Akash Network - Economics Special Interest Group (SIG) - Meeting #1

Agenda

  • Establish the purpose of Economics SIG
  • Discuss AKT 2.0 draft Proposal
  • Establish a Working Group for completing AKT 2.0 paper.

Meeting Details

Participants

  • Adam Wozney
  • Andrew Gnatyuk
  • Anil Murty
  • Boz Menzalji
  • Cheng Wang
  • Damir Simpovic
  • David Tse
  • Greg Osuri
  • Hannah Hamilton
  • Hartmuth Gieldanowski
  • Jeet Raut
  • Joao Luna
  • Marco
  • Max
  • Scott Hewitson
  • Tyler Wright
  • Zach Horn

Notes

Greg kicked off the meeting by revisiting AKT 1.0 and explaining the primary utility of AKT being for security of the network, governance, and value exchange. Greg also went into limitations that are addressed in 2.0

Key Limitations for AKT at Present:

  • Price Volatility: Tenants that have long running workloads such as website front ends are unable to accurately estimate how much AKT they will need over time. There is the possibility for over or underpaying. Short running workloads such as batch workloads are not an issue.
  • Akash is in Early Product Market Fit: Some users are very happy with Akash and others are seeking improvements. When users complain about an issue it means they are actually using it and a good sign overall, however this means demand for compute resources from Providers is not as great as it could be.
  • Lack of Incentives to Providers: Running large amounts of compute and high quality compute is expensive. There needs to be more compute “liquidity” on the network which means that potential tenants have the ability to access the compute they desire.
  • Community Pool is Small: The pool as it stands today is great for funding small public goods but there needs to be more resources to tap into to fund development activities and contributions in terms of code.
  • AKT Value Accrual Beyond Inflationary Rewards: There is a hard capped supply of AKT with ~180 million pending rewards to be distributed. After those rewards are distributed, there needs to be incentive for stakers to secure the network.

Potential Solutions:

  • Use any IBC token for settlement on the network: When other currencies are payment options it reduces friction to use the network. AKT will maintain its value by taxing other currencies for use. Another option is to introduce a fee discount where the network can incentivize users to use a certain currency (i.e. AKT with 100% discount won’t incur fees)
  • Make Fee and Take Fee Model: The fees generated by this model are directed to the Incentive Distribution Pool and supplement the existing inflationary rewards. This allows for an assessment of the network security budget, and potential use of funds for other purposes.
  • Options to Subsidize Providers: -Cover operational costs and amortized cost of the hardware for a period of time. -Incentivize based on the workload they host, similar to the Filecoin Plus Program. -Use an “exponential discount model” described in Evolution of the Akash Network Token Economics blog post.

The end of the meeting included a presentation from David Tse, founder of Babylon Chain. He and his team are working on bringing Bitcoin’s security to transactions in the Cosmos Ecosystem. David will be involved in the Working Group and Bablyon Chain could be a potential solution to reducing AKT’s network security budget.

Babylon Chain Presentation Notes:

  • Babylon Chain is delivering Bitcoin security to Cosmos
  • Bitcoin economic security: Confirmation time is proportional to economic security. As time increases, security increases.
  • PoS chain economic security: Near instant finality provides level of security but level does not increase over time.
  • Babylon economic security: A combination of the base level of PoS security and adds Bitcoin security on top.
  • How does this reduce the cost of security?
    • Some transactions may not need instantaneous finality. For those that do not, you can rely on Bitcoin
    • Deployment transactions could benefit from lower initial security while token transfers and staking would still rely on Akash’s existing network security
  • Mesh security - How does this compare with new development efforts within Cosmos?
    • Mesh security is a tradeoff between security and sovereignty
    • Babylon Chain’s tradeoff is latency only
    • Efforts are complementary and it is possible to do both

Notable Questions:

  • Question: There has been a regulatory gray area for AKT in the past in regards to Take Fee, can you give an update?

    • Answer (Greg): This comes down to what determines a security or commodity. If the value accrual is tied to a single entity or small number of individuals, it can be classified as a security. If the value accrual isn’t tied to a single entity but rather a community and value capture is in a public setting then it is a commodity. By radically open sourcing Akash, the number of community contributions outside of Overclock Labs is increasing. This is a good trajectory to classify AKT as a commodity.

  • Question: Can you explain the difference between the Developer Fund and Community Pool? Right now, it seems a little ambiguous.

    • Answer (Greg & Cheng): The goal of the Developer Fund is to solely fund developer efforts. Right now, the Community Pool can be used for anything and we want to focus on splitting developer and non-developer activities.

  • Follow up Question: Can you give an example and how it works?

