Watch

Transcript

Disclaimer: The transcript that follows has been generated using artificial intelligence. We strive to be as accurate as possible, but minor errors and slightly off timestamps may be present.

Jacob Steeves (00:23):

So why Finney? It’s been a year of expansion and there have been some pain points. I think that, you know, focusing a little bit on the pain points, it helps people understand, especially if they’ve mined over the last year. And one of the one of the biggest pain points, one of the political battles with inside Bittensor has been around the UID count. How many miners can we sustain at any one time?

(00:44):

And there’s been a number of limitations around that because we run Yuma consensus and Yuma consensus is limited to 4096 UIDs based on the size of the chain, what we could sustain in the blockchain at any moment in time. So Finney was designed to allow us to scale beyond that UID set horizontally and horizontally, sorry, horizontally and vertically, as you’ll see in a bit.

(01:13):

But the major push for Finney was not to release two of the major updates we’re going to talk about here, subnets and delegation. These are the two add ons. The main thing for Finney was increasing the tensors degrees of freedom with inside of the Web3 ecosystem so that we could integrate into into the Polkadot world so that TAO could be moved across borders. Finney is about BitTensor’s freedom in Web3. Freedom from centralized control so that we’re no longer fully controlling the validator set and we’re able to push the protocol outside of the hands of the foundation and into the Web beyond any individual’s control. Hitting into Polkadot, getting in that relay chain position, that parachain is number one step for that. We don’t, moving beyond a single validator set that we control as a foundation means that there’s no third parties. It means that the token can be moved across borders between chains without anybody getting in the way. And these are really important because if we can’t allow TAO to be liquid, if people can’t choose how they want to use the token, it no longer has market value with inside the Web3 ecosystem. And it’s that market that we’re using to drive the creation of machine intelligence.

(02:37):

So that’s Finney, but where we are today is not quite the Finney that is connected into a decentralized ecosystem of Polkadot. Today, we’re just launching Rococo, Finney Rococo, which is a bit of a test network. And we did it early so that people would have time to prepare so that people have time to prepare for these updates to the chain. They know how to use them and they won’t be sort of caught on surprise.

(03:08):

They won’t be surprised when we launch the chain in a month or so. So what are the major updates? And I think people are really excited about them. So the number one one is subnetworks. We’ve been talking about this for a really long time. Subnetworks are basically just networks. Subnetworks are basically just running multiple Nakamoto consensus across the same chain so we can scale horizontally.

(03:28):

So that we have multiple networks and we’ve expanded the we’ve really increased the speed at which we can run Yuma consensus, allowing us to possibly scale up to 64,000 size networks. So we will be able to have 4,000 scale networks, multiples of them, plus even 16K networks across the chain, running concurrently.

(03:51):

This will mean that the inflation in Bittensor is split between these networks, and this will affect the incentive that each miner gets. So there’ll be a hard task of selecting where that inflation will go and into which networks. A little bit on that later when we get to delegation. With subnetworks, we also introduced something called subnetwork connect, and I’m going to talk a little bit about this, maybe even demo it partially at the end of the call.

(04:20):

Where we have networks that are connected to each other. So Nakamoto 4K was the original network that many of you are familiar with. We’re going to open a Nakamoto registration network with 16K UIDs that allows people to prove that they’re worthy of entering the 4K Nakamoto by performing well on the registration network. It’s a hierarchical aspect, similar to a promotion or relegation network, or promotion or relegation league.

(04:51):

Yeah, go for it. Yeah, cool. So we can have a Nakamoto registration network of 16K UIDs. This means that we’re going to really reduce the difficulty and proof of work, which was, over the last year, was all the proof of work. It was who could push out the other users with that difficulty, and less of a focus on was their miner actually performing well from a machine learning perspective. Obviously, those two things were interplayed, but by filtering UIDs first with the Nakamoto registration network, we’ll be able to reduce the demand on POW, and hopefully expand the machine learning side of the network. Subnetworks also allow us to build hierarchies where we’re working with different modalities. This is really exciting innovation for what we’re doing at BitTensor because we’re going to expand into different modality subnetworks. This is something that happened in Q1 or Q2, where we’ll open up an image network for BitTensor. But with Subnetwork Connect, what we can also do is connect those modalities so we can have a text image network that will be able to connect those modalities so we can have a text image network that requires understanding of both image and text. So we can push towards multimodality through the incentive mechanism directly how people are registering to the network.

