Lukas Seel: 0:00

big welcome to everybody listening. This is helios horizons, episode seven, with mark tilleman, the director and contributor at Pyth data association. And, Pyth, of course, if you don't know, it is a blockchain oracle that specializes in prize feeds. What all that means, we'll talk about with Mark himself. Welcome to the show, Mark.

Marc Tillement: 0:29

GMGM, I almost struck myself, couldn't unmute. No, thanks a lot for having me Very excited to see what we're going to talk about. And yeah, just thanks a lot to Helios and I hope I can. How do you say clear the horizon for the attendance?

Lukas Seel: 0:48

Yes, what I want to start with is probably a little bit about yourself and your background, because it's always interesting to see how people ended up in this space. Of course, you're in a very particular niche in the Web3 blockchain space, so curious about your journey to get there. So maybe give us a little bit of a background of how you first discovered Web3, what you did before and then what led you to Pyth in the very end there.

Marc Tillement: 1:19

Yeah, of course, I think the first time I must have heard about it was, I mean, 2017 top. I must have had a couple of friends that either lost a bunch or made a bunch, or made then lost a bunch.

Marc Tillement: 1:34

But I didn't 2017, so I was 23 years old. I wasn't, no, no, 20 years old, so I wasn't much into it. I was studying in England but I got into crypto in 2019 when I was living in Hong Kong. Out of luck, I changed apartments and the guy in the other room it was an OGOG was one of the early kind of BitMEX employee and we bonded a lot and for maybe like six months, we very much just talk crypto as in the actual, let's say, tech or motor behind it and it was super cool. Like this whole like cutting the middleman out, removing like false trust towards rather untrustworthy people or agency very rang with me and, for like months for yeah, I was just reading about defi, just docs of applications etc. Covid crash happened I was like this sounds, feels like a good time to put a bit of money. And as soon as I started putting money into it, I just like whenever I wasn't doing my previous job my normie job, let's call it I was just looking at crypto stuff. So I worked for three years in Hong Kong in a radiological company, if you want all the deets, and early on. So it was early 21, One of the few people that was working in crypto Hong Kong, it's such a crypto hub.

Marc Tillement: 3:14

Back then heard about Pyth getting launched. Early days connected me with the initial let's call them founders and had a couple calls and he was like, okay, let's join the ship. We need some kind of crypto-facing guy and I guess it's been up from there. So for Pyth, I started mostly on the community side. So the whole Discord, the whole Twitter, the whole more marketing as the product went live. The team was very tech-oriented. I also had to start doing a bit more of the BD and now that the team has grown like fast forward three years, I spend most of my time doing some DeFi BD. So either convincing any type of DeFi app to integrate with or existing user keep them happy, keep improving the product. So we have great apps that use a great Oracle and more recently, with the whole token going live, the forum and the governance going live, I'm spending a bit more time on this whole aspect, kind of representing the Pyh data association within the Pyth network.

Lukas Seel: 4:27

Yeah, super interesting I think, a bunch of these things we'll hopefully get to touch on. I just want to ask when you first heard about this thing the crypto thing, let's say, the blockchain thing you mentioned that that sounded interesting to you you were very young, I guess also and then you kind of became this, I don't know community manager before even you know, getting into it for real. Was there ever like a hesitation of the technology? Did you hear anything that maybe dissuaded you in the beginning to get in, or were you kind of immediately taken by this ideal and vision of the space and Web3?

Marc Tillement: 5:07

So I think I started loving this idea. I mean, when I started loving the idea and I was anyway not knowledgeable enough back then to fully understand is this blockchain type Bitcoin power, blockchain actually the solution for this or Ethereum or what not? So it was very much for me like a do you say, means to a goal. So very much the goal was, I mean, we I'm not a fan of using banks, et cetera and it very much felt like oh, another thing actually was more on the inflation part of things. Like it's something that I also grew into. I used not to be the most financial guy, but getting exposed a bit to especially Bitcoin, I started to understand inflation, that it's kind of the hidden tax. It was kind of a mesh of this technology and it was just a pure belief because I was too dumb to actually verify if it was working properly or if it was viable long term. It was like this seems like a great tool to achieve things I'd like to happen.

Marc Tillement: 6:19

And the other thing that kind of convinced me is all this while having my previous job like every night. I mean my housemates were working full time in crypto and on weekends I mean I was just hanging out around crypto people and what shocked me positively was how committed they were and how, let's say, relatively young everyone was and coming from like the previous companies listed in the French, like the French NIC 3,000 employees. Like you go on holiday for one or two weeks, you come back and it feels like as if you like nothing happened and I mean everything kept running without you. So it was kind of a combination of growing tired of working in a big company, plus, like I was spending most of my free time just reading stuff. So it was like just felt natural, kind of trying to pursue this full time.