    • Answer (Greg): The vision is that a requirement must be presented to the SIGs, go to a Working Group to create a PRD, and finally to the Steering Committee. The actual funding may be through a governance proposal or via some other authority the Steering Committee may have to spend. The Steering Committee would be advisory to any proposal and any funding would be much more public and on chain than before.

Action Items:

Greg will create the Economics 2.0 working group

Transcript

This editable transcript was computer generated and might contain errors. People can also change the text after it was created.

Scott Hewitson: All right. Welcome everyone to the first economics Sig. It’s a big nice group, we have today. The agenda we have lined out is to establish the purpose of the economics SIG discuss the Akt 2.0 draft proposal, and then establish working group for completing the Akt to 0.0 paper. So with that, we’ll just pass it over to Greg.

Greg Osuri: Thanks so much. Good to see you all David. Thanks so much for joining. for those of you that don’t know, David, we’re David is the founder for For Babylon Chain and he’s somewhat of an expert when it comes to Bitcoin and Blockchain security. He has done quite a lot of work with Ethereum, especially we can change. I believe he solved two of the three, critical problems with the ethereum proof of stake. and I invited David to join our discussion because, As part of the new revamping of Akash economics, I’d love to explore how we can optimize the security budget. I think so far with proof of state change, especially with app change. We have modeled the security budget.

Greg Osuri: you know, after a typical layer one, And there hasn’t been much consideration as to how an app change can look at security versus a Layer 1, to create a security. And that means we more often than not. My assumption is we overpay for security and so if we can optimize the security budget for Using more of shared resources just like how we use a shared code base in Cosmos.

Greg Osuri: We can therefore allocate more budget for growth and other related activities. So, love to hear David’s been working quite a lot in, you know, in enhancing security, really optimizing the security for for, you know, app change and Babylon. Chain is one of the first I would say class of blockchains. That’s very Cosmos Native that offers security as a service, right? So we don’t see this sort of utility chains. I would categorize them or probably wait for massari to come up with a new name. But I think there’s a new category of blockchains. They’re offer critical services to other app change and I would classify Babylon Chain as a security as a service chain.

Greg Osuri: So I’m super excited to hear David’s thoughts on how we, you know how we can incorporate some of the work. He’s been up to if you haven’t read David’s paper, I highly recommend. Go check out the Babylon children where it’s a little technical. I would write Here with me, I do need to put my glasses already. How do you read that? That really enjoyed reading? David’s work quite a lot. Thank you so much David for taking time to join us today.

David Tse: Great. Great meeting everybody here.

Greg Osuri: and of course, you know, Everyone else here in the call. Thank you so much for coming. It’s in. This is our like Scott Open. This is our first security, sorry, economics, especially in this group, the goal is to, it’s really establish an economics climate for a cost to be number one secure and number two grow, right? So as part of this, you know, you should have we want to start reviewing what we have with the Akash Economics Paper. And all the economics situation in general. And look at some of the shortcomings, and look at how we can improve the shortcomings with a new proposal that I published yesterday.

David Tse: If?

Greg Osuri: Which is intended to be the starting point of discussion. So without further ado, I’d like to kick that off, unless anybody has any thoughts. Or any ideas or any questions? Now would be a great time.

Greg Osuri: Cool. Damir.

Scott Hewitson: Go ahead, simple.

Greg Osuri: Please go ahead.

Damir Simpovic: Evening, thank you. I’m gonna ask. I mean straight away. What’s on my mind? Because I’m I’m running your provider. and the, the Cost of running of provider is quite substantial and the cost of running, the high quality provider is even more and to be able to offer some of the services.

00:05:00

Damir Simpovic: On a really high level. I read in your paper that only in your proposal that you’re planning planning on. Like kind of subsidizing providers. In such a way to increase growth and the quality of the service. So if I misunderstood you please correct me or maybe you could tell us a bit more about us.

Greg Osuri: if you’re happy to cover that as part of this discussion,

Greg Osuri: Absolutely, we’ll cover that. That is a big topic of a discussion, how to optimize growth. In the other questions concerns before we start off.

Greg Osuri: Great so well. The first version of cash. Economics introduced the notion of Akt as a talking. The purpose of Akt is to secure the Akash for stay consensus mechanism. And to govern the protocol all aspects of the Protocol. Is governed by the akd holders by means of delegators and these these delegators are essentially token holders that delegate power to validators who then validate the blockchain. And each token holds one vote, and that’s how the consensus works. Akt also is a primary way to exchange value on a class network.

Greg Osuri: And the reason for that is obviously to increase demand for akat which in turn increases, the the value of Akt which in turn increases the security or a cash blockchain. So akat is integral to security the blockchain. and, all the But when used as a payment mechanism, it’s not very optimal. Number one, being volatile. So, where that really is problematic is for long running workloads. So, for give you some background, a card supports, both long running, as well as short running workforce. Short-running workloads, we would consider are something that, you know, usually ends in an hour. Or less usually are considered to be batch y’all. So background workers. So they’re essentially timed applications.