(06:11):

So that’s subnetworks. It’s really exciting. Subnetworks are all about scale. It’s about how can we scale BitTensor to more users, to more domains, and it’s a multimodality AI. How can we really showcase that we’re not just solving only a single problem just around text but working on problems that any client wanted to come and use BitTensor for? The way in which we’ll select those modalities, we don’t really know.

(06:36):

Right now, we’re just going to focus on the core ones, image, text, and audio, and start there. But potentially, clients could come and ask for a time series network. I know there have been some people in the community that have asked specifically for that. It’s really exciting for us to expand the AI research that we’re doing in different dimensions of space within the industry.

(07:00):

Right. So this is more about the AI side, about scaling. Finney is about core Web3 technology. How are we launching ourselves into the ecosystem? This is about scaling the compute power of the network. How can we build bigger networks and really truly become a neural internet? Delegation is our foray into autonomous organizations, the ability for people to band together and work in BitTensor as groups.

(07:32):

So those that have used BitTensor before are pretty familiar with the cold-key hot-key structure in BitTensor. We’ve expanded this very simply. And that allowing people to stake their funds to hot-keys that are not their own. This is totally trust. So if you stake funds to a hot-key that’s not your own, you won’t lose access to those funds. You can always get them back. You just can’t use them in validation. But the hot-key that you stake them to, owned by a particular other individual, by another cold-key, can use them. So we’re making this distinction or expanding the potential for people to move TAO in BitTensor so that we can build hierarchies or clusters of individuals that then can group their power and use it with inside of BitTensor.

(08:26):

Over the last year, a lot of people have come to the realization that they needed 1024 TAO in order to be a validator. This is changing. We’re no longer going to be picking a raw value for the validator cutoff. It’s going to be based on the number of UIDs. I’m sorry, it’s going to be based on the stake of the validators. So we’ll take probably the top 100 in Nakamoto 4K. So there’ll be 100 validators. And so in order to validate with inside BitTensor or to participate in validation, you’ll have to delegate stake to someone or have a top 100 key.

(09:05):

There’s a number of reasons why we’re doing this with delegation. So delegation is effectively a liquid democracy where individuals can build organizations and move TAO around with inside the ecosystem and let that TAO be used to vote on things, especially the direction of the incentive flow and with inside BitTensor. So delegation is our foray into what I would say is a liquid democracy.

(09:34):

And it’s also a way for us as a foundation to make revenue as well as other groups of people with inside the ecosystem to make revenue because we’ve built in a flat rate of 18% as a take from the delegators. So if there is 100,000 TAO delegated to one particular hotkey from another individual, 18% of the emissions or the dividends from that stake, the 100,000 stake will go to funding the organization.

(10:07):

We hope that people invest in the foundation so that we can continue work and hire engineers, but we’ve made it specifically so that there’s competition for ourselves. Anybody can build, anybody can become a delegate and anybody can nominate stake wherever they want in any piece, in any splitting of that stake they want. So this really hopefully integrates a little bit of competition at the beginnings of competition at the organizational scale for BitTensor. So delegations about organization, it’s an on-ramp for us for people so that they can begin validating without having to run a miner, they can begin making dividends without running a miner.

(10:45):

It’s a funding mechanism for the foundation and it’s also a democratic tool. It’s a way for people to gather behind representatives and have those representatives work for them in order to make BitTensor move the way that they want them to. So those are the two main, and it’s a real shame that we missed the rest of these slides because they were quite informative, but the majority of the information about what we’re doing here with Finney and the release is in this and I can release the slides to people that want to take a look at them.

(11:20):

So Finney is open and to follow on with this presentation, I thought that it would be great to open Finney and give people the documentation to start playing with it so you can use it if you’re on a computer right now, you can connect to Finney. And why don’t I just show people how that’s done. So if you can see my screen, we’re all good, and you’re on my terminal here. Alok, please give me a thumbs up.

