diff --git a/sig-economics/README.md b/sig-economics/README.md index 88f147a8..9630ac03 100644 --- a/sig-economics/README.md +++ b/sig-economics/README.md @@ -24,8 +24,8 @@ Meetings happen every [First Wednesday of the Month](https://calendar.google.com | #10 | Wednesday, November 1, 2023 10:00 AM PT (Pacific Time) | [Link](https://github.com/akash-network/community/blob/main/sig-economics/meetings/010-2023-11-01.md) | [Link](https://github.com/akash-network/community/blob/main/sig-economics/meetings/010-2023-11-01.md#transcript) | [Link](https://6cjhlgaiutfckbasf6sqjbypkrhyjgll463t67f7aw3af5gtghea.arweave.net/8JJ1mAikyiUEEi-lBIcPVE-EmWvntz98vwW2AvTTMcg) | #11| Wednesday, December 06, 2023 10:00 AM PT (Pacific Time) | [Link](https://github.com/akash-network/community/blob/main/sig-economics/meetings/011-2023-12-06.md) | [Link](https://github.com/akash-network/community/blob/main/sig-economics/meetings/011-2023-12-06.md#transcript) | [Link](https://czik4rsznmpaxems7ffmui4fiaar7waib6gtwy5jt4pptosygpla.arweave.net/FlCuRllrHguRkvlKyiOFQAEf2AgPjTtjqZ8e-bpYM9Y) | #12| Wednesday, January 10, 2024 10:00 AM PT (Pacific Time) | [Link](https://github.com/akash-network/community/blob/main/sig-economics/meetings/012-2024-01-10.md) | [Link](https://github.com/akash-network/community/blob/main/sig-economics/meetings/012-2024-01-10.md#transcript) | [Link](https://gbxszp7zaubg3dskphdcgwpqkjtz3mb3zgtq46ka2yyzzkfij7na.arweave.net/MG8sv_kFAm2OSnnGI1nwUmedsDvJpw55QNYxnKioT9o) -| #13| Wednesday, February 07, 2024 10:00 AM PT (Pacific Time) | | | -| #14| March, 2024 10:00 AM PT (Pacific Time) | | | +| #13| Wednesday, February 07, 2024 10:00 AM PT (Pacific Time) | [Link](https://github.com/akash-network/community/blob/main/sig-economics/meetings/013-2024-02-07.md) | [Link](https://github.com/akash-network/community/blob/main/sig-economics/meetings/013-2024-02-07.md#transcript) | [Link](https://6rzdyy2kxzn4qpilczq7vdvfd52kxb65rk4cnhor32f4cibhkqmq.arweave.net/9HI8Y0q-W8g9CxZh-o6lH3Srh92KuCad0d6LwSAnVBk) +| #14| Wednesday, March 06, 2024 10:00 AM PT (Pacific Time) | [Link](https://github.com/akash-network/community/blob/main/sig-economics/meetings/014-2024-03-06.md) | [Link](https://github.com/akash-network/community/blob/main/sig-economics/meetings/014-2024-03-06.md#transcript) | [Link](https://l7relb4v4vhxhw4shqui4iyoyyeqdt6ogwpx2thybmneridemfnq.arweave.net/X-JFh5XlT3PbkjwojiMOxgkBz841n31M-AsaSKBkYVs) | #15| April, 2024 10:00 AM PT (Pacific Time) | | | | #16| May, 2024 10:00 AM PT (Pacific Time) | | | | #17| June, 2024 10:00 AM PT (Pacific Time) | | | diff --git a/sig-economics/meetings/013-2024-02-07.md b/sig-economics/meetings/013-2024-02-07.md new file mode 100644 index 00000000..5e2f15e5 --- /dev/null +++ b/sig-economics/meetings/013-2024-02-07.md @@ -0,0 +1,215 @@ +--- +title: "Economics SIG Meeting #13" +key: "sig-economics-meeting-13" +date: 2024-02-07T18:00:00Z +contributors: + +- Adam Bozanich +- Adam Wozney +- Alani Kuye +- B S +- Cheng Wang +- Denis Lelic +- José Henrique Targino Dias Góis +- Max +- Rodrigo Rochin +- Scott Hewitson +- SGC | DCnorse +- Tyler Wright +- Zach Horn + +--- + + + +# Akash Network - Economics Special Interest Group (SIG) - Meeting #13 + +## Agenda + +* Updates since last monthly sig-economics meeting. +* Updates on provider incentives. +* Open up to community for ideas or concerns. + +## Meeting Details + +- Date: Wednesday, February 7, 2024 +- Time: 10:00 AM PT (Pacific Time) +- [Recording](https://6rzdyy2kxzn4qpilczq7vdvfd52kxb65rk4cnhor32f4cibhkqmq.arweave.net/9HI8Y0q-W8g9CxZh-o6lH3Srh92KuCad0d6LwSAnVBk) +- [Transcript](#transcript) + +## Participants + +- Adam Bozanich +- Adam Wozney +- Alani Kuye +- B S +- Cheng Wang +- Denis Lelic +- José Henrique Targino Dias Góis +- Max +- Rodrigo Rochin +- Scott Hewitson +- SGC | DCnorse +- Tyler Wright +- Zach Horn + +## Notes + +**Introduction and Updates** +- Scott initiated the meeting, gave a recap of meeting #12 and any outstanding items in discussions or on-chain items related to sig-economics. + +**Discussion on Provider Incentives** +- Provider incentives is currently being voted on-chain via [Proposal 246](https://www.mintscan.io/akash/proposals/246). The spend is only for Phase 1 of the program and is ~$3.5M worth of AKT. +- Cheng lays out the three phases of Provider Incentives Program (PIP). Phase one consists of a large committed pool of resources that require 1 year contracts. This is to ensure that there is a baseline level of supply that is not subject to individuals or companies adding/removing resources regularly. +- Phases two and three will be more closely like a "liquidity pool" where rewards are disbursed based on configuaration, uptime, and other metrics. This phase is still in progress and requires more thought on ways to operationlize effectively. +- Alani asked Cheng who owns the providers/hardware that is going on the network? Cheng responded that Data Centers and other individuals that go into the 1 year contracts are the ones who actually own the hardware and any depreciation or tax implications would be from their end. +- Cheng also states that while these providers are being paid for by the community pool, the revenue collected from these machines will go directly back into the community pool. +- Scott also calls out that everything that is being done will be be reported on as information arrises and funds are spent. There are no hard dates for reporting but Overclock Labs will be as transparent as possible given some information is commercial contracts. + +**Other Discussions** +- Alani emphasized the point that individuals need to actually be reading proposals both in Github Discussions and when they go on-chain. He has concerns for when people read a few lines and sharing their opinions without fully understanding what is being proposed. +- Cheng brought up the point that the Cloudmos team's six month contract for continuation services ended, but that the wallet used for the acquisition still has a significant amount of funds (due to the run-up in price of AKT). The proposal in [Github Discussions](https://github.com/orgs/akash-network/discussions/472) would repurpose lefrtover funds to continue paying the Cloudmos team for an additional 12 months. + +### Action Items +- All are welcome to contribute! Anyone who wants to contribute please visit sig-economics in the [Discord Server](https://discord.gg/akash) + +## 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 everybody. This is the 13th I think economics call second of the year. + +Scott Hewitson: last month's call went a little long. There was quite a bit of discussion about provider incentives and a few questions on the Community tax and inflation props that it passed so I would encourage anyone who did not go to last month's call to read those notes and GitHub. And yeah, I'm not sure if we have too much to talk about to today. Let's see from the discussion side. There wasn't a whole lot going on outside of the provider incentives is currently on chain and being voted on scheduled to pass at the moment. I'll drop the link right here. So if you haven't seen it, please vote whichever way and yeah. + +Scott Hewitson: And then let's see if we had so in Discord not a whole lot of chatter just kind of, follow-ups from last session. So again check that out if you can. And I think for this call, maybe we can just kind of wrap a bow on the last bit of Provider incentives now that it's on chain and then kind of open it up to the community for any questions and concerns or anything that they want to talk about regulated to cig economics at Large. So with that Cheng, do you mind talking shortly of one Probably the fourth time that you've done it but a little bit more about provider incentives just to … + +Cheng Wang: No, I love it. + +Scott Hewitson: give everyone the lay of the land and maybe talk about some of the things that we incorporated since the last call before going on chain. + +Cheng Wang: Yeah, there are quite a bit of work and ideation