Greg Osuri: And you don’t have much problem with timed applications because of all utility in an hour, is not that great of an issue. And most wall tile events, we see Are, you know, structural events that happen and we probably see a 10% drop in an hour, but that happens very rarely, and that is not really a big concern. Real concern is long running workloads such as websites or API servers that. Yeah, I’ll find forgets. For example, my blog, which runs for over. Year. You know, that I don’t really update much and I kind of want to just, you know, fire and forget.

Greg Osuri: Situations like that because the way Akash setup where you agree on a set Akt in the beginning of the lease as the leaves, you know, is executed the value, or the Akt might differ, and we all saw that as you saw, we all saw that last year, how much of volatility that we have experience with atk? So this ends up in a situation where where the users either overpaying or underpay and similarly the providers either receiving more or receiving less. It creates for uneasy economics climate, that we intend to solve with this paper.

Greg Osuri: Second akashes in early stages of product market rate. What that means is we do have users and we have usage and all users are not happy. Some of them are very happy, users are not happy and because they obviously use a product to a point where they have, you know, experience some short giving and this shot giving you need to be solved and, you know, usually in product market fit, The indication is when users actually complain about your problem. What a product, because actually care and want to use. So we saw early signs of PMF. and that also means there is the demand is not as great as it should be for providers, to be able to commit a large amount of compute or high quality computer

Greg Osuri: And all the computer has been provided so far to a Cox network. We have roughly around 45 providers globally, distributed. All of them has been organic. Right. So which is a great sign but the organic supply also means provider support or provide whatever they can provide instead of what the network really wants. So there is enormous miss alignment behind what the tenants are asking for in terms of storage requirements, in terms of computer, requirements versus what providers are providing because providers are usually providing and use capacity that they’re not using.

00:10:00

Greg Osuri: So creating early incentives to establish a demand side that can, you know, have good liquidity. When I say liquidity is really the demand, you know, meeting the supply and vice verses supply me, man, right. So improving compute, liquidity and compute resource. Liquidity is very, very critical for a car to really see the next Phase and sometimes bootstrapping is super important. So I want to discuss in mechanism to incentivize providers and Damir like pointed out, You know, it’s very hard for providers like Damir to commit to high quality without the demand. And third is our community fund, which is to be spent on development of network or really spend on the community is too small.

Greg Osuri: It, You know, it does a decent job at funding public goods but not enough to really evolve the network when it comes to growing the actual, you know, contributions in terms of code and involving the network so far. So we want to propose a fund essentially to support public goods funding.

Greg Osuri: And fourth Akt needs a crew value beyond inflation rewards. As we all know, Akt has a fixed supply of 389 million and so far, about 207 million have been released in total and with about 180 million in pending that will be released, right? So once these tokens are released in a, you know, in a period of time, this defined in our old paper, there are no more incentives and we cannot exceed the maximum capacity that we originally propose. So we have to start accruing value for Akt. Why? Because you know, why? Because the network security is proportional to make any value,

Greg Osuri: You know, like we discussed Akt is a proof of stake Akashiqua, stick blockchain and that’s determined by 880 and ekit. He has one board so it’s very, very important for the their for for Stickers, the ones that actually secure the network to have right incentives to secure the network that go beyond tokens that are currently time distribution. So, what we propose is a make and take and make few more Similar to the take income model, we proposed in the first version but this is a lot more clear. So what I mean by cake and make is really comes down to liquidity. So we look at compute as a liquidity.

Greg Osuri: And we look at compute users tenants as taker from the liquidity and provide us as makers for the liquidity, right? So both of them have something of value to gain and when there is something a value to gain, then network can effectively charge a fee for the value their quality.

Greg Osuri: In order to ensure the security and the growth of the network. So this is a very clear value alignment and very clear, you know, proposition that essentially. It says, Hey, you know, you As a compute receiver should pay a fee called Take fee. And you as a computer provider, should make a fee, call call make feet when you are, You know, when you are using the Akash network and these fees will be collected into something called incentive distribution pool that will be distributed in a Collins to a proposal that will be discussing later. Second is And by the way, this make fear rate and take free rate will be set by the governance on chain, right?