(11:51):

Right, so developers out there, if you want to connect to Finney, you have to be on the Finney branch. So this is Finney right here. So you have to pull this branch from master with inside your BitTensor directory and reinstall it. So I recommend reinstalling with Python. And this will get you totally connected into everything that we’re working with on Finney. So registration, these networks are open.

(12:34):

So for instance, if we were to run a metagraph, you’ll see there’s three networks to choose from, 0, 1, and 2. Right now, these are just numbered, but they will be ASCII based in the future. We’ll actually have a monogram for them. So there is network 0, 1, and 2. And each one of these networks has about 4,000 UIDs in them each. So with Finney, just on the test network, we already have three times as many UIDs.

(13:06):

And we want people to register into these networks right now and really, really test the shit out of what we’ve built. We want people to break what we’ve built on Finney. So then they have one UID. Now, this is maybe a little bit like those early days when there was only a couple of UIDs. And the first time we presented an open network on BitTensor, it was at 100 UIDs by the end of the call. So I highly doubt that that will happen today. But there’s a couple other things that I’d want to showcase here with the CLII that are open for people to do. So one of them is nomination.

(13:44):

So you can nominate yourself to be someone who can accept who can accept funds through delegated of stake. And you simply need to call BtCLII nominate and then you need to share your address with people. So when we launch Finney on Kusama or Polkadot, we will show what foundation address people can use and that’ll be available within the code. So if people want to delegate stake to us and so that we can take an 18% set of so to speak tax on the dividends, we’ll be very, very glad for that because we can use it to continue funding development. But if you want to build your own organization, delegating stake will be the beginning of how we move forward with building those organizations. So BtCLII delegate is the other command which you can call and you simply passed. Let’s see, let’s delegate. So as you can see, the key that I’ve nominated is here and can accept a delegation from other peers, other hotkeys and other cold keys. And so it’s important to remember that it’s the cold key itself that is performing the delegation. Right now, we’ve fixed the take to just under 18%.

(15:08):

So it’s 17, 9999%. And I’m sure there’s people out there who are familiar with that. So it’s 17, 9999%. And I’m sure there’s people that have questions about exactly how that’s gonna play out. The network’s open and so people are welcome to get on Finney and continue playing with it. We hope we do play with it and break it for us, honestly. And that’s basically it. That showcases the two extra additions to the code base for now and maybe we’ll break to questions at this point now.

Ala Shaabana (15:46):

Yeah, I think that’s a good idea. Okay, Mike’s asking, will validators be able to adjust the take? I’m not sure what you mean by that, Mike. Can I clarify a little bit?

Jacob Steeves (16:03):

Yeah, I know what he means. I know what he means. No, so we’ve fixed the take on purpose and instead we want competition to occur at the level of awards that each of the delegates is accruing to their delegators through validation.

(16:25):

So if you have a hotkey that you’ve delegated to and that minor is not actually validating correctly inside BitTensor, you’ll see that affected in your returns as somebody delegating stake. So the fee is the same. It’s fixed across the board. It’s what the delegates can use that for, how they can use that stake that momentumally between which pool you want to join. When can we expect Mainnet to launch?

Ala Shaabana (17:01):

I think we can probably expect it more towards the end of Q1. Is the idea that small holders get a cut of the dividend awarded to validators? Yes, I’m exactly right. Let me back. The idea is that your small file holders will get a cut of the dividend that the validator will be getting. Can a validator reject stake from, say again?

Jacob Steeves (17:27):

You’ll get the proportion of that validator’s dividends, which would reflect the amount of stake that you’ve staked into that validator. So as an example, if there is a delegate with 100 stake from me and that’s it, then I should get 100% of the dividends that that delegate is making minus an 18% take. So they’ll take 18% off the top. They’ll take 18 tokens. Let’s say they made 100 tao. They’ll take the first 18 and then the rest of the 82 will go directly to me.

(18:19):

Alex, you want to continue? I thought the original plan was for users to build their own validators to access the network.

Ala Shaabana (18:22):

In limiting validators to 100, are you also limiting the number of businesses to 100? Well, we have a limitation that’s based per network.

Jacob Steeves (18:35):

So we can have some networks that have more or less validators. So it really depends on what the networks are working on. And whether or not we want them to have larger validator sets or smaller validator sets. It’s something we can pick specifically for each network. You mentioned there’s a Kusama-based mainnet.