went into the last version that you saw go on chained for provider incentives program. We're calling the PIP for sure. We like to make things nice and snappy with the naming abbreviations there. So the primary change is I'm going to be is the phasing so phases one two, and three and each phase corresponds to particular bucket of funds that are allocated to GPU resources. So the biggest in phase one is currently what's on chain for approximately 3.5 million USD. All right, things fluctuate a little bit given market prices of akt, right, but either way the phase one is going to be for the committed pool of resources for those of you I'm sure most of you have read through it for those of you who didn't commit a resources. Takes into account some of the feedback from the community which is if we run a short lived. + +Cheng Wang: Provider incentive campaign what happens when those resources drop off? How do we ensure continuity for tenants? All that kind of stuff. Right? So the committed pool is how was expanded right relative to presentation terms to the total phase of total first presentation program in order to sign one year long commitments with various data centers in order to bring those high-end resources from data centers who are professionals at their job and do this for living so that those really resources are reliable. And number two is to ensure that those providers are adequately paid right and not dependent fully on, hosting workload. So we do anticipate these things to sell out very very quickly and they're not being an issue there and third is just continuity, So if you're someone doing some serious work, whether it's AI ml driven or something else and you want to deploy your applic + +Cheng Wang: And your use case you're back in whatever. It is on a caution Network you want continuity, You don't want these resources to disappear in three months. You're not out here just to kind of experiment potentially you might run proof of concept type of workloads, right just to get things started. But ultimately you want them to persist. So that's what the committed pool does. + +Cheng Wang: There for one year long leases and it'll be specifically with data center providers Not just like anybody out there the other two pools phases two and three are what we'll call liquidity pool. So that is going to be your very familiar with how pools are constructed. We'll have specific configurations that will map out. They're more likely be in the meeting like more the consumer grade chips or slash medium lower and data center great chips, right like that save you 100 30 90s 40 90s that ilk of Chip will be in that with those particular pool with us specific configurations necessary right for the network and that'll just be completely Community Driven and anyone who want to contribute those resources can contribute and that's that tier and then there's a final tier face three and these things don't have to occur. Sequentially they can occur concurrently, right just depending on timing and various other factors of when these resources + +00:05:00 + +Cheng Wang: on how they can be secured Etc phase three in the last one we'll call Next Generation R&D pool reason being is AI. mean technology in general moves extremely quickly and AI even more so + +Cheng Wang: And these chips really fall in and out of favor very very quickly. the life cycles are super duper short, 100s a 100 for example, I'm not sure if some of but those chips Nvidia has already stopped manufacturing in 100 in favor of the replacement architecture l40s. And so the l40s chip is replacing the a100 and a lot of people out there even those who are and I'll haven't even really, in their brain retrained it if you will that's a borrowed term to using l40s versus a 100s. So, the Next Generation pool allows the core team to go out there and stay on The Cutting Edge right like AMD support for example, a key core feature. We've been working on for a while. So, amd's enter the chat. So that's one other field of group of chips that we need to really keep an eye on right and in addition to that down coming down the pike is h100 right now. + +Cheng Wang: In the Krav as far as Nvidia is concerned and just generally right h200s are coming out. Then there's the whole Blackwell architecture b100s are in the next two years or so or year potentially and then there's the X Series right that's coming out after them lay around 2025 for NVIDIA and I'm not even sure what amd's roadmap and times like timelines look so suffice to say the R&D Next Generation pool allows us to really experiment in small bursts with some of these Where our chips and architectures to see what the actual supply and demand curve for potentially would look like and to encourage people to use these chips right and get data for the community for a caution and work so we can understand in subsequent iterations of providers incentive for two pip part three, etc. Etc. Can be more tuned to actual data, so + +Cheng Wang: Those are kind of the three phase one is already on chain proposals slated to pass as Huey's outlined and shared and that's what we got there any particular questions or thoughts around that? + +Scott Hewitson: Yeah, just a quick clarifying question. So we got to the committed pools 12 months the next one phase two. The liquidity pool says four months. obviously some of those things can be running concurrently, the liquidity pools probably more interesting to some of the people at home who's got their 40 90s. Do we have any rough timelines on when we're thinking to implement something like that, or is it going to be and after that after the committed pool? + +Cheng Wang: I don't think it'll be after so. + +Cheng Wang: the committed pool once I passes we will like a lot of my new show So sorry if I'm a little boring. I'll try to go over quickly. So once we get the funds we have to get into a point where right now we're talking to, on the core team side a bunch of data center providers as far as you're availability as we get the funds just because get the funds doesn't mean the chips will be ready right away. These things are rather scarce in the environment at this point in time. So, as we get them onboarded in some significant quantities will be able to bring them online. We're gonna bring them online as quickly as we can as far as hotline goes. I can't say with absolute certainty, but I definitely say measured in weeks lower number of weeks, as far as getting them on and it might be within phase one some of those resources might be phased as well because we can't just go to the store and grab a bunch off the shelf and put them on right those in quite work that way so we would love to get the first sub phase if you will face one launch ASAP and more + +Cheng Wang: information will be available as soon as we get that but I don't really have too much to share at this point in time too much concrete. So sorry, I can't directly answer that but What we want to get these things on online and we want to showcase or that continue to Showcase rather a caution networks capabilities, right? We've already demonstrated the entire animal stack right from inference to fine-tuning to training every other, deep in projects out there. As you know said they can do one of the other but no ones actually proven anything in the wild other than us. So we'll continue to work on that. Yeah, Alani I go for it go first. + +00:10:00 + +Alani Kuye: Here. hi Cheng. and part of me. I'm not as informed on this year. or not, a 100s are they? across owned overclock own provider owned data center owned who actually is the custodian of the infrastructure + +Cheng Wang: and Great question. So who physically owns the hardware will be the data center. The actual machines themselves will be data centers and the data centers themselves will become providers on a caution Network for putting their Hardware onto the network as far as who's going to manage the providers for data centers who are comfortable managing it