Marc Tillement: 7:21

When I first told this to my parents, they were like yes, Mark, stop it, you're crazy. But no, so I guess it was half based on the people pursue this full-time. When I first told this to my parents, they were like yes, mark, stop it, you're crazy. But no, so I guess it was half based on the people I think I was very lucky into how I discovered crypto, like bunch of very knowledgeable people, committed people and very nice, and all this kind of pushed me on the edge of all all right, let's just do this full-time.

Lukas Seel: 7:46

Yeah, I can identify very much, even with the parents part, of course. An interesting thing and really interesting, the different ways that people get involved with. I want to switch to kind of the main discussion and talk more about Pyth and oracles and DeFi in particular. What would you say? You mentioned the general ethos of decentralizing finance and the hidden tax and all of these things. What attracted you to you know dive deeper into this DeFi ecosystem and all the ways that that works? In particular, was there anything that attracted you to Pyth before you started working there and and you learned about those things? Or how did you get deeper into the DeFi ecosystem?

Marc Tillement: 8:36

I think it was no. So when I first started I wasn't that like, oh, this oracle, like idea, is super like, and even today, when you see like no, except your developer that in caring about your oracle arguably no one cares that much, I mean, as long as it works well, people don't care so very much. What shocked me, I think, and hooked me on, was this ability to have and I started kind of degening early on on on solana, it was, I mean, one month post their launch, during it was during Covid crash or around this and the months after like a few so and playing around being able to move money that fast in between friends, do trades like so everything without kind of an account, without asking permission, like I I don't have a, well, I'm not, I'm not American, so I can't I mean I cannot choose robin hood, but I would, even today I'm not interested at all in getting such kind of web two traditional like financial product, but I don't know why, just maybe I think, yeah, the, the permissionlessness was very much something I was, and I think today, and we all like, Pyth is built for user, for builders, as a fully permissionless oracle, and it's one thing I feel is very, very like addictive when you're like, okay, I go on the website, you just connect your wallet and whether you want to long or short whatever coin 100x you can do it. If you just want to supply your stablecoin for safe parenthesis, safe yield, you can. So it was very much the ability to kind of manage your finances however you see fit. I think this is what hooked me. I did some bad trades. Maybe I should have a I don't know how they call it in the banking world Maybe I should have an advisor. I think. Just the ability of doing whatever you want without restriction, whenever you want.

Marc Tillement: 10:43

I was in Hong Kong. I had a french bank account. I mean I used a fair bit transfer wise and other stuff. You're like it's. It's not possible, it takes four days, you pay 50. If back then I had stable coin and some kind of credit card that can on ram and off ramp, I would have saved a ton of money and ton of time. So I think it's very much.

Marc Tillement: 11:06

When you actually try a fair bit of things, you'll see that you have still plenty of pitfalls on many DeFi projects and primitive, but the potential is so big compared to what we have today in the traditional world, I felt, yeah, getting astonished by this is what convinced me, like defi is what I love, and, for example, I mean I have a few nfts but I kind of understand the value behind the tech of nfts who, like, compared to deiy and defy would be like it's big, and I'm not gonna say all of defi is perfect, but well, yet I think to find the killer use case for nfts has its use case so pretty much sticked around this and I think, yeah, as I used more and more, I mean I got to learn chaining back Chainlink back day when I started. I mean, my first prep trade were on synthetics v1. You could trade tesla on it back in the day. It was the worst ux experience ever, but so kind of discovered this infrastructure part and um, no, very much like the when hearing about path, like it could have been another type of infra protocol.

Marc Tillement: 12:22

But the fact, like one angle we take very differently with other Oracle network is like the actual premise of data. We don't think that, like the best data is publicly available, like financial data is even like it. It's eight billion dollar market and when I first heard about pyth it was kind of like today, this market data is monopoly of a few big US companies, whether it's centralized stock exchange or bloomberg or whatnot. So it felt very much like, and back to why I initially liked crypto is kind of killing monopoly, killing the middleman that can extract as much value as he wants. And I mean when I got this idea of like Pyth and the TLDR I tell my parents is Spotify, but for financial data you have data owner, data creator come contribute and if you have users that use your data on chain, you'll get rewarded. No cut goes to Pyth or no kind of overpower we can exert.

Lukas Seel: 13:48

super interesting, funny, funny enough, I going you how you explain uh oracles to your mother, and then you, uh, you already provided it.

Marc Tillement: 13:58

Maybe so and maybe we should, I mean, we should not call Pyth an Oracle, because the first Oracle out there was, let's say, chainlink and a few other bands Waziri, supra, et cetera. The one core difference and maybe I'm going to ask another later question, but I think one core difference is it's not generalized infrastructure. So Pyth is Spotify, let's say, and the other Oracle. Or maybe you take the Airbnb Same Airbnb, I mean without the $100 cleaning fee, I guess, but it's Airbnb. Pyth is Airbnb, other oracles. It's more generalized infrastructure, it's more like booking, it's more like aggregator of other people's data. Pyth is all about proprietary data.