Ala Shaabana (19:08):

I can probably answer this one. So I’m not sure if anybody’s kind of been keeping up with the Kusama options, but I’m not sure if anybody’s been keeping up with the Kusama options. But there was a bit of a hiccup on our side of the Kusama option. And what ended up happening is we’re now on the current one. That will be ending, I believe, on Thursday. So when that’s available, we’ll be going on to, like we’ll own that slot. And what we’re going to do at that point is we’re going to test on that slot as much as we can. And then we’ll provide an update whether we are going to be launching on Kusama right away or on Polkadot. The reason for that is because the Polkadot slot opens on February 13 anyways, which is coming out within a month. It’s coming out in about four weeks, give or take. So we might just launch it on Polkadot and skip the entire Kusama setup.

(19:60):

Are rewards automatically distributed? Yes, they would be. Will Nakamoto Mainnet be dismissed or not? Nakamoto Mainnet is sticking around. It’s not going anywhere for the time being. Until we launch fully on Polkadot or Kusama, Nakamoto Mainnet will continue running. So Nakamoto Mainnet will change there for the time being.

(20:24):

What does Kalman look like for stepping into Finney? And how does it go from here? Finney testing, test approved, migration Finney. Are UAD cases on that Kalman line also or different? So yes, as mentioned at Torx, the Kalman line basically is that we just launched what is it that’s off-launch for Finney. We’re going to see how it performs on Rococo. This Rococo has actually been hosted by us entirely. So if something goes down or an entire thing breaks down, it’s totally fine. Not done in the world. Then as we get closer to the Polkadot lease period, so our Polkadot lease begins November 13th. We’re actually going to be going through a lot of valuations including on Kusama and on Polkadot to kind of double check to make sure that we are going to be able to safely onboard everybody over including their balances, their registrations, and kind of literally take a snapshot from the last block on Nakamoto onto the first block of Finney. Once we get that process down, we’re going to have a much more concrete update for you on when we’re going to do it. But so far, it looks rather promising that we should be able to hit our goal for Polkadot.

(21:25):

And then as for your question about the UN increases on that timeline, I think Ponce, you’re probably more on that side than I am. So we’ll give you an update on that as we go.

(21:42):

Let’s see. So the UN 100 value threshold go live in two days for NFP1. So it’ll be more like whenever we launch Polkadot basically. So Finney does go live with Polkadot at NFP1, not in two days. In two days, really, the only difference is we’re just going to have a Kusama slot that we’re going to be testing on, but that’s about it. And obviously, we’re not going to be moving anything over without in front of the community. So once we have more, once we know exactly how they’re going to go, either it’s going to be Polkadot, or if it’s going to be Kusama, we’ll let you know. But for now, as mentioned, nothing changes on Nakamoto. Nakamoto is going to stay the same. We are going to be either launching on Kusama or Polkadot, depending on how this Finney soft launch goes. So we’ll be providing regular updates weekly, seeing how everything goes, and then we’ll go from there.

(22:39):

And yeah, so I’ve got a lot of captain today. So I’m probably just being a little bit wired. Ticac is asking how to become a top 100 validator, just take the amount, or what? Sorry, can you say it again? How to become a top 100 validator.

Jacob Steeves (22:60):

It’s just stake, for now. Yeah. Right, so this is the aspect where people, if they want to become a validator, they need to collect TAO from people that are willing to delegate it to them. So that aspect is a little bit like representatives seeking votes. Yeah. Ala, next question?

Ala Shaabana (23:32):

Let’s see, it was asking, do you mean end of Q1, March 31st of February? Basically, we’re aiming for mid February, unless something catastrophic happens. So we’ll move to March only. So ideally, fingers crossed, keep us in your prayers, we’re aiming for mid February. Can you address the issue raised by Elder J about open-source validators? We will address this in the coming week. To be frank with everybody, we have been head down on this launch for a long time, and now we’ll go back to regular business, as well as maintaining.

(24:16):

Stand by for that. Is there slashing if a node goes down? Even by slashing?