Greg Osuri: Second, we also talked about how Akt is not a great value of exchange, why? Because of its volatility, especially when it comes to long running workloads. So what we propose is the network, implement, a multi-tocking settlement mechanism, and optimizing for a stable payment mechanism. So something like a usdc or a uisd, which is interchange. Um, stable token by Agoric. In fact, any IBC enable talk it can be used for settlement. So this also not only gives stability in terms of pricing but also gives other blockchains to be able to use their own tokens to white label akash as a hosting service has offered exclusively to them. So you can imagine a scenario where, as Moses, for example, could incentivize, as much as validators to Iran Akash workloads, their workloads are on Akash and pay you, so you can have this incredible.

00:15:00

Greg Osuri: Fact that I’ll solve the user experience problem, at the same time, build and strengthen the ecosystems around the IBC. You know, broader and broader IBC ecosystem by enabling white labeled a hosting solutions so, What that means is. Well, The big question is like, Well, if you use usdc, how is that going to

Greg Osuri: Um, how is that going to impact on, you know, on occasion demand? Because now it’s going to demean the the need for akash, right? So because that’s in not the best interest of the network especially when it comes to security. We have to ensure a cache network Akt has value or our cruise value. So what we propose is a is a tax on this, right? So the take fee and the make fee will be the tax that you pay to use Akash network especially the tokens. And that take me and make P will be eventually distributed to the akash holders that pay for security. This will go obviously center distribution for that, we’ll discuss on how that could be could be distributed as well, or and we can also introduce something called a feed. Discount for specific talks, right? And this is particularly important because

Greg Osuri: This allows us to optimize to certain tokens to be used a lot more than others. So you can imagine a scenario where the take fee for using aka discount could be hundred percent for using Akt. That means if you use Akt, you don’t pay fee and then encourages people to use Akt more or and you know, and vice versa. So I could imagine a scenario the token holders might make a case saying that because we’re incentivizing, you know, Giving heavy discounts on on fees, they’ll encourage more people to use a car. So this was very prominent in D&B where B&B chain essentially where bananasis could get a discount in the in the training fees while using airbnb versus other tokens, right? So this workout really well for B&B Now, we got to see how that’s going to work out but having an option there to to assign a discount

Greg Osuri: That could be used whenever we use is,…

Hartmuth Gieldanowski: and,

Greg Osuri: I think it’s going to be beneficial to the chain. And Max has a question. What are the legality over the take? And we take me and income has been figured out they were concerns about akbm security. I believe the new way is different as it puts feelings incentive distribute. Yes. So, yes, to answer your question, our best analysis from the best legal minds that we are able to talk to really came down to what determines if I talk in a security or a, you know, or A Or a commodity right now. Again, I’m not a lawyer, I’m not a legal expert. This is my purely, my opinion, and this opinion has been rooted in a lot of research in the past and that, that we were able to, you know, both do ourselves as well as get from experts. So really comes down to

Greg Osuri: The responsible party for value. Accrual Akt talking. Right? So if the value of cruel for 880 token, comes from an individual, you know, individual made the company or individual a group of individual or they’re working together.

Greg Osuri: Then that value crew will maybe consider security. So because, you know, it’s really one single company. That’s, you know, that’s building the value of it, okay? But the value comes from a commune from people that are you know, that do not work for a single company. But are doing things by themselves. Yeah. And be able to create the value and also capture the value in in a public setting. Then that asset could be a commodity. The difference is gold, for example, right? So even though, you know, some but people purchase gold and expectation of profit, you know, purchasing goal, not because I have I’m going to use gold for anything. I’m expecting the goal to sort my value and increase this value over time big, The expectation of profit alone does not make goal is security.

Greg Osuri: Because I and anyone can actually mine gold from anywhere in the world, right? It doesn’t have to be a single company. A gold company per se.

Greg Osuri: So so expectation of profit, I’ll be taking them or you know, or are dividend really doesn’t matter. When it’s not the fact that determines a token as security, or commodity really comes down to who’s responsible, for occurring, value the cautious new radically open model. There is no single company anymore. In fact, the the Overclock Labs the company that was responsible for, for, you know, founding Akash network in and being the, you know, the early contributorial network is no longer the major contributor. In fact, most contributors Akash has about 228, contributors, almost majority of the contributors come from public. And at that number is actually increasing. So the number of overclock contributors. The number of contributors that work for overclock, is minimizing the number of contributors that do not work for office max. So that puts akashes in a much, much better.

00:20:00

Greg Osuri: Place when it comes to the security versus commodities discussion. So akash it could be considered as a very well because there is A And we also have other major platforms like Kraken that it’s considered very, very conservative has has, you know, listed Akash in the US, right? So that also gives quite a lot of confidence for us that Akash, you know, being looked at from legal minds is a commodity Um, Boz. and made a comment and get her discussion on how to incentivize entities to participate or integrate to grow the network based on performance fees would appreciate your feedback Max. I will, I did not take a look at that. Comment, absolutely.