Lukas Seel: 14:48

Yeah, I mean you already mentioned Chainlink. There's also Switchboard and other things that were on Solana, I believe. What sets you apart as Pyth? What of differentiating yourselves from other players in the market? Is there a different philosophy? How would you sort of define that mission of Pyth?

Marc Tillement: 15:22

Yeah. So mission overall would be. So, technically, Pyth network, the network has kind of two customer types the data publisher, the supply side, and the user side, the demand side. So vision is to and it's kind of the chicken and egg but the whole idea of attract all those who own like create and private data, like to like contribute them to the oracle and doing this it's almost like a distribution channel whereby all these trading firm exchanges sit on so much value being this data. What if they could? And overall, in crypto, no one sells its data. What if you'd sell your data? Like US exchange 30% of no one sells its data. What if you'd sell your data? US exchange, 30% of their revenue is selling data. What if you tell Binance oh, by next year I can increase your revenue by 20%. You just have to gate your API and sell it or whatever. So it's Grand Vision is supply side, like have so much user, like demand side side of we want the highest quality data possible and we're willing to pay for it, and so it's kind of we need to build that demand side to show publishers this oh, actually there is lots of revenue to be made here and so what sets it apart?

Marc Tillement: 16:51

So supply side, specifically, the idea is so data owners, so it's exchanges, crypto exchanges. It can be centralized, decentralized, it can be US stock exchange or trading firms. So today we have 100 data publishers on the network. We have so, mentioned US stocks. We have three out of 13 US accredited exchange or data publisher to Pyth. Cbo, who makes three billion a year yeah, two, three billion a year selling data, is actually now contributing their stock data to path. Why, for them, it's kind of a new distribution channel, Binance kind of the same, Kucoin, all the big centralized, the decentralized exchange. Also, like they already do fees on their trading but they own technically, like I mean it's kind of public data because it's on the blockchain, but they kind of own or can provide their pool information. Trading firms like, if I'm a trading firm, I pay every year 2, 3 million in data fees to the exchange I trade on. I can actually resell and make as much distributing via Pyth. So this idea also like getting those, let's say, professional participants that know what the price of an asset is to contribute to the Oracle helps also growing this demand side.

Marc Tillement: 18:14

So demand side, why are people using Pyth? What's that good, based on all these data publishers. We can have like sub-second latency on price feeds. So today we have 500 price feeds on the Pyth network. All of them update two to three times per second. In this we have about 350 crypto price feeds.

Marc Tillement: 18:37

We have 70-ish US stock equity prices, we have FX commodities and so this all the oracles arguably cannot get access to such deep data, and especially on the US equity side. And if other Oracle kind of offer equity data, I mean I don't want to say it's not very legal, but I would wonder where they find the data. Because, like stock data for the, like US stock data is not only expensive but it's one of the most regulated markets in the world. With Pyth getting all the biggest traders in the world on board, we can get access to this data and so from the user I'll get pretty much it's almost like long-term Bloomberg. We want to have 10,000 price feeds, whether you want to trade Korean stocks, the weirdest crypto out there, US yield curve, et cetera. We want Pyth able to provide these data

Marc Tillement: 19:38

Initially it's targeted to DeFi apps, but we do see off-chain utilization and I'll just add one thing on the demand side. So one thing how we built Pyth and like the whole design is very, I think in line with defi is Pyth work as an on-demand oracle, meaning that all these 500 price feeds can get access permissionlessly by anyone on more than 60 blockchains. Let's say I'm a perp protocol on optimism. I can fetch off-chain the piece price the current price of Tesla, of Bitcoin, of ETH, of whatnot send it on optimism, the Pyth smart contract will verify it and then I can fulfill my user trade at this price, and this without contacting anyone at PEF no legal contract, no business deal. It's fully permissionless and on demand and we can do this because the user and by user I mean the application has to do the transaction to send the price and change. So overall it's a very cost-efficient Oracle design and we can actually have a revenue mechanism embedded. So I went a bit deep here, but I'll stop now and I'll let you want to rebound something.

Lukas Seel: 21:04

Yeah, I'll actually go a little bit deeper, because you mentioned stocks, real-world assets coming on chain and all of that, I think, is probably where the future is headed, especially with, in Europe at least, legislation becoming a bit clearer, with Mica coming out soon, how do you see that fitting into what Pyth is doing right now? And also, how are you guys preparing for, for those changes in, maybe, legislation, but also just planning ahead for things happening in the US, where this kind of stuff is much more complicated still?

