique of the early Ethereum grant
in the world where everybody just got their money
and they're like, wow, I got the bucket money.
Why should I have to do it?
And then BitTens are sold it
by making everything this perotious contest,
this hunger games like environment
where you don't get anything unless you're the best.
The other thing that BitTens were did,
which I really liked is when they decided
they were gonna do subnets and have subnet tokens,
they could have just sort of allowed whales
to buy the subnet tokens right away, right,
off the early market.
And dump on everyone's head later, right?
They could have pre-mined all the subnet tokens
and just day one, okay, you can buy them.
They're super cheap, here you go, have at it.
But instead they decided that they would
emit those tokens into a liquidity pool very slowly.
And so you basically started off with like,
one token, one town,
and two subnet tokens are something like that.
And then with each block, there would be then two town
and four subnet tokens and then three town and six tokens,
right, it was something like that.
And if you bought right away, you got really screwed
because the value of the tokens were,
the subnet tokens were extremely inflated.
They were unnaturally high
because of the way that this liquidity boot up happened.
But what that, even though it was sort of weird and strange,
I remember looking at it going,
why are they doing it this way?
It was the bizarre thing I've ever seen.
But what it did was it was an anti-wale mechanism.
It completely defeated the normal dynamic we had seen
in crypto, VCs buying 80% of Solana and hanging on to it,
right, and then at some point later,
dumping on everyone's head,
that scenario was completely impossible a bit tensor.
The price discovery happened completely naturally
as a liquidity boot up happened completely naturally.
And so that too was, we've never seen this
in any other ecosystem.
I've never heard of anyone doing it this way, right?
So even though it was super weird,
it was actually super good and right for a spec.
And now of course, there's enough liquidity
that this all makes sense.
Shizzy, we haven't heard from you yet.
What's going through your mind, sir?
What's up, guys?
I don't know, it just feels so good.
For some reason, I'm just so happy today.
And I just don't know.