Greg Osuri: I will take a look at it. I should have reviewed before we saw the meeting.

Greg Osuri: Cool.

Greg Osuri: So so we covered take fees and make fees and we covered how certain tokens can implement a discount rate that encourages or discourages talk is to be used. We could see a scenario where ist you know which is a interchange security or stable to again could have a timed sort of like incentive. Like you could say like for six months IST Users will have 50% discount and and so on so forth. So we can also create several, you know, programs and encourages, you know, our ecosystem tokens to be used a lot more cash. Now, the instead of distribution pool, like I said, the pool comprises. I also propose the network rewards, go into the central incentive distribution pool, that will be further distributed out.

Greg Osuri: and so now we have essentially 180 million tokens that are still to be minted, all that minted into the into the IDP Now we propose IDP distributed to five different pools further. One is the most important thing, which is a developer fund, or we want to call this a public utility fun, which is essentially used to our public goods fund, right? Which is essentially used to fund, all the public goods on a cart network, be development, or crash network, create development of clients, be development of the entire

Greg Osuri: You know, blockchain. So this Dev fund, the mechanics are how the developer farm is distributed, will obviously be determined by the working group. But initially we propose a state committee to be able to have quite a lot of say. And of course, every fund, you know, every funding request will be on chain and fully transplant. In fully denial also will be on chain and fully transplant, right? So,

Greg Osuri: Um, now next we come to provide incentives. We have several ways in cineworld providers, right? So we can all the way from straightforward, just covering the cost of hardware to incentivizing based on like the, the kind of workload. They, they were able to serve a file called Plus is a very interesting program where in, you know, a verified deployer or deployment could earn 10x rewards for for a provider and we also have another proposal that I propose three years ago, I believe, yeah, three years ago on using a model that sort of like a game that encourages providers and incentivized providers and and, you know, heavily discounts the the user. I don’t want to go into the details of it, but, you know, happy to discuss this after, as we end the call that I want to take too much time away, right?

Greg Osuri: But there are several approaches on how to subsidize providers, and these approaches will be discussed thoroughly and modeled by the working group, right? That will come after the this economics proposal to actually see what it looks like. Stakers portion of the fee goes, There’s nothing new. We have it today, it’s very similar to what you will be getting as a sticker. But how to sort of distribute the speeds? of, this is the area where I want to start exploring using, you know, using Babylon chain or using

00:25:00

Greg Osuri: Um, using, you know, mesh security or maybe combination of both to as a security budget, right? So this is, it’s gonna be quite a lot of work here. It’s gonna be a lot of exciting work. Love to see David’s contributions community pool, Nothing changes here but community pool here, really incentivize We want incentivize community rather than developers, right? So the development fund comes in developer fund which is which we should be substantial after tomorrow. And community pool should be purely for community activities, right? And then we want to for things that are not distributed. We want to burn it all games, right? I think burning allows for essentially, the way it would work is Simpson Center distribution, pool has a vascular tokens because of multi currency. So if anybody’s paying in usdc the cake fees and make fees will be collected usdc and stored on chain in the IDP as usdc and any remaining us, you see that’s not distributed back to

Greg Osuri: Back to these four pools will be exchanged for Akt off market using a decks. And then burnt and this increases the evaluated. If a NON-STAKERS, This is incentive for holders which we have a good portion of them and we need incentives for holders, because because we need liquidity, right? So there’s too much token for a stake. That means there’s less float that affects liquidity of Akt, right? So, we want to incentivize liquidity providers and we do that by providing tokens so, of course, there’s quite a lot of, you know,

Greg Osuri: Math. That needs to be figured out here in terms of how the account for gas fees and slippage room water calculations. We started working on them at a very, very early. We started defining some of the, some of the, some of the functions and some of the, some of the variables and obviously these calculations will be further, you know, worked on to give you the model by the working group With that. I want to open the floor for questions, comments concerns, and, and thoughts. This is my proposal for a car shooter Akt.

Greg Osuri: Absolutely. So David love. Love to see a presentation.

Greg Osuri: There, you still there.

David Tse: So yeah, maybe I don’t know, maybe after some if you get some feedback from the group first on your presentation, then if there’s some time left over, I’m happy to present some thoughts on how Bitcoin can potentially reduce the Akt, security budget just some thoughts. Or maybe you guys can discuss your your Very substantial proposal here. First.

Greg Osuri: Awesome.

Greg Osuri: Any thoughts concerns? From the group. Max.

David Tse: and,

Max: Hey, Greg really cool proposal. Just another question on like the difference between the developer fund and community pool, you mentioned like Community pool would be only used for community things, which is kind of like ambiguous in my my head a bit and Is the developer fun going to be something like like unchained? Well it’s going to be. Is it going to be like just about it? A multi singer?