Marc Tillement: 21:41

so we have, let's say about, so we can have already preparing for it. Because I mean I agree with you, at some it's going to happen, I don't know when, I don't know where but until then we can gradually work towards by the day it happens, so that the Pyth network is the best place to fulfill that demand. So that's why, for example, we have, let's say, 100-ish price feeds that are very stratified, 90% of them at least. When I say usage, I'll refer to on-chain usage. So whether it's perp et cetera, there's pretty much zero usage on this. But if tomorrow you can safely launch a perp exchange on any chain, anyone tomorrow can launch pretty much easy on-chain with Pyth as the oracle.

Marc Tillement: 22:33

For example, we have all the Bitcoin ETFs I think there are 13 or 14 of them. All of them have their price fees on Pyth. So if at some point Vaneck, blackrock or any of the 11 other guys want to launch, like one thing I mean I think that was the hot take or the hot prediction I gave like a month ago on another space was what would be kind of the most bullish thing or the most other thing I'd like to see is blackrock launching kind of its own l2. I mean it will be permissioned for legal reason, but with all these etfs like, make the trading happen on chain, like crypto people, I mean it would be amazing. So I think it's very much like no one knows exactly how it's going to happen, but, from our perspective, as long as we have the data ready, we'll be able to fulfill the demand. So it's gradual because, as we've seen, like we have stocks, price feed from day one. We had those, but till date, and maybe here and there we see usage, but it's going to be by unknown teams because of the current regulations.

Marc Tillement: 23:47

And, sadly, while in Europe and I was at Paris Blockchain Week on I don't know if I was the best fit for this panel, but I was on the regulation panel and with like a couple of peeps that work on Mica and things and sadly, like the USA, like the, the US just whatever they said happens. Like if, even if Europe says you're fine, if somehow the US wants to find you bugs, they'll find your bugs. I mean it. Yeah, we're all very dependent of what the US will do, but after, I mean, you always have very DeFi, degen guys launch purpose perps on Tesla with a hundred leverage. But I mean, if the people want it, that's actually, I think, maybe more than all these bonds tokenized on chain. The real world asset parts where we could get massive adoption is through this stock trading, etc.

Lukas Seel: 24:59

Yeah, interesting. I saw an interview with you at Paris Blockchain Week where you were talking about USDC, and I think this is also an interesting topic to touch on, at least briefly. I gathered from that conversation that you weren't the biggest fan of that asset in particular. Let's talk a little bit about bringing assets on chain, though that can help retail. Your mom and my mom understand what they're doing, because it's not a speculative asset. What they're doing because it's not a speculative asset, and then perhaps also your vision of having a currency like that, a stable coin that isn't usdc, and what you might not like about it or the pros and cons of it, perhaps

Marc Tillement: 25:53

Yeah I think, bottom line argument is, I mean most of my stables are actually in USDC. Maybe I should no, actually not, I use something decentralized nowadays, but it's as everything can have its use case, then am I the target of this use case? So USDC, for example, I mean they haven't done it much compared to Tether, but tomorrow, for whatever reason, no, and actually Circle, they're going to IPO in the US. If the USA government wants to blacklist your address, you can lose everything. So it's like I understand. So who's going to use USDC? So I mean I think it's going to be very successful. Like we might see trillions of USDC at some point in circulation. But I think it's going to be very successful. We might see trillions of USDC at some point in circulation. But I think it's going to almost be like USDC.

Marc Tillement: 26:39

I talked to Bitcoin ETF issuers to launch their own chain to trade. Good chance they'll be settled in USDC. So there is a use case here and a reason Okay, it's regulated, licensed in the us, all good, but as a, non-american, as a non-us institution, the like. My argument is more like why would I take the kind of censorship risk. After, where you make a good point is, for example, my parents. I'm not going to onboard them. I mean, if I try to onboard them through Maker, go mint with Maker or with this, I'm not sure they would do it With USDC. There's a better chance. They just send the fiat transfer, get some USDC on it. So I just hope that the coming generation I mean our generation I don't know if it's bad to say the past for my parents, but for the coming one I hope that, and I feel most people care more about these things that arguably the government's not out there always looking out for your best. So everything will have its use case. I mean I'm in DeFi.

Marc Tillement: 28:01

It would be kind of weird if I were heavily relying on a centralized stablecoin. But for example, I mean we work a lot with them on Pyth, it's called Engel. But for example, I mean we work a lot with them on PEF, it's called Engel. They do Euro and how do you call this Euro and USD-backed stablecoin? And actually here you still have some centralization aspect because some part of it is backed by T-bills at one of their brokers, but no such thing as freezing et cetera. So I guess it's always. I try to not be black and white, but it's always kind of fine. If there is a use case, I'm fine with it. If there is actually no use case and we should be trying to get rid of it, usdc they are get. Yeah, it will be the settlement whenever the NIC is on chain good chance it will be settled in USDC, for example.