Jacob Steeves (24:25):

I think it’s really just the same way it’s been all this time. If your node goes down, and you’re a miner, you’ll be slashed by the validators, of course. If you’re a validator that’s had a bunch of stake pooled to you, and you’re performing badly because you’ve gone down, you can imagine that people will pull their stake. So it’s important for validators to stay up online if they want to track investment.

Ala Shaabana (24:52):

We mentioned that the idea is to incentivize validators to compete for delegations by optimizing to validate better than the competition. Will there be more information on how validators can be optimized?

Jacob Steeves (25:09):

There’s a number of ways that we want to introduce competition between these validators. One of them is that we want validator pools that exist to showcase that they’re using that take, that 18%, to actually benefit the network. The foundation is creating our key, and we’re going to be outputting a lot of work from the funds that we make. We would hope that people that invest in other pools expect the same performance from their validator pool, that they actually perform necessary duties into the network. That’s what I meant by introducing a government’s layer into BitTensor. We’re going to work hard to make sure that that’s visual and transparent to people on how that works. It’s just really, okay, how much money is this validator making for me? The margins between validators is actually not that high.

(26:10):

They won’t be that big of a difference between validators, so it will be minor percentage points, as far as we can see. We’re not looking for people to, say, find a way to create better validators right now. We’re more looking for people to pool stake as a way of building organizations.

Ala Shaabana (26:39):

This is asking, are we safe on Nakamoto until February for our Polkadot launch? Yes, you are really safe on Nakamoto unless otherwise stated. That means that until we make an announcement about it, Nakamoto does not change. It’s still the main net. Everything else, Finney, is a testnet. The one thing to keep in mind for everybody here, though, is that we will be winding down Nobunaga in the coming weeks. I will post up an announcement about it as we do it, but Nobunaga is becoming obsolete in the next couple weeks. Nobunaga is a testnet, so it doesn’t make much of a difference.

(27:12):

To be a validator, you are an active manager and engineer with a focused vision of how you want machine learning to work. Hang on, maybe I’ve had too much coffee. To be a validator, you are an active manager and engineer with a focused vision of how you want machine learning to work. Basically, as a validator, you have to make sure that the direction the Bitcoin store takes is the one that, A, that your delegates agree with, and B, that is good for the network. You have to walk that line. How many registrations will be on Finney at live launch?

Jacob Steeves (27:48):

There will be 4,096 registrations on the Nakamoto clone. Everybody’s stake, their balances across the network, and their UIDs will be fixed and a deep copy from the current Nakamoto. It’ll look exactly the same and feel exactly the same as Nakamoto. All the hyperparameters will be the same. It will be what you expect if nothing were to change, really.

(28:18):

But Finney will also have other networks on that same chain, so that will be exciting. There will be other places for people to register. We may open it up immediately with a 16K registration network so that in order to get into Nakamoto after the launch, you’ll need to perform well on this registration network to do so. We’ll reduce the difficulty for registration on Nakamoto because people have to perform well in this registration network first.

Ala Shaabana (28:50):

If the validator is out of the top 100, would there be deregistration?

Jacob Steeves (28:59):

Once it’s out of the top 100, it falls under the same deregistration issue as others. However, you don’t immediately lose dividends. You still maintain your dividends while you have bonds. Basically, there’s a slow decay of your dividends to zero, at which point you’re still performing well and you’re not going to be pruned from the network. There are slight differences there in the way in which we’re pruning nodes that people can read in the code. I think you’ll see it’s much better than it was before in Nakamoto.

Ala Shaabana (29:43):

Well, you guys created an easier way to stick to validators as in a user interface.

Jacob Steeves (29:49):

Yeah, this would be really easy because the way we built delegation is that you don’t even need a hotkey. You don’t need to register anything. You just need a normal BitTensor wallet. Delegating could be just a single click on a normal wallet. So we could, for instance, and we intend to add the ability to delegate stake as something that’s integrated right into our wallet. It’s going to be on the website. So people will be able to just hold, if they’re holding TAL, they can immediately delegate it and pull those tokens down.

Ala Shaabana (30:28):

Mr. Paco requires more documentation. Can we explain why PAL is still necessary? So Mr. Paco, it’s good to give a bit of a history lesson here. When we first started BitTensor, the idea behind the typical work would be a form of civil resistance. To keep people from deploying too many miners and flooding the network.