Greg Osuri: Right. So we haven’t determined the action mechanics but the goal of the developer fund is to fund develop, right? So there’s a lot of public goods that are on a car like Cloud Mosque. Example you’d say it’s a free open source tool. we have all the client libraries we had the preparers we have you know even fleek we have lots of Public goods that are developing Akash, that’s not getting the sufficient funding that. That is that that should be right and we want more public goods. We built in Akash So the goal of Devon is to have more formal as you know, allocation of funds. and, you know, And the community pool is to use. I know it’s ambiguous. Maybe we can. There’s a good point about what? Exactly that’s community pool do and what versus developer fun. Maybe there’s an opportunity to do to merge these both.

Greg Osuri: Um, I was my original invent thought was to, to do non-developer activities, you know, like that could be funded by the by the community fund. We could also do certain liquidity, provisioning mechanisms using community funds. So we haven’t really thought through like, I suppose like, Our finalized like what kind of activities but that’s the working group. A charter. Okay.

00:30:00

Cheng Wang: Yeah, I think one of the comments was posted and in the discussions last night. I think that goes that exact sentiment right now. I think with these participants more more defined, it kind of chunks out with more specific specificity for the elements that were all rolled up into the community pool and this big and this big kind of opaque pool, right. So the developer fun. I think I agree with Greg in that respect for the community. We do have a community sake as well, so there are elements of this that would ultimately. So, for example, we’re in the community in our world say yesterday we’re talking about, if there are boundaries for example, for certain contributions that we would like for example, taking notes and minutes etc for meetings like this and others that could come out of that particular pool but for further the finding exactly how it should work. And and those mechanisms completely open to to discussion, I think it’s it’s worth while saying that this proposal very proposal here does add a lot more specificity such that

Greg Osuri: Here.

Max: Okay. Yeah, makes sense. I guess things can be like fleshed out during the working group sessions and stuff like that and just one last question before and…

Greg Osuri: It.

Max: I stop talking Can you give me an example of you see it working? For example, developer from Require a requirement that like it. Let’s say I would put out the proposal to get funded for Whatever something that I developed. Would it be because like the community pool is straightforward, you put up a proposal and then you ask a community pool spend, and then it sends it automatically. But, and how do you see it working for a developer fun. If it’s, let’s say a multisig would be kind of like the the past way of doing it,…

Greg Osuri: Yeah.

Max: with the Previous programs. The community pool management.

Greg Osuri: Yeah, so so my envision is the in order to get the Devon the proposal needs to be presented to a SIG. So, essentially, Cloud Moss wants to present a proposal SIG SIG. Clients is a proposal is the special interest group. Once the SIG clients, you know, approves and discusses, it goes to a working group that will work on the proposal to create a PRD. And establish some sort of guidance, in terms of estimation how much this proposal would cost, how much this proposal would be. You know, would what recommendation to the student committee and how much this would cost? The same company will eventually will approve it.

Greg Osuri: And once that’s approved, depending on the amount, it could go on chain as a proposal or it could, you know, a Syrian committee could have some authority to spend Akt, that’s less than a certain threshold, right? So for bigger projects obviously it’s going to go fully on chain very similar to community fun and that the proposal really to what’s real community. Thoughts are right and week. All so work out a way where instead of like having the student can be an approval or a denier, we could have another advisory, the sounds like, look this proposal. you know, Cloud Moss made the proposal, you know we worked through usually like you know, we want to

Greg Osuri: Make sure that there for anybody can, can can build that was just not the proposal who did the original work on on working group, right? So…

David Tse: and,

Greg Osuri: if cloud moss, if you put a proposal, we the steel comedy of a should be responsible to ensure that anybody in the community can pick up the work, right? So? So in that case, the anybody can post a proposal and still recommended will recommend as to who the right person who the best person to write this if there is any and if they have any concerns or post that, or they have the strong reasons why they should be funny posts on approval, communities, more like a advisory committee. That’s what how you are thinking right now. Of course, a lot of the nuances as we worked out the working group, but but a little different from what you used to in the community group. Our community.

00:35:00

Max: Again something, yeah, a blank in my head. It was a the cab. It’s kind of like the following of that or the evolution of the cab. Slash accelerator.

Greg Osuri: Yeah, Cab. This is a lot more public, it’s a lot more on chain, right? A lot more like a lot more direct.

Max: Yeah.

Greg Osuri: A cab is still like the approver, right? We don’t want an approval process, what we rather want and advise any mechanism, if they make sense.

Max: Yeah. All right, thanks for answering.

Greg Osuri: Always Anil, you have a question. If not, I would love to give the floor to to David here.