Lukas Seel: 28:59

Yeah, yeah, interesting. I think it's very interesting to see. I always think about my parents, right, and like what their interaction with blockchain and Web3 is like and what they would actually need to come on board, and this is also something I wanted to touch on later, but we can just kind of fast track that Now. How do you see this sort of onboarding or mass adoption happening? Do you think our parents' generation is just not going to be part of that? Do you think it's going to be super abstracted away and they're on it but they don't know that they're on it and it doesn't matter anymore?

Marc Tillement: 29:36

And also, perhaps, how does Pyth fit into that vision of that process. So will my parents ever own coins? Whether they know it or not, I mean practically knowing it, I doubt they're ever on a ramp. I mean, if we have to improve one thing to bring both worlds, it's the whole on ramp off ramping issue. One thing I very much like, and I don't know why I haven't even yet ordered it, but it's the whole. Gnosis released their card and pretty much you can and it's not everywhere in the world, but you can already pretty much have a bunch of well, they have DAI, xdai or USDC analysis. You top up your card and you can pay at the grocery shop with it.

Marc Tillement: 30:20

So I think if we go this way, I mean there could be a way. I mean, this is one way where we can bring both worlds, like the people not knowing and just using that. But I think it's too much of a leap and we're also so far away from actual, true abstraction that I mean, is there a lot of crypto application you've used where you're like oh, this is almost a traditional website UX? Very few, like less than five maybe come to mind, to mind. So I think my bet would be my parents or like the 4, second generation or the doom generation, of never going on chain. If they go, it would be purely from investment perspective where? And back to the inflation mentioned earlier.

Marc Tillement: 31:19

So I'd go this way and how could piff help? So here it's more like we being infra. It's almost all right, it's not like a current abstraction, but it's delivering. If you go trade on robin hood, whatnot, um, it's like speed of light. Um, whole idea of piff is if all our price feeds are speed of light, also, the DeFi apps that use those will feel like using Robinhood or stuff like this. So I don't think we have a rule in there. But it's kind of like abstracted away, as in the sense it's like 10 layers down, you have the only lower layer than us would be the blockchain, then you have the oracle and after I guess it's all about the actual teams, like protocols, shipping and caring more about ux and user flow yeah, super interesting.

Lukas Seel: 32:21

I also maybe this is a good segway into this as well kind of building this ecosystem out a little bit more and having perhaps more user-facing applications that feel like Web 2 or like Web 2.5. You guys have started a few things on your own community grants, research grants, development grants let's talk about that a little bit. What kind of projects do you want to see apply for that and how does that fit into building out this DeFi ecosystem?

Marc Tillement: 32:59

Yeah. So I think on the developer grant side I'll try to take each vertical one by one. So and after it will like I'll say my kind of personal preference on what I'd like to see. I might forget some. But for example, developer grant, it's not so much a we pay out for you to integrate path. That's not the goal, because I mean if you do it once, you have to do it twice and ad vitam aeternam. On the developer side, I am more interested in seeing contribution from random and it can be actually already Pyth users that improve the developer UX when integrating the Pyth oracle.

Marc Tillement: 33:48

Pyth overall it's not that complicated, to be fair, but it's kind of different for developers that have experienced other oracles, because it's built the exact opposite way All the other oracles. Pretty much you don't have to do, or the only thing you have to do is sit on your hands and let the oracle do what it's telling you it'll do . Pyth, and here I refer to the push model where, let's say, chainlink will tell you I'll update the ETH price on Ethereum mainnet every one hour or one percent price deviation. If you're a developer, you have nothing to do but point your smart contract to this account where they'll update the price. Pyth should have to request the price of chain. Send it on chain so you can have this extra step. It brings tons of benefit. But people in Chainlink have been here for five, six years now. There are so many integrations and developers have. Their first discovery with oracles for developers was Chainlink. So we do have a lot of education side of things, but this would be more on the research side.

Marc Tillement: 35:03

But back to developer. It's like a better tooling, like better interface. Make it as easy as possible to integrate Pyth In the same space actually I was referring to earlier. It was with the dev rel of ICP and he said one thing he was totally right is if you're a developer, you have your console, you try to integrate path. You shouldn't be like if you have an issue, you shouldn't have to exit your console to find the answer or troubleshoot that issue. Like don't have to go in the docs, don't have to go in discord, why? And just this developer UX of of, however complicated the thing could be, a day one learner could troubleshoot himself into a working product.

Marc Tillement: 35:55

This is almost my dream goal of the developer bounties Everyone kind of. I mean. Am I going to find a good analogy. You're the first one, it's hard, but I mean you have a good heart and you'll get some piece token. Try to make it easier for the next one and keep this flywheel going and the easier it becomes, the more people will try help it and improve it. So on the developer side, I am personally very excited and looking forward to this aspect.