themselves. They will do so for those who are not comfortable managing themselves the cash 14 and ocl core team will work with people in the community as well as using internal resources to manage those providers. as far as not the I guess custodian might be a way to look at it right at the administrator of those resources as So there's one path. + +Cheng Wang: Data center will be the owner of the resources. They will either manage it themselves as a provider or Outsource the provider management to ocl core team VZ like the community right in some way shape or form in that way. then so that's how those resources be put on to the network. + +Alani Kuye: So from a tax perspective who gets the carpets and who gets the Opex. If I'm gonna depreciate that over four years for example, right you sir,… + +Cheng Wang: So yeah. + +Alani Kuye: I'm going with it, right. + +Cheng Wang: Yeah, so the benefit the capex depreciation will definitely come out of the data center. so let's say I am data center, AI machines.com, right? And I have a bunch of a 100s or H 100s and I decide to sign an agreement with ocl to put these resources on the caution Network. So then at that point, it's simply a commercial agreement from the data Center's perspective. That's all it is. Right. They get paid their money for putting resources on to network and that's all gravy there and thats not That's simply just a very plain commercial agreement between overcog collabs and a data center or several data centers. And so the data center in this case will operate normally would they incur their Opex right and their capex? It's all baked into whatever they charge overclock labs and the kind of Comm. + +Cheng Wang: Right over a year or if it's on demand and other whatever the scheme is, right the duration is that's all baked into the price. So they take their capex depreciation on those chips. And obviously they incur the Outbacks of it as well as far as but the other operation side the only nuances in terms of Provider management for the sake in the instance where caution Network core team manages the provider right? That's where the operational cost will come out of the provider incentive pool, right? Because there is some Elementary. So for example, there's a Community member. Let's just say who manages a provider or set up providers? They would have to be paid for their work right? So that there's of the pool. + +Alani Kuye: Okay, I was just trying to get a better and thanks to clarifying that because I was thinking if I 100 K and it costs 120 for your depreciation plug the economics, cognit on the books and either way I get it back somehow right? Someone's gonna you put the taxes I can offset that. + +Cheng Wang: exactly + +Alani Kuye: From an accountant. + +Cheng Wang: Yeah. Yeah,… + +Alani Kuye: Alright. Thanks. + +Cheng Wang: totally. So if you get a hundred dollar chip, Let's just say and the lifetime value of it is let's say five years and do linear depreciation. You can write off 20 bucks out of the hundred per year for five years. And offset, whatever your income is right in those five years with that particular thing and then whatever your Optics is, obviously that takes your bottom line down as and then whatever revenue is your Revenue so that all go against your state at Revenue there so for data centers, in this case. they're only revenue is directly going to be the contract that they get paid on and so in the event, this is something really interesting too. I'm not sure if we were able to put that in discussions, I think we might have which is + +Cheng Wang: Data center for gives sign data center science contract with ocl bring resources on Kosh Network right save those resources are $100 right costs those overclock Labs $100, but that's been funded by the community pool that passes then to the data center that agreement is tied off all clear then from the tenants perspective, right? They're going to use resources on the network that is managed by who Clash Network, Minister administrator, shall we call it and those revenues that flow through ultimately + +00:15:00 + +Cheng Wang: Where the zacho to we're going to all the excess net of paying expenses. I think this is what we talked about. We'll go back into the community pool. So if we buy resources in the community pool for a hundred dollars, right and obviously we're going to ensure that the prices are highly competitive and not overtuned right but having said that once the inflows come in net of whatever the expense are. Let's say, Alani is managing a provider and he gets paid, five percent out of this whatever is coming through. I'm just throwing out random numbers right then 95% of those revenues would go ocl a cautionary core team that manages provider will be putting that at the end of the program back into the community pool. So it'll be this pretty virtuous cycle where the community pool will be ultimately refilled by usage. Yeah anytime. + +Alani Kuye: Pretty solid. Thank you. + +Cheng Wang: B's had two questions as the PPA elected by Syrian receive a salary at our parent did that I don't think that PPA receives. A salary per se like so we're gonna talk through it on the steering committee. the ask there might be some consideration for administering the program. Like I said before as far as costs go those haven't been really super clearly defined. for one thing as far as who is the PPA I think + +Cheng Wang: What is it provider program in administrator as we're calling too many acronyms here. I think that administrator I personally would nominate ocl because we have the resources and Personnel to do it and we've always been extremely transparent to the community there. So that's the number one. So, yeah, in terms of costs, they're likely will be something but likely if ocls elected as the primary administrator, we likely would not take anything as far as a fee goes because that's just kind of cost of doing business for us. Number two, there's transparency reporting requirements. They're actually how regularly I'm too apologies already. No, so we don't know what the Cadence ought to be. You can definitely expect them regularly inconsistently. It won't be regularly every six months right be something like that, but we just don't know exactly what the Cadence is because we don't know as far as reporting how the date is gonna be aggre. + +Cheng Wang: how long it's going to take us to Crunch that information put it into something that's shareable but suffice to say I think you can believe and it don't take my word for it. I just look at all the community governance proposals. We've submitted in the past and all the reports that Scott Huey here has been at the helm of and pushing those opposite Community as far as putting money back into the community pool as far as access funds go number one number two showing, all the reporting what was called Google. + +Cheng Wang: Data Studio things called looker studio now so all those reports are out there and shared with the community to view and we do Endeavor to do the same thing here. Thank you Scott for sharing that so something like that. You can expect something like this and obviously if there are questions and things like that, we'll post updates because this is such a huge thing will likely post updates maybe and GitHub discussions should there be enough for it at the very least. We'll be talking about it on seeing economics it provider, across a whole bunch of things and working groups as well because this is the biggest program launch today and they're gonna be a lot of questions and a lot of curious mind. So we'll definitely share the data very very ask consistent as we can but also not So frequently, let's just say daily or weekly that it kills us on the administrative side,… + +Scott Hewitson: Right. Yeah,… + +Cheng Wang: Because we do have a lot of things are juggling. Yes, Scott. + +Scott Hewitson: so I think just to add to those two questions from BS the PPA if any work is coming from overclock. We're