(30:50):

What it evolved to as well as civil resistance is basically a form to say that you have the compute to participate in the network. I was coming in with some small Raspberry Pi, you’re coming in with something that’s powerful enough to actually run the models that are required to enable BitTensor to grow. Now, you were asking, somebody was asking, it was you, I’m not sure who it was, but somebody was asking earlier if a comfortable work will still be necessary once we go on think. And I explained that it is, but it’s going to be less intense. So less intense meaning because of the more UIDs and more subnetworks, it’s going to be a little bit less crazy than it is now. It’s going to have like a hundred trillion difficulty anymore. But it’s not necessary because it does still allow us to kind of have a form of civil resistance and a form to identify that this person is able to run the models that are required.

(31:42):

It seems like a good question, but will TAO get easier to buy on any launch or is it staying peer to peer? No comment. We can’t really comment on this at all because this is not a question, period. Let’s see. Okay, again. HeavyStacks, I’m pretty sure I answered your question earlier. You were asking, there’s a plan within the building system to build their own validators, check the network, emulating.x100, are you also emulating.x100? Konstantin, if you answered this already, just because there’s so many questions, you can feel free to go through the recording later.

(32:22):

When can we expect increasing of UIs on any? More or less, of course, not exactly. I think Konstantin mentioned this earlier. Was it upon launch or a little bit after launch? If I’m not mistaken.

Jacob Steeves (32:35):

Yeah, a bit after launch.

Ala Shaabana (32:37):

The breakdown of what we will use the 18% for exactly.

Jacob Steeves (32:44):

Yeah, so.

Ala Shaabana (32:45):

Go for that one.

Jacob Steeves (32:48):

What the foundation. Yeah, certainly what we intend to use the 18% for, I think that’s a whole other conversation, but we can touch into it. So, the foundation, like the developers that are behind the scenes building the tensor right now, or committing most of the work into the repo are funded by early day miners, primarily Allah and myself who sold TAO in order to fund the nonprofit in Canada, to hire engineers to continue developing a Bittensor so in order to continue that organization, moving forward, we want to, we wanted to build a mechanism for us to fund it, and without having to actually give up TAO directly so to solve our own problem we built delegation really so that there was a way for us to delegate stake into a shared pool and then use some of those funds over time to fund the development to fund engineers researching and building on top of protocol.

(33:51):

We didn’t want to just make that for us. Right. That would just be realized in the foundation so we built it so that it was competitive so that other people can do the same thing. So if there’s other people in the community who want to build a pool and take 18% from that pool so they can find Bittensor and projects on top of the tensor. We want to see that we think that that’s the best route forward for for the whole system. And, and that’s what we’re hoping that people do. And that’s where the 18% will go so 18% will hopefully go to, you know, funding these organizations and their ability to help us continue with our mission here.

Ala Shaabana (34:34):

Will there be an immunity period for those your ideas that would have been copied so we don’t get registered before incentive has been built.

Jacob Steeves (34:42):

Yeah, so what we’re going to do is we’re going to do a big difficulty and difficulty will be like 190 trillion, and there won’t be won’t be possible for people to to register. Given that the network limit is right around 41 trillion so the trickle in of new registrations will happen very slowly until the difficulty adjustment pulls it down and then there’ll be no registration so you’ll have time. And depending on what number we pick, you’ll have more or less time. We haven’t really figured out exactly what that’s going to be. If you’re going to have a week or two weeks or three weeks. We’re also going to open up to networks, likely the same time so Nakamoto will not be the only game in town.

Ala Shaabana (35:33):

Are we talking to the chief 1000 wallet to implement their solution after the polka dots. Yes, chatter has started but we haven’t actually learned or anything yet so I’d rather not kind of take into more detail there. When will sequence length be increased, and will flash be integrated to reduce memory usage before the next increase.

Jacob Steeves (35:55):

What flash flash attention.

Ala Shaabana (35:57):

Yeah.

Jacob Steeves (36:01):

It’s not on our priority right now, too. It’s a maybe Carol says maybe. Because it’s not high priority for us to get flash attention in board, if you know and I said numerous times before, if you think that’s a competitive advantage for your miners. I recommend implementing it faster than us because, you know, it’s in our interest to make faster miners and the benefit network but in a way it’s more in your interest to do that faster than us. So, we welcome open source contributions to the miners.