Marc Tillement: 36:34

Another one and I mentioned we don't want to pay people to integrate. There could be some developer grants for actual new products, but this would require almost like a new primitive, like, for example, uniswap kind of invented the AMM. I'm not asking you to reinvent AMMs, but if you reinvent a completely new primitive within DeFi that you space, we'd love to support that. So it's either moonshot on, this guy's a genius, he's shipping what's going to revolutionize DeFi, or it's making path the easiest Oracle ever to integrate, the best UX, et cetera. On the community side, our research. So research, and I mean it can be so wide, but for example it could be as, like I mentioned, push against pull Oracle. So little people know about it. Have people write blog posts, do research on like why maybe is it better, what's the downsides of it? So just kind of it starts from general knowledge. But the other thing I'm very much more excited on the research side is more about the stats usage of the Pyth Oracle. So we're live on 60 blockchains today. I think there is in between 3 to 6 million price requests every day by apps on those 60 chains. Go would be like create the nicest Dune flipside dashboard ever, so that all this usage can be visualized in a very, let's say, easy way, digestible way. And the second effect of this first I mean it's good monitoring practices. But the second effect of this is linked to how pith is designed. So I mentioned full Oracle.

Marc Tillement: 38:41

Pyth can monetize. Whenever someone does a price request, actually the requester has to pay a fee to the Oracle. So in this case, so pretty much you can say that we have every day 5 million revenue generating events. Today the network is in growth phase and the fee charge is one way for all the EVM chain, one lamp port for Solana, so pretty much dust. It's like a cent exponent minus 50. So we'd need like a trillion update to make one ETH today.

Marc Tillement: 39:17

And it was kind of the kind of reasoning at the beginning. It's growth, growth, growth. Kind of the kind of reasoning at the beginning. It's growth, growth, growth and so with these, like research into, for example, okay, on optimism, there is about 10 000 requests per day, x amount for this price feed, maybe 30 is actually used for by f, for fx or commodities, two-thirds by crypto and from there decide the pricing, because the pricing, the price feeds, is going to be decided by the Pyth DAO. And so to make intelligent decision, we need to have good data and not we.

Marc Tillement: 39:56

When I say we, it's everyone, all the stakers. We have like over 100K stakers, this token in governance, so our idea is for everyone to be able to understand. Okay, so if Pyth starts charging 10 cents tomorrow on Optimism, maybe it will lose all of its customers because it's too expensive. So, finding that right balance of what should we charge to the applications, because in the end, the application is going to charge it back to the user in some various ways. But on the research part, this thing in between the research usage and the economic parts, you know the flexibility of demand I think that would be very nice. And after, more big brain, I mean this is already, this is more economic, but even bigger brain stuff. I mean, if you want to ship some tokenomics we've mentioned this like various types we could imagine within the network could start using the Pyth token in new ways.

Marc Tillement: 40:59

So in research it's either kind of general education or very much like help us, help the community kind of know the best of what's happening within Pyth and how we could, let's say, retake the example of charging in the future, because that's super important and community is very much to. It's an Oracle B2B business. Even many developers aren't that excited when talking Oracle, so non-developers like are the last one to care usually. But it's all about. I mean, we've had three years pre-token so we found various fun ways to keep people active. And one thing I like with Pyth it's kind of a window towards DeFi. Whatever chain you like and are active on, we'll work with people there. We'll get you exposed to plenty of new things and so very much communities to get that flywheel effect of like don't come to farm prospective airdrop, it happened already but kind of nurture this defi lover community yeah, I really like that and I think it's it's interesting.

Lukas Seel: 42:18

This is, this is something perhaps, the the last point before we open it up to community questions as well. Quick reminder this is episode seven of Helios Horizons. We're here with Marc from Pyth, and one thing that I noticed when looking at the Pyth website is the team and I think this ties back to this sort of Web 2.0 versus Web 3.0 spirit. It ties back to the governance aspect as well is you guys all have the same title and it's contributor. I'm super curious about that choice, that conscious choice, and then also how you see these different use cases for Pyth, the token, now that you have a government instrument, a governance instrument.

Marc Tillement: 43:08

So I mean that's one thing. We started introducing title. Actually Not many, but no, for the longest time we had zero, because it was either, I mean you kind of have the lead engineer and a bunch of engineers below and same on the business side one lead and after like eight people and after it's just very much whether you're more focused on marketing or more the bd. And arguably, if I have, if let's say I have to do some marketing, I'll do some marketing. If the marketing has to do bd, he'll do bd. I think this was also kind of a necessity because the team has always been super small. I mean, I think today, if you include both, the association, we are like about 12 people, and the labs entity that is contributing to the network is, I don't know, 12, 25-ish, 30. So let's say total 40.