not going to be taking money for effort. if there's other folks in the community that need to help there that can help out then that would probably make more sense because people need to be compensated overclock Labs. we have our own payroll we have our own thing. So we're not using Community funds to pay that and then the reporting side I dropped in this dashboard from events. my view on this is kind of as things start coming in and that's when we'll get the information out. So yeah, I don't know if it's a quarterly probably is, I can count on that for sure just because + +00:20:00 + +Scott Hewitson: we should be doing that but really, the second I learn about an expense or it's realized or whatever I'll put it up and hopefully just have it running. So there's a lot of transparency behind it. everything in the community pool it's the community pool. So it's kind of like a public programs like governments. It's like we should be showing what we're spending money on and how that's growing the network. So hope that helps add some flavor. + +Cheng Wang: Up Bravo. Thank you Scott. it perfectly said + +Scott Hewitson: awesome Does anyone else have any questions or concerns related to provider incentives prop 246? + +Scott Hewitson: I'll take that as a no if you do if you think of something or you don't want to talk about it here feel free to drop in Discord and we'll be sure to get you a good response. At this time, I'd like to Alani. What's up? + +Alani Kuye: Yeah, just very minor commentary, thank you guys for all the work you've been doing. I also just want to emphasize that it's important for people to read. these proposals in their entirety before quantificating I think oftentimes It's very concerning when people just read four or five lines without really comprehensively understanding. What a proposal is about and just firing off a grenade and opinion basically, so I think we should really spend more time within these things and really understanding it. So we have literacy around contacts and where it's going because we want to be careful. We don't get so bogged down and tactical mindset and not meet the overall strategic objective. That's just import from my side. + +Scott Hewitson: I think that's a great Point Alani. I mean we live in a world where everyone reads headlines and makes their decisions off of that. So, I mean, it's easier said than done but for areas where there's pontification or grievances or I think we've been able to do a good job to alleviate any concerns or at least address them and explain, An altering point of view,… + +Alani Kuye: No. + +Scott Hewitson: so I agree with that 100% + +Scott Hewitson: awesome + +Scott Hewitson: We still have quite a bit of time. So if there's anything you want to discuss any ideas you have anything related to Sig economics or anything that's growing the network the floor is yours if you want it. + +Cheng Wang: I did have one quick one if I can shoe horn myself in there for No,… + +Scott Hewitson: I suppose. + +Cheng Wang: so I will not be seating my time someone else now in this case. I think the other big thing is we put in to get up discussion to wanted to give a little bit Time air time for the provider incentives want to marinate a little bit. But one thing we did want to and the pain of transparency is for the continu. + +Cheng Wang: Of cloud most right and they're continued contribution to overclock Labs. I put up a discussion there just to keep things on total up and up and that we are extending our relation Cloud notes because the team is great and obviously they've worked for years. Now. I'm within a caution Network in a fantastic fashion. So we're continuing that for at least another 12 months hopefully more but the slated agreement right now is for 12 months and we're repurposing the funds from the original funding of the six months and as well as the cloud most acquisition there is some access funds left over to support that and I think there was one set of questions from ZJ, which is he has some questions. Will it be using you denominated payments as far as what the expectation the Costco it's kind of very little bit based on hours worked and stuff like that, but it'll be slightly above what the + +Cheng Wang: I'm run rate was for the first six months. That's number one and number two the funds vest over the next one. So it'll just be monthly payments. So it wouldn't be a whole big shot all up front with the acquisition that vests over time. It'll Simply Be work that's been done. it'll be like a contract So a salary if you will continue to work being done and I'll be paid out over time on a monthly basis and that'll be denominated. Whatever the Says at the time of payment we snap it there and that's historically how we've done it. + +00:25:00 + +Scott Hewitson: Awesome. Thanks for covering that a little bit + +Scott Hewitson: is there anybody else that wants to bring up anything can be related discussions proposals any ideas? Yeah. Leave it open for a minute. + +Scott Hewitson: No takers this time because Andrew G's not here to ask me five questions to keep me honest. I love that guy. No, I mean last call was I think was an hour and 20 minutes. We're at 30 minutes now. So, I think a lot of stuff got hashed out in the last call. If I think we can wrap up at this time. Unless anyone has any last things they want to say and if you do have any, comments concerns questions, whatever you want. Feel free to drop them in Sig economics and Discord. I checked that very regularly and try to be very responsive there. so with that, I think we're going to end the call and we'll see y'all next month and + +Alani Kuye: All right. Thank you. + +Scott Hewitson: alrighty + +Cheng Wang: Thank you everyone. Thanks Alani good to see you. Yeah, right. + +Zach Horn: thanks guys take + +Meeting ended after 00:26:37 👋 diff --git a/sig-economics/meetings/014-2024-03-06.md b/sig-economics/meetings/014-2024-03-06.md new file mode 100644 index 00000000..f08adc1f --- /dev/null +++ b/sig-economics/meetings/014-2024-03-06.md @@ -0,0 +1,309 @@ +--- +title: "Economics SIG Meeting #14" +key: "sig-economics-meeting-14" +date: 2024-03-06T18:00:00Z +contributors: + +- Adam Wozney +- Andrew Gnatyuk +- B S +- Cheng Wang +- Moss Business +- Robert Del Rey +- Rodrigo Ronchin +- Scott Hewitson + +--- + + +# Akash Network - Economics Special Interest Group (SIG) - Meeting #14 + +## Agenda + +* Updates since last monthly sig-economics meeting. +* Updates on provider incentives. +* Open up to community for ideas or concerns. + +## Meeting Details + +- Date: Wednesday, March 6, 2024 +- Time: 10:00 AM PT (Pacific Time) +- [Recording](https://l7relb4v4vhxhw4shqui4iyoyyeqdt6ogwpx2thybmneridemfnq.arweave.net/X-JFh5XlT3PbkjwojiMOxgkBz841n31M-AsaSKBkYVs) +- [Transcript](#transcript) + +## Participants + +- Adam Wozney +- Andrew Gnatyuk +- B S +- Cheng Wang +- Moss Business +- Robert Del Rey +- Rodrigo Ronchin +- Scott Hewitson + +## Notes + +**Introduction and Updates** +- Scott initiated the meeting, gave a recap of meeting #13 and any outstanding items in discussions or on-chain items related to sig-economics. + +**Discussion on Provider Incentives** +- Cheng Wang provided an update on the Provider Incentives Program (PIP) since it has passed on-chain governance and funds have already been used to secure contracts. +- As of the time of this meeting new resources such as 4090s and H100s have come on the network for available use. +- Andrew asked if these resources are already being leased or if there are use cases that are coming on to the network soon to use them. Cheng responded by saying that there are multiple tenants that the core team has been talking to and expects the providers to fill up. +- Andrew asked for further clarification on whether we are incentivizing resources while they are being used or just to have them up. Cheng responded by calling out that these new providers are signed contracts so they will be up but obviously don't want them running empty. Any