Anil Murty: I know I was just gonna say what touching already said regarding the say The community pool, It makes sense to have it separate because we have identified a significant amount of work that the community needs to do metal requirements for it.

Greg Osuri: Absolutely. So David we have 10 minutes. We do have a hard server 11 so if you want to take

David Tse: Yeah, cool. Yeah, five minutes should be enough. So I want to present a few thoughts and that those thoughts started from my discussion, I twitter space. I had with Greg. And Greg said, Hey you know, since Babylon is delivering Bitcoin security, the cosmos, maybe we can use that to reduce the cost of The budget, the security budget and as Greg mentioned, as the inflation rate of Akt goes down,…

Greg Osuri: If?

David Tse: it’ll become less attractive for stakers to stick and so the The problem is that there may be a reduction in the security and so here’s one. Thought on how bitcoin security could potentially help on that regard. Okay? So I created a few slides. While you were talking Greg, so

Greg Osuri: You want to share the slightly presentation?

David Tse: Yeah, so this is a spontaneous. Okay, let’s see…

Greg Osuri: Oh, okay.

David Tse: if this works.

Greg Osuri: Professor say science.

David Tse: All right, so I have this thing here, okay? People seeing the The screen.

Greg Osuri: Yes.

David Tse: Okay. All right,…

Scott Hewitson: Yep.

David Tse: so just to give a background. So we are Babylon and we are creating a project that just because security to the cosmos change. So, We’re launching our testnet next week. And on this test net there is Babylon. Which is our chain here in the middle. And there are seven chains right now integrated with Babylon and Akash is a very early support of our project. So thank you very much. All right, so this is kind of abstract, right? What does it mean by bringing bitcoin security to Cosmos because, you know, Cosmos change that their own security already? What does that really mean? So, I want to explain a little bit about that and then get to the question. I want to raise, so just a little bit background.

David Tse: So every chain has the notion of economics, security and Bitcoin itself has the notion economics security. So here I am proposing one. We’re thinking about bitcoin economics security is through this plot. Okay, so on the x-axis is the confirmation time that is a time. You wait to confirm a transaction and on the y-axis, is the economics security of that transaction, and on bitcoin, the The time you wait to confirm a transaction is economics. Security is proportion to that. So in other words the longer you wait the more security you get. So for example coinbase, when they confirm a bitcoin transaction that usually wait until it’s two block two block people, they caught two confirmations, and that gives a certain level of security. If you want more security, you can wait until four block people six block deep.

David Tse: And in fact, six blocked, it was a pretty standard number people use as well. So as you increase the confirmation time, your security increases, okay? So this is a typical economics, security, so profile of Bitcoin this week. Now, if you go to approve, a stick change, like Akash then the profile looks quite different. Okay? So, again, I’m plotting as a function of the confirmation time, the economics security, So where’s Bitcoin you get essentially zero security when you confirm instantaneously? In a proof of stick chain, like Akash. When you confirm instantaneously, you already get the security level, which is proportional to the amount of steak 80kt value. Okay. So therefore the more stake value, the more security you get. That’s one observation that’s pretty standard but what I want to emphasize today though is that

David Tse: This stake value in contrast to bitcoin security. Is flat as a function of confirmation time. So in other words, you do wait longer to confirm, you don’t get more in security, you don’t get more security, all you get is a security given by the validators, the stakers of the akash network. So you get instant financial, you get a good look some security, but you don’t get more by waiting and like Bitcoin. so, what Babylon does is if in effect, it allows a composition, Of security of Bitcoin worth a cash, okay? Because Akash Is connected to Babylon and connected to Bitcoin. Through the protocol. The details of which I don’t have time to talk about today. But at the from the user level, or from the security level one can think of, I’m just adding the bitcoin security with the cash security. So if I add

00:40:00

David Tse: The first two plots that show, then I get this new line here, which is now the sum of the two lines that we plot earlier. So here you can see that. if you confirm transactions instantaneously, you get the stake 80 Akt, value security, Which is a standard, but now. As you wait longer longer, okay? And confirm instead of instantaneously confirming. Then you start enjoying more and more bitcoin security added on top of a car security. So you get this nice green line here. Okay, increasing All right, so this is the n effect in terms of the security as a function of confirmation time. Now going back to the question of how to reduce the cost of security of akash. Okay, so here’s one thought so If you think about the transactions on a cache, okay? Maybe some transactions.

David Tse: Don’t really need this instantaneous latency. So if you can shift some of the transactions are cash to longer latency, okay? Then that means these transactions can receive the benefit of bitcoin security. And this will reduce the reliance on the security. Through the stake attached. So therefore, this can potentially reduce Because you’re shifting some of the transactions on the couch to longer latency. Okay, so this is the subject, that’s it, that’s the thought. Are there any questions about this?