Marc Tillement: 44:05

And many DeFi apps not many, but a few DeFi apps from last cycle have actually as much as people as, let's say, here, contributing full-time. And it's pretty crazy because, when you think about it, we're on 60 chains, 500 price feeds, we have 100 publishers to also manage, and I think we've ended up being kind of lucky in always finding people that just work 6.5 days per week, not because it's forced, but I think it's so small. Everyone has so much responsibility that in the end, you don't care, like there is no politics you don't care about. Oh, what's my title, a s we grow more and more like, by nature, we'll need it, but I think this is one thing also I've learned from the guy that got me into crypto A couple of them have launched their own projects and he was founder engineering team, non-engineering team and then just do your stuff. Don't politicize everything, just don't break, don't lose people's time and just do stuff. And this is one thing I'll keep pushing on this end is very much to stay as lean as possible, because again, I mean, at some point we might again have a bear market or whatnot, and so I think it's very much like mentality of like staying sharp, staying a startup.

Marc Tillement: 45:39

Overall it's not always possible. I mean mean, if it's, if you're succeeding enough, at some point you might become a middle-sized enterprise. Let's say, um, the like goal is not to have 25, 25 people, hr department or something like this. Um, like. If I see internally stuff like I saw in my previous company, I'd be like, yeah, I didn't join for this and and after.

Marc Tillement: 46:10

Also, it keeps everyone on its toes and even on the product design engineering side like you don't have endless process meetings, whatnot. It's very much down to the point. And what's actually very fun is the non-tech people are kind of always challenging the tech people, so sometimes you open threads here and there and you're like almost a CM. Yeah, that sounds pretty dumb to another engineer and this is very much what we want because it's very kind of a two-sided market, even in market. Or we look at two different things internally engineer most of the time look at the code and we get the either the community or the user feedback. So it's it's not that we don't care about the same thing, it's we don't care about directly the same thing. Ultimately it's the same, but it's always good to have this, none, like not that structured stuff.

Marc Tillement: 47:14

I kind of forgot the second part to that question how we are evolving in the future, or was it another angle it was. I know it's governance. Okay, so governance, and actually this is super cool because you won't find maybe one other record that has actually a governance working so arguably when you, whenever you see, oh, this decentralized oracle, it's likely just a five out of something of 35 multisigs that you don't know the signers for. With the token, I mean Pyth was like this until token launch. With token launch it was actually possible to give back all the power to the token staker. So initially it launched very much with just the staking program. You'd stake and you'd vote on program updates. So the big Oracle, if I don't know, we managed to improve latency by 100 milliseconds. It's like a whole DAO vote because changing Oracle contract is likely the most risky thing. So it was also very much from a security perspective of you never know what can happen internally.

Marc Tillement: 48:21

Now token holder with the airdrop, sufficiently decentralized to reach something that makes sense. So overall we had this back end of last year. We had achieved, let's say, on the pure technical perspective, what we wanted, decentralized, up to the token pota. But then, like the social layer of the governance wasn't existing. Earlier this month we actually launched Pyth forum, no forum pyth network, which is kind of the discourse where actually now the pyth DAO can guide the like forward-looking development.

Marc Tillement: 49:01

and it ranges from and actually that's super cool to see because we already have two proposals or two ideas that say they haven't reached a formal proposal stage. They aren't so much, I mean they are not touching at all the Oracle code, the price feed, the product, but they are more community-oriented. And the two proposals they are pretty much the same. It's the two competitors doing their bid. It's to offer the domain name service for the Pyth community. So you have the eth sol on Solana and you can actually have pyth on Solana, and it's so those two company, that kind of came to the forum and now do their bidding.

Marc Tillement: 49:54

Oh, we offer this product to pyth. Like whole community chimes in, and at some point next week I guess we'll pick where we're going. But having this um like community driven is, I think, very cool and it, I mean, I kind of joined DeFi for this, so, you'll see me in the how do they say in the governance arena, like a lot more. And after we'll go on. And I mentioned the fees, so I mentioned the fees like we need to know the usage, we need to know the elasticity of the pricing we can have and after it's going to be in the forum that we decide this. I don't think there is any other oracle out there that, like it's the token holders that choose how much they're going to charge, and it just makes sense. It's like how it should work, I feel.

Lukas Seel: 50:46

Yeah, super interesting. I don't see any questions in the comments, but I know that John, who's hiding behind the official account right now, has a question and also an invitation. Again, if anybody else wants to step up and ask a question of Mark, who's giving us his time today, please request the mic. John, go ahead.

JB Carthy: 51:07

Yeah, Marc, really enjoying all the insights and I suppose, just as you were chatting, I was wondering, as a team that are uniquely positioned. You know you work with so many chains, work with so many different types of DeFi primitives what is maybe the one aspect of DeFi, or a couple of aspects of DeFi, that maybe excite you most going forward over the next few months and into the next few years?