funds they make will also be sent to the community pool. + +**Other Discussions** +- Moss Business, a small content creator, asked about security issues related to the network or if there were some SIGs on security. Cheng suggested going to [sig-providers](https://github.com/akash-network/community/tree/main/sig-providers) as some of the questions that Moss Business asked are being worked on there. +- Andrew asked where we are in AKT 2.0 and what still needs to be done. Scott responded by calling out the [AKT Token Page](https://akash.network/token/) that shows the main features being: + - Take and Make Fees - Live + - Stable Payments and Settlement - Live + - Incentive Distribution Pool - Not live but make/take fees are currently going to the Community Pool + - Provider Subsidies - In progress with PIP + - Public Goods Fund - Not implemented but Community Pool has been the placeholder and where contributors can seek funds to grow the network. +- Andrew brought up Io.net raising $30 million to compete, how is the team thinking about this? The core team's stance is that the pie is large enough for many players and competition is welcomed and good for the space at large. + + +### Action Items +- All are welcome to contribute! Anyone who wants to contribute please visit sig-economics in the [Discord Server](https://discord.gg/akash) + +## 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, all today's call. This is the cig economics monthly call for March. We've had quite a few of these I will call out that I did not get the notes up from the last call. So that's squarely on me. We've had a ton of things going on. So I will make sure to get that done shortly after this along with these notes, but that is on me and my bed, so kind of kicking things off. I was looking at proposals that have passed since the last call we had the provider incentive pilot, which we talked about. It was actually on chain while we had this last call which tasks and then also kudos to the folks doing the developer onboarding proposal via zealing that passed as well. So excited to see what comes of that. + +Scott Hewitson: And from discussions there wasn't anything major that I saw and then from the Discord chat. It's not too Lively. But sometimes we get a questions a few answers. I think today we can kind of talk about obviously the provider incentives program and where we're at with that since the proposal is passed and I'll pass that over to Cheng in a minute and then as always we can open it up to community members if you have Any questions concerns or just want to jump on your soapbox and talk about economics. So with that Cheng if you don't mind you could give us a little. + +Scott Hewitson: quick recap with where we're at with the PIP phase one phase one is the only one that went on chain and passed so we can just talk to that or if anyone has questions about the other phases we can talk about that as well. + +Cheng Wang: Yeah, I'll go through and talk about phase one real quick and touch on the others and we can open up for discussion from there. phase one past resoundingly and of course it's been great since then, prices have appreciated thus far the market I think is responded the economics environment I think is improving. So we'll see, it's going really well. We have thus far engaged to Providers one and that's already existing one. So we work with both of these before as soon as the information is finalized. We'll of course communicated to everyone publicly but we are throws of getting those resources signed and spun up and I think it's now coming on starting to come online. I think yesterday a bunch of h100 came online. So I don't know if anyone saw that the latest and is located in Quebec and Canada, and so we're starting to get all those fun up. So in terms of the resources that we + +Cheng Wang: Be targeted based on the discussions there a 100s a6000 and 40 80s or 40 90s. Excuse me. We are getting all those spun them is just gonna take a little bit of time to nail all those things down. and really looking forward to it. I think the resources are all we currently in terms of pace as far as the allocation the budget goes for phase one we spend about I would say 60% year marked about committed committed about 60% of the phase one budget ready to these resources some have come online some are in the process of coming online. And so there's still a little bit of ways to go as far as finding the additional resources that we want to Target. So that's phase one. + +Cheng Wang: Phase two is still in design the stuff around that we were talking about it internally in amongst according yesterday the composition of that and the execution and especially operation of it it's gonna be challenging for sure because it's not on chain deck stuff right where we're doing it fairly manually to start. So there's that and then phase three obviously will be a little bit more kin to phase one. So low on the operational load and just gonna be more so on the back end work that the core team has been doing + +Scott Hewitson: Thank you for that update with Pip. Yeah, does anyone have any questions or concerns or anything else? They'd like to know about provider incentives Andrew. There we go. I got my man and asking questions. + +Cheng Wang: a price of PR + +Scott Hewitson: He's here to keep us all honest,… + +Andrew Gnatyuk: Yeah. Not just a quick question and… + +Scott Hewitson: and I love it. + +Andrew Gnatyuk: maybe I missed it somewhere on the previous six for this h100 so I saw them on the stats do we have clients to use them? do we have some will use cases that are gonna on board and we will see that they are already released or something that + +00:05:00 + +Cheng Wang: Yeah, so the resource I think just came online yesterday. And so we do have some organizations that are looking to use them already. And so we're gonna get that spun up. That's number one. And number two is we have some Partners who I don't want to necessarily announce you get in and I say Partners likely but we're working with a couple of companies on the kind of client side who are developing toolkits themselves to enable fashion appointment and all that kind of stuff, right? And so + +Cheng Wang: They're integrating Akash as we speak and we'll make a caution Network an option for people who are using, what they're building on top of just the raw machines themselves, And so once that comes online, we should see a lot of these things fill up. So as far as in my head what I know, I think they're at least two to four two for sure and probably four if not five very high potential use cases for these machines. We definitely won't have them up online running, vacant right? Because we haven't really pushed it out. And so once Step One have these resources set to connect them with the folks who we know already working on connecting them. And so that's really in progress step three is work on the partners where I guess + +Cheng Wang: Yeah, Partners those just say who need these resources and continue develop them so they can use them and the final bit would be to not the final bit. But the next thing is advertise, right really promote this as soon as we know the resources are stable and tested up online running all that kind of good stuff. + +Andrew Gnatyuk: Some got it. Thank you. + +Scott Hewitson: Does anyone have any other questions regards to the pit or one phase two? Yeah, we got some background chatter. There's a TV and Cheng's background. + +Scott Hewitson: awesome Yeah, I mean I don't know if we have a ton to talk about on this call. I did want to call out a few folks. I see b s I believe you've come out on a couple of these calls and then I see moss business so welcome to you folks, if you want to introduce yourself feel free. If not, also that's fine. And we're always looking to having people join these calls, we get a ton of great ideas out of these calls from coming from wherever so yeah Andrew, what's up? + +Andrew Gnatyuk: If this one more