Ala Shaabana (36:37):

Will there be measures to stop large holders putting across multiple validators by multiple or 100 validator spots by single entity.

Jacob Steeves (36:48):

So, in the same way in this. Yeah, so I mean, I mean, that would really come down to how much time you have if you have all the time, then you that you could take all hundred slots. Yeah, exactly. So that that is an issue.

Ala Shaabana (37:01):

And, but in doing so it’s a little difficult right because let’s say that you do have there’s very few people that have enough to actually have 100 slots and stay competitive in all the 100 slots. Right. It’d be rather hard to keep that up. And after a while it’s gonna get harder and harder and harder as the network grows and as more talent.

Jacob Steeves (37:19):

And then also hundred is also just we threw that number out. It’s, you know, on Nakamoto right now we have over 200 right so we don’t intend to, to make it a small number like 100 intention.

Ala Shaabana (37:34):

Just to start, really. What stops everyone from just delegating to foundation or only one validator, 18% across the board.

Jacob Steeves (37:46):

People are welcome to do that. I mean, I think the point is that, that when you delegate your stake you want to make sure that that those funds are being used to improve the network, and people should do the due diligence about who they’re staking to,

Ala Shaabana (38:04):

you know, always, you can always take somebody with a lot of a lot of towel and that’s fine. But at the end of the day, if that has not been used in the network, then really it’s it’s kind of a self depicting properties. So this is also kind of an incentive mechanism for something to join big cancer as long as they’re developing on it, and kind of be funded to do that too so it’s kind of incentivizing. What are your thoughts, as a restriction of service where people find you to compete creating honking. And it’s a rather interesting business, but we haven’t really done it yet.

Jacob Steeves (38:45):

What’s the question The question is whether or not we look

Ala Shaabana (38:47):

into the business, what are your thoughts on restrictions. Yeah, we’re able to sell a huge compute for getting hockey’s rich equipment.

Jacob Steeves (38:55):

You, I mean, I don’t know if it’s been done, but but you can do it, and it’s it’s something that the protocol allows because you can make the registration call from a key that that is not the one that’s being registered. So, you don’t even need to give them the hockey’s in order to do that registration and then that’s that’s actually built right into the base layer. If you want to use it, extra service mean we were watching that pretty close on Finney will have the same transfer fee of 0.125.

Ala Shaabana (39:34):

I know it’s down.

Jacob Steeves (39:40):

It’s dropped by 100 fold. Yeah, it’s it’s 100 fold smaller.

Ala Shaabana (39:49):

Okay, I’m good. Did you know when ledger compatibility will be. We haven’t looked into that quite yet, but probably will be sometime in 2324 I’d say, maybe even sooner.

Jacob Steeves (40:07):

Like, I might be on the call but he was saying that we may even have the integration now if you do a little bit of wiggling around with the polka dot JS extension which can hold our token ledger works directly with that I believe. Yeah, but, but, but I mean we were not working on ledger ledger integration directly. Yeah.

Ala Shaabana (40:35):

It’s a great idea. It’s, it’s. Yeah. 18% will go to the delegator you take with right so for someone if somebody takes Mike, his node, then 18% goes to Mike, or with the foundation.

Jacob Steeves (40:52):

No no no inverse inverse inverse inverse 18% goes to the valid to the validator. So the foundation.

Ala Shaabana (40:59):

So that’s that’s what that was a question right like somebody sticks with Mike on his nose. Okay. So yeah, so unless you sticking with the foundation. It won’t go to the foundation.

Jacob Steeves (41:13):

Yeah, exactly. So you have to, you’d have to delegate stake to the foundation minor. Yeah. In order for the foundation to take the 18% stake, take the 18% is for the delegator. Sorry, it’s for the validator. So the validator.

Ala Shaabana (41:35):

Yeah. So just to confirm what the 800 100 is an example of the plan limit. It’s, it’s an initial limit I think investors. More than, it’s not something that is going to be hard limits, and it’s not something that’s just going to be respected, just something that will start there and start. How many different victims of foundations are there’s either choose from at launch for initial competition. We don’t have a number on that yet. We will share that with you as soon as we do, as soon as we start moving forward on. We’ll have all those details for you. Did you use it already.