Greg Osuri: It’s amazing. So to summarize, obviously borrowing bitcoin security means Increasing the latency means longer transaction times, which makes total sense. For certain transactions.

Greg Osuri: That like staking or unsticking moving money of some sorts, right? You can have longer security or short security, but for certain transactions for like, deployment and whatnot could be, could be lower security, right? Because the Someone attacking an individual deployment. You know, versus someone attacking or trying to, like, steal money from from the state value is very, very, different profile of transactions. You could argue that someone, you know, attacking a deployment. Per se, you know? Could could essentially be presented with the lower security model because they’re down by the cost but I can totally see where this is going. The I really like the idea of being selective per transaction for security and that’s something we have never seen before happen. This gives us a much better.

Greg Osuri: you know, Model This is a very heterogeneous approach to a very homogeneous thought point that we have right now, Very cool.

David Tse: Yeah, okay, so this is just thoughts to get the discussion going. And will be in person I guess in the week or two to trash this a little bit more.

Greg Osuri: That’s awesome. Hello. So Boz has a question. How does it work with mesh security? What’s the relationship?

David Tse: Yes, yes, yes, very interesting. So, so much security by Osmosis team and Juno they’re working on this mass security. Basically allows the stickers. To cross-validate. Okay, cross validate. So in other words, for example, osmo, Stickers can stick on a cash. Okay, so that would increase potentially the economics security of Akash because now, you can think of that plot, the Akt will be Akt, plus some osmo stick. Okay. So that’s one thing. Now, however, there are some concern that, hey, you know, what happens if like all the osmos, they could decide to stick to the cash, then you can’t lose the sovereignty of Akash.

00:45:00

David Tse: And so what? Sunny proposes. Okay, let’s put a cap on how much you can cross stick. Now, once you put the cap though, you see that the gain is actually not a whole lot, not a whole lot. So, in informal system, there’s some calculation, you’ve this cap is like, 10% or 20% Then the overall improvement security is relatively little. So there’s a trade-off between sovereignty. sovereignty and security there in the across, the King solution of Sunny, I think We’re here in this method. There is really no trade-off in sovereignty and security because a costume has the full sovereignty. So you can certainly can, there is so in other words yeah. Because you know, So that’s sort of the advantage of having this notion of sort of composing security through Bitcoin.

Greg Osuri: So, really the trade-offs with Babylon is latency with trade-offs. With mesh is sovereignty. That’s really…

David Tse: Yeah exactly,…

Greg Osuri: what we’re looking at.

David Tse: exactly, perfect perfect. Yes, that’s a perfect description of the, the various trade-offs.

Greg Osuri: Great.

Greg Osuri: This is incredible. I love this so much. So much Explorer. I’m really excited. If Akash implements a share security, use Babylon, I think we’ll Set a new standard as to how app change, you know. Are evolving When it comes to modernizing security.

David Tse: Yeah. Yeah,…

Greg Osuri: It’s really cool.

David Tse: I want to highlight the fact that it’s actually complementary is completely complementary is not it’s not either or it’s not like you either do magic security or you do because security you can do both. And in fact, even with like, ICS in touch interchange shares security Security.

Greg Osuri: Initial security.

David Tse: Like for example, You know, these are new usdc chain coming up, which is we can also get bitcoin security that so it’s not so. So in some sense what we’re trying to what we’re seeing here is that there’s a lot of different types of security but you can mix and match compose them and find the right trade-off points to operate at

Greg Osuri: Perfect. Where we got five more minutes. We have a hard stop at 11. But Pacific any questions, any concerns. Thank you so much David. This amazing

David Tse: Yeah, it’s great.

Greg Osuri: Fantastic. If there’s nothing we can wrap this up and I assume we are, we’re okay with creating the working group, but I want to get a confirmation from everybody. In the, in the group. and we will obviously, you know,

Greg Osuri: Release the details of the working group, the goals and be very, very clear and specific as to as to how we can do that. Obviously David’s gonna be part of the working group to define the security Babylon for Prakash. And if you’d like to be part of the working group, please do let me know how to get. Hold of me. Happy to include you. And we’ll working group will meet obviously in a much regular and a different setting. Um, then this and the working group will be responsible for rent paper. So, With that. Thank you so much and see you in the next thing.

Scott Hewitson: All right. Have a good one, everyone.

Andrew Gnatyuk: Thanks. Bye guys.

Joao Luna: Bye.

Scott Hewitson: Here.

Meeting ended after 00:49:21 👋

Last modified on September 21, 2023
Edit this page on GitHub
← Economics SIG Meeting #2
Economics SIG Meeting #1 →