question to clarify, so if I understand correctly, we are incentivizing the providers when their gpus got least right not when they just go on the market. + +Cheng Wang: no, so for phase one because for phase one the way we're doing it is we go to the providers for these high-end resources and we sign long contracts with them and then pay them for it. So in this particular instance, we cannot just say hey put these resources on a caution and we'll pay you if there's usage we have to commit to it and it's just one of those things where it's a chicken and egg challenge, right? We have to have those resources online in order to continue to attract the use cases that come on a caution Network and grow the network. So that directly answers your question there. And so the other part what that means is because they are already being paid. + +Cheng Wang: Contract right? So then whatever that gets generated to those particular wallets, as far as least Revenue that will cycle back and go back towards the community pool. So it's kind of like all this virtuous cycle, So let's just say I'm a data center, I sign a contract with overclock Labs because we have to be the entity to face off against someone to sign a contract right for hundred thousand dollars for whatever resources right? Overclock Labs through the community pool pays me a hundred thousand. I'm good. I don't care what happens you can use it you cannot use it when those resources get used. It goes into a wallet that ultimately we'll require those providers to send the funds back to us and it's all on chain and public and auditable so they can't really, do anything squarely and get away with it. So once those funds come back into the wallets ultimately at the end of everything will aggregate at do all the accounting and the reporting to show the community. Hey, here's what the usage was Etc. And then the + +Cheng Wang: access will be of course sent back to the community pool. So that's for phase one phase two will be a little bit different in that. Yeah face-to phase two be a little bit different because it'll be more free for all and open to everyone in the community who has, resources that meet the requirements and so they'll be essentially the same thing. They'll be paid for those resources and uptime right as functions of how we calculate and then + +00:10:00 + +Cheng Wang: For that. I think the guidance will be since you're getting provider incentive to just have these resources online, you're essentially being paid for it anywhere regardless of usage. There will be I think requirements for prices. So you can't essentially egregiously double dip, So let's say the market can bear four dollars per hour. We would likely have that price push down, right if you're already getting some sort of compensation from the pool, but we were so ironing that out in detail. + +Scott Hewitson: Yeah, and go ahead and Moss business. + +Moss Business: Hey, can you hear me? Okay. Awesome. + +Scott Hewitson: Yep can hear you good? + +Cheng Wang: on the clear + +Moss Business: I'm a small content creator. So I just have kind of a discovery question. Mostly interested in are there any sigs where a discussion around security like Security in the Akash Network would be discussed appropriate. + +Cheng Wang: Security in what sense from a deployment sense from a blockchain side of things deployment side of things. + +Moss Business: Correct. Yeah. + +Cheng Wang: Yes, so I would say Sig providers might be where you want to go for that. So when you say security it's kind of like if I deploy something is my deployment secure is the provider trustworthy and all those kind of things So I think for that particular thing, we have one of the first features we started working on has been one provider thanks to you for dropping us. So the auditing right the ability for providers themselves to audit workloads that are running on their machines so that They can then adhere to their own local jurisdictions and requirements. Whatever those are right. So for example, if someone comes in and deploys. + +Cheng Wang: I don't know if Kim Jong-un wants that deploy some workload on to Andrew's machine and Andrew's beholden by law to prohibit those types of things as a provider. He should be able to crawl through and see what's actually running and secure himself, right? that's one part. Another part is then how much can the provider see or cannot see that's another component of it and a final part of it is on So this is from the tenant side if I deploy an image and that image is open and public right? I'm less likely to want to push forward. + +Cheng Wang: Push excuse me proprietary stuff that I've worked on because I don't want someone to just steal my secret sauce, And so provide then Secrets right is another important thing. how can I as a tenant secure my Docker image so that someone can't just easily go in and look at what I'm deploying. So those are all things that are being worked on. If you look at in our community board that Huey just drop there and you click into projects like the product engineering roadmap in there. You'll be able to see the items that I just talked about in great detail and all the kind of discussion history and development history associated with each + +Moss Business: Thanks so much. That's a exact conversation that I was interested in was the tenant security securing a Docker and the ability per provider to look inside of the container and stuff like that. So I'll definitely take that combo to the sake the providers say that you mentioned. + +Cheng Wang: For sure, and then follow along in me and the GitHub it's all open. So if you want to look like in history the development Pace what's been done? What hasn't been done what's on whom that stuff should all be fairly transparent there. If not, just hit us up in one of the things we're working groups. And Community will be more than happy to talk about it. And if you have a question just drop into one of those groups in Discord paying the question and off you go. + +Moss Business: Thank you so much. + +Cheng Wang: Pleasure. Thank you. + +Scott Hewitson: Thank you for that question Moss business and thank you for joining the calls. when I was at East Denver I get a lot of questions right all these projects doing kind of the same thing as you guys what sets you apart and I truly believe that something that sets us apart is this open framework where everything is being built in the open? Everyone can verify everything everyone can jump into any call and see what we're talking about? it's pretty awesome. And if you've been in crypto for a while, everything's open source, and the folks that go down that route end up having success where some of the closed systems may not have as much. So I just want to add that little tidbit because it was something fresh at East Denver. I was just there last week. But yeah. so + +00:15:00 + +Cheng Wang: said + +Scott Hewitson: what that does anyone else have any other questions? I mean we can move along from the Pips stuff. Does anyone have any questions just regarding anything or any ideas or things that you might want to see moving forward chain? + +Cheng Wang: Yeah, it's not Andrew No, so I think the one thing that I want I wanted to exactly what you mentioned regarding what you've observed in East Denver and everything. You said regarding our open framework. I want to emphasize to this group and everyone can do their parts and smaller big however, it is right is there's there as this space in right continues to develop there will be and there are and have been a lot of attempted copy cats where people who want to go into the space because that sector is hot. It's kind of like defiles hot. + +Cheng Wang: through the nose, there's so many networks out there. I don't even care anymore with some stuff like I use what I use. I know what I know. I really don't care about the other stuff. It's just all kind of like the regurgitation of the same idea over and over again having