(42:18):

I think we’re caught up on questions so far, folks, for anybody. Well, I guess everybody’s here but if there’s anyone asking you if they weren’t part of this chat, feel free to let them know that we’re going to be uploading the video as well as the slides.

Jacob Steeves (42:42):

In the next 24 hours.

Ala Shaabana (42:43):

Awesome. Are you asking again when the next sequence like increase. No, that are before.

Jacob Steeves (42:50):

Nothing, nothing planned.

Ala Shaabana (42:53):

Yeah, nothing financing.

Jacob Steeves (42:59):

That’s great for for basically the network to reach an equilibrium before we start hitting with increases like that. We might also create sub networks that have we can and we will actually create some networks that have different sequence like etc so they’ll you know some places where you don’t need sequence length and others you do and and people can choose to work between those, those consistent mechanisms.

Ala Shaabana (43:32):

Yes. Saver there will deFinneytely be half fixed dates number two monitors.

Jacob Steeves (43:59):

Is anybody registered.

Ala Shaabana (44:02):

Oh, I haven’t seen anyone yet.

Jacob Steeves (44:06):

Let’s, let’s see.

Ala Shaabana (44:08):

I already already says he can’t use metagraph or any.

Jacob Steeves (44:28):

Okay, well we’ll get into technical and know some people are registering, I’m actually seeing, I’m seeing six on network zero so that’s great. So, things are off, and people are registering that’s awesome. Okay, so there’s still more questions come in here. Could we make a network Finney channel Yes, absolutely we can. Let’s do that right after this call. My understanding is that Finney will be on polka dot or kasama not both. Is there any scenario where there will be two economic potential networks running simultaneously one on kasama.

(44:60):

And when within the more stable copies on polka dot. No plans to we don’t want to have to incentivize networks. I don’t think that would just complex things and complexify things so we’re just going to keep it pretty simple.

Ala Shaabana (45:14):

If we do feature on sama and polka dot, the summer will be a season.

Jacob Steeves (45:23):

Yeah, yeah. Yeah, possibly we do do that.

Ala Shaabana (45:29):

Oh,

Jacob Steeves (45:41):

time again.

Ala Shaabana (45:42):

Time. Fantastic.

Jacob Steeves (45:44):

I think the questions flowing in or is that it I think maybe we should end the call.

Ala Shaabana (45:48):

I think we’re good. We’re going to end it. I’m Jacqueline I will work on the video, get it done and upload it to you guys like tomorrow ideally.

Jacob Steeves (45:59):

Fantastic. Okay. Thank you everyone. I’m really glad to have everyone on the call really, really huge call today. Very exciting time and biggest possible. Yeah, and we’ll see you around.

Video Description

This is an introduction to our newest network upgrade: the Finney Update.

Link to the Finney section of our docs: https://docs.bittensor.com/Finney.html

Subscribe to the The Bittensor Hub for everything Bittensor!

Bittensor is an open-source protocol that powers a scalable, globally-distributed, decentralized neural network. The system is designed to incentivize the production of artificial intelligence by training models within a distributed infrastructure and rewarding insight gained through data with a custom digital currency.

Discord: https://discord.gg/Qv3fxVaXyE Website: https://docs.bittensor.com/ Whitepaper: https://drive.google.com/file/d/1VnsobL6lIAAqcA1_Tbm8AYIQscfJV4KU/view Network Map: https://bittensor-explorer-staging.netlify.app/

Socials: Bittensor: https://twitter.com/bittensor_ Jacob Steeves (Const): https://twitter.com/unconst1 Ala Shaabana: https://twitter.com/shibshib89 On The Chain: https://twitter.com/on_the_chain

DONATIONS: 5DHYPY9P8VBSpp6Q8m8hHrq4ME5UTiXfXVFbH32MNV4JcSxH

HASHTAGS: #BITTENSORTAO #BITTENSORMINING #BITTENSORCRYPTO #BITTENSORNETWORK #AI #artificialintelligence

Sponsors

Podscript is a personal project to make podcast transcripts available to everyone for free. Please support this project by following us on Twitter.