said that there will also be a lot of potential projects and projects that come out that are Very vague not transparent we'll make claims and not back them up all that kind of stuff. Right? And for here at Akash, everything is open. We can't hide behind b s right if it's out there it's out there and we have to just kind of deal with whatever the actual realities of it are. So I think for us as a community everyone here if you can do one thing that'll be incredibly helpful is if we can all do our part in this spelling. + +Cheng Wang: False information misinformation that's being spread about a cost potentially out there wherever Twitter Etc where people say this versus the other and just to let you guys know in our conversations with even funds right who are interested in taking a position on a caution and stuff like that. They come to the core team and ask questions. Hey, how do you differentiate between yourself and project XY project Z whatever it is, right and I have to tell you. + +Cheng Wang: The level of sophistication of research from these funds oftentimes. I would say 80 to 90% of the time is pretty poor. if I were a professor grading these assignments, I would probably give them a d to be kind like they should be able to distinguish these things and I think part of the challenges the information isn't readily necessarily out there as loudly and as frequently as we can hope so, I want to definitely ask everyone whenever you can have the opportunity to say something like hey, here's some information.'s whatever here's what I see, back it up with facts and just continue to push that out there and not get into the kind of the Tit for top on vague. Yes, right. So That's the PSA I have. + +Scott Hewitson: Thank you for that ng. Yeah, as you dropping a cheeky note that we can't hide behind you b s your name is BS so, said that throw that out there. Does anyone else have any ideas or concerns or comments in regards to get economics or the token or how things are running? Yes, that's right. + +Scott Hewitson: doesn't look like we have any there is. Go for it. + +Andrew Gnatyuk: Yeah, That's me. I've missed you. So let's meet you all so let's talk. + +Scott Hewitson: All right. + +Andrew Gnatyuk: Yeah, just a question. you do fraction to memory. Where are we right now with the attached economics 2.0? + +Scott Hewitson: 2.0. Let's see. We've done let me just pull this up real quick. did + +Scott Hewitson: so we got take rates implemented right we've got + +Scott Hewitson: I'm just pulling up everything on this. So the + +Andrew Gnatyuk: Take her and make her face, right? Okay, and + +Scott Hewitson: Yes, those are not flowing to stakers yet that all flows to the community pool. So, something that will need, there's the incentive distribution pool planned and that kind of is like the pool for everything. And then from there it goes out to public goods fun providers and is staking potentially burning and then community pool. So right now, In the interest of Simplicity, We've done those take rates to the community pool. Obviously the community pool has been the source of funding for a lot of great initiatives and + +00:20:00 + +Scott Hewitson: it really just is a bottleneck of engineering efforts. We've got a million things going on. So implementing all of this stuff from akt 2.0 is takes time and it kind of shuffles with importance, when we have something for an engineering task. That's the higher priority that gets pushed up probably simply because we're having success. from the economics side and being able to do everything that we need to do is but yeah, I mean there's a lot of things from it, I think. let me just pull up because I want to say the token page on the website has. + +Scott Hewitson: The five things so the main features make it take fees stable payment and settlement those both are live, the incentive distribution pool that is not provider subsidies kind of with this prop and in the public goods fund which is kind of like the community pool. So I'd say it's a Technically two out of five live, but you could probably say a half from some of the other so maybe two and a half to three out of five complete and we're effectively, working most of them are complete but there's some engineering efforts that we that go behind Those modules of creating the different pools and all that other stuff. I hope that that helps any of those little bit of a ramble as I was going back and forth through resources. + +Cheng Wang: That was great I think one thing that I would love to add to that is. + +Cheng Wang: Of the primary features that contribute most to akt 2.0 those have been done. Right? So the take fee being the primary element of it that in and of itself I think is the biggest economics to be frank like the incentive distribution pool and the sub pools. I suppose below that those are all kind of counting stuff like operational things. That'll be nice to segregate going forward when things can be more tidally organized by potential by group and use case. So that's kind of a nice to have at this very moment in time as the community pool and fun continues to grow that will become more important later on because we don't want to have one big bag of money in the community pool slashing around then the mandates aren't that clear Etc though? We've I think work really hard in terms of when proposals the sequ. + +Cheng Wang: Of getting things up on GitHub discussion answering questions and then on chain Etc writing all these things out. I think we've been really good at that. So that's what I want to say. what we've done so far are the most important things that enable economics 2.0 to start running and that has been done the next big things. I think from a feature and usability perspective is some of the behind the scenes stuff then that will make stable payments more useful because right now all the IVC Channel jumping crap, it just makes USC not as usable user friendly as we wanted to be. So that's I think one thing that's currently on the back burner and as he Illustrated when we have enough engineering band with will be tackling that as well. So yeah, and then in terms of the priorities mentioned previously the secret stuff like provider features Discovery, that's another thing that was just pushed out that is a huge blocker that we needed to on unblock in order to have phase two of the pit be possible right because we need to be able to + +Cheng Wang: Providers and say what resources have you brought on and be able to see all that stuff in a database and record and all that good stuff. but yeah, I think the progress there as far as the organization of the IDP and accounts a little bit definitely lower priority because it's just organization at that point and the big thing after that is just again here we mentioned distributions of stakers. that's going to be a big piece. But as both I think we can cross that bridge once usage climbs up, Quite significantly, right because I think we're just at the tip of the iceberg at this point as far as usage on the network goes. + +Scott Hewitson: Yeah to add to that having the take fees flow to the stakers and validators. I think from a retail perspective that's something that people want that we don't have implemented at the moment but to Cheng's point, There's good Network spend but until it's a point where it's significant. It's like all right, I got my hundredth of a cent, from the take fees for staking a bunch of akt, it's not gonna be as a people aren't going to be like, wow, this is awesome until there is significant network activity Network spend. + +00:25:00 + +Andrew Gnatyuk: Yep. Got it. Thanks. + +Scott Hewitson: first Alrighty, does anyone have any other questions or concern Has anyone want to bring up anything Vlogs? Thanks for jumping in. I don't know if you want to throw any words of wisdom or anything. + +Andrew Gnatyuk: Yeah, I still have one concern. I want to ask chance basically let's See it this way on this bull run. + +Scott Hewitson: Okay. + +Andrew Gnatyuk: Let's say akt is pumping right? We're not talking about much on the exact prices, But let's say eight ten bucks 15 whatever. Yeah, and it turns from a security talking right for the network the more office speculative talking and let's say we have lots of unsticking or unbonding and let's say our network has just 30% or 20% in staking. because right now we have 60 and before that we have 75. + +Andrew Gnatyuk: Is why I was asking about the incentivize for the stakers because at the beginning everybody was like we have great APR and we need to stay because that's just like 380 tokens one. once it's just pumping up, right? But the market shift for example in two years and we will have to deal with with this, lots of tokens just freely looking around the success indexes and that stuff and that's what makes basically concerns me. + +Scott Hewitson: Yeah, I think you… + +Cheng Wang: Gotcha. + +Scott Hewitson: it's important to note that there's a Min and Max inflation parameters and it's basically the amount that you're gonna get and from staking rewards is dependent on the amount state. So the less + +Scott Hewitson: staked there are the staking percentage is gonna be that they're all parameters that can be changed subject to governance and they have been changed with governance proposals. But in general, that in the cosmos ecosystem, the less tokens that are staked the town or balance to that is the staking APR goes up. So if you have everybody staked then it's those rewards of split between those that are stake when there's only three of us staked we're gonna get all of those rewards. So there is that counterbalance to come back that but I definitely understand your concern and some chains even they deal have, low staking percentages and low State bonded percentages and really high aprs and it still doesn't incentivize folks to stake. + +Scott Hewitson: So I think that's kind of on the whole network at large to be able to get to a point where we have that big network activity that big Network spend to where ideally, someday most of the rewards that you're getting are from Network spend versus just the inflationary rewards because that everything is in speculative without Revenue coming back from the chain and the network. + +Andrew Gnatyuk: Yep. Got + +Cheng Wang: Yeah. Yeah, I would add to that is. + +Cheng Wang: Just as Huey said and I wouldn't necessarily like just zooming out a little bit. I wouldn't. + +Cheng Wang: Place too much stock in the micro right now from the day-to-day perspective, because right now the market is still trying to figure out where things are right is akt gonna continue to move one way or the other. I don't know the market will do what the market does and on top of that. There are also governance parameters. We can definitely toggle as well. Right? So the inflation Max and all those kind of things that are things that we can definitely toggle to ensure right to if staking rate Falls below certain perspective and point rather and it becomes an actual concern that is shared by, the core team in the community at large something that as a community everyone right can submit proposals to let's say bump inflation up such that, stakers get a little bit more return right on akt though. I would say if you're holding akt return has been pretty good as play not gonna lie. So the knobs right do exist for us to toggle these + +00:30:00 + +Cheng Wang: Particular things one thing I would have to say is finally on this point. If you look at even though bonded Ray has fallen from the 70 percentage to 60% ish the total value Bond, it has gone up significantly, So the economics S in context it's not just the wall percentage necessarily that matters. + +Scott Hewitson: security + +Cheng Wang: I would argue that by having 60% Bond date at this point in time. We're more secure now than we were at 75% simply because the value of network is grown so considerably so it's a lot harder to Do stuff. + +Scott Hewitson: Yeah, the cost to attack in raw numbers has gone up. + +Cheng Wang: for sure exactly it exactly + +Andrew Gnatyuk: Going agree on that. Okay. Thank you guys. + +Scott Hewitson: No, thank you. + +Cheng Wang: always good to voice us. Thank you. + +Scott Hewitson: Alrighty, does anyone else have any last thoughts or comments or concerns? Hey,… + +Andrew Gnatyuk: Yeah, yeah, I do like one last question what direction for you? Let's go. Let's + +Scott Hewitson: We got some time if you got a couple it's all good. + +Andrew Gnatyuk: So the last news from Twitter ionet has raised 300 million to build the largest decentralized GPU Network in the world. And so the AI compute shortage, what's your thought on that? It's basically like GPU Marketplace and Solana. It's not like I'm advocating Savannah right now. + +Scott Hewitson: Yeah, I mean so it was 30 million. Yeah, yeah. + +Andrew Gnatyuk: I'm fully Akashi, I just have this. Yes. It's Because they give it to me and I was just wearing it today. Yeah. + +Scott Hewitson: Yeah, so no, they congratulations to their team. I think it was 30 million that there is but some prominent backers. + +Andrew Gnatyuk: Yeah. + +Scott Hewitson: I mean, everything we do and Greg is pretty outspoken about this. he says anyone who's In the same direction, building for the same long-term goal, it's competition let the protocols, the proof is in the pudding I guess so, it's not that it's a distraction or whatever, we're here to build and we're here to do our thing. if anything it's a validation that people are looking at this there's a need for this and people are putting money behind it. So if anything I think it's good news for Akash that other companies are raising to do what we are already doing. + +Cheng Wang: Yeah, I mean this is a way thing about it again zooming out a little bit. It's trillion dollar market. Let's just be frank. Right and so there's more than enough space for everyone to Palo Al around and so if let's just say you look at the actual true market leader in this space as far as the cloud space goes right and then GPU space if in AMD for example is no slouch and they're still kicking but even though Nvidia is dominating sucking most of the oxygen out of the room, If AWS despite their market dominance is your and gcp are still tremendous players in their own, So there's why I say that only to say the market is very big and it's ripe or people to come in and try to see what they can do and more attention on the space is always a good thing. What I would say is, we've historically shy away from intentionally avoid the language big is fast, whatever it is, right because it's all qualifiers and highly subjective stuff right and again, + +Cheng Wang: The best represent putting right and so we'll see and the market is big enough for many companies to thrive and it's just about getting product Market fit and getting adoption all this kind of stuff. So yeah, it's good to see the space growing. I would just say do your own research, open systems. all of our information is out there information on a lot of the other companies projects, excuse me. Just don't exist. + +Andrew Gnatyuk: Yep, totally agree. + +Scott Hewitson: awesome + +Cheng Wang: and it's cute for them, too. I would just say, we're pretty serious a just saying. + +Scott Hewitson: There's Andrew have any other questions he'd like to ask. But awesome. + +Andrew Gnatyuk: No, I'm good. + +Scott Hewitson: Again, as always it's a pleasure to have these calls pleasure to hear everyone's thoughts. And yeah, I think we can wrap up and As was since he's here. We'll see all in the metaverse. + +00:35:00 + +Andrew Gnatyuk: Yep. you… + +Rodrigo Rochin: See you. + +Andrew Gnatyuk: Bye. Thank. + +Scott Hewitson: Take care. wedding + +Robert Del Rey: You guys thanks. + +Andrew Gnatyuk: I guess. + +Robert Del Rey: Scott & Cheng. + +Cheng Wang: There on thank you. just + +Robert Del Rey: All right. + +Scott Hewitson: Have a good one. + +Meeting ended after 00:35:16 👋