Marc Tillement: 51:31

So I think the one thing, oh yeah, actually I have a good one I'm excited about those who actually try very new things or things completely differently. So, for example, whether it's Monad or TVM trying to rewrite the EVM code actually to achieve like 10k TPS and fast finality, but it's not like just a gift fork, all right, and we strap some token on it and you get a new chain. It's actually let's rewrite the code. The chain runs from scratch and they managed so far to create a great community based on this. But it's things that excite me are these type of things. Maybe it will fail. I hope they'll succeed. Same with, for example, yesterday I was on a space with Eclipse, so it's a Solana virtual machine.

Marc Tillement: 52:32

Layer 2 on Ethereum Like this is pretty cool and I wouldn't be shocked if they have a great traction compared to EVML tools, because I mean, the more we go, oh, and I think maybe the counter, the other side of the coin. I don't like about people just redoing stuff a bit the same way as all these Bitcoin data tools All the ones that have traction today are just EVMs. I don't think personally EVM is the end state of the coding world for blockchain. It showed it's very safe, secure. It's been around for seven years but it's. I mean, there isn't any fast EVM. What not today? So Monad, rebuilding it makes sense.

Marc Tillement: 53:22

Those Bitcoin layer tools, I mean from an economic perspective. I am shocked actually I didn't think about it before because it's actually even better than Ethereum layer tools, because you can actually tap into a trillion dollar asset copy paste curve code, have a code, whatever successful EVM project you like, copy the code, deploy on a Bitcoin layer to EVM and boom. You can then get half a trillion of asset into your app. And while it can be successful for a few years, this is the type of stuff that I'm not very excited about. It's just all right. We just strapped some EVM on Bitcoin and that's it. So people that try new things, and now I guess after. One thing I'd love to see is maybe a big and actual enterprise, and ideally on the financial side, launching their own chain.

JB Carthy: 54:30

Yeah, yeah, for sure. The EVM. At what stage do you see, maybe, and do you see the future going like in the next few months and in the near future, medium term future, and do you see anything moving towards? Or, let's say, will it continue to be scalability through Ethereum layer twos or do you see something moving away for the focus being more towards scalable layer ones, and what do you think will initiate that shift, if something will initiate that shift,

Marc Tillement: 55:16

I mean they have zero market share but I wouldn't be shocked if they become have majority of the market share. So SVM, it can be, move VM. Yeah, so I think it's in, by what year are we 24? Yeah, I wouldn't.

Marc Tillement: 55:41

Within five years, let's say, non-evm layer 2s on Ethereum will be, let's say, less used than new VM on Ethereum. Ethereum layer 1, and I mean it's the whole kind of also network effect of you have a half trillion asset or don't know how, maybe less. Um, that is technically deflationary when things go well, so you can have the reach of stage where it's almost too big to fail. So I think you'll still have bitcoin ethereum, longterm I wouldn't be shocked if simple Ethereum fork die. This I wouldn't be shocked. And I mean there isn't if you check DeFi Llama, there isn't like a fork of Ethereum. I mean Fantom they completely rebuilt. They pretty much died. On the product side, they have infinite treasury, so they managed to keep building and now they're actually completely rebranding their VM. So I think, yeah, you live long enough until you're like I need to differentiate myself.

Marc Tillement: 56:52

So usually, whether it takes five to ten, I think it will take five to ten years After I mean, and I guess we're in a good state, kind of already now. I think it will take five to 10 years and I guess we're in a good state already now because we have the move chain where they're up to a century and each one has its focus. One is more DeFi, one is more gaming. Solana, it feels like it's trying to do everything at the same time. So, yeah, so I'd say five to ten years. You have a consolidation of just Ethereum as an EVM layer, one. The rest would be like Monad, let's say, which is technically EVM, but EVM under Stellar Raids, let's say, or a completely rebranded one and L2s and after ZK. I mean ZK is going to be a big thing, but it feels like we've been talking ZK for the past five years and very few ZK things are actually used in production and if they are already used, it's not that scale. So, and this again, it's another five, ten years.

Lukas Seel: 58:01

Thank you, it's nice to have the Oracle guide be a little Oracle for us in the future.

Marc Tillement: 58:06

I'm bad at prediction because I do love doing predictions, but I lose more money than I make. Yeah.

Lukas Seel: 58:16

I don't see any more questions from the audience or requests, so we'll start wrapping it up. I'll have one last question for you. Today is a special day, of course. As we all know, it's May 22nd, so it's Bitcoin pizza day. Do you have any special plans?

Marc Tillement: 58:33

I had a pizza for lunch. I didn't pay $700 million like Laszlo did. When was it? 12, 13? No, what year was it again, Lukas, do you recall? I think 2013. Okay, so yeah, 11 years later, I paid 11 euros 10 euros for my pizza today, and I actually didn't think about it when I had the pizza.

Marc Tillement: 59:01

But no, I mean it's a great story. I mean, have we heard about this guy since? I'm not sure. But no, I think it's all these one time far in the past, weird story where people just reminisce too and you bond over that. For example, three days ago, on the 19th, and I mean, maybe you've seen it on your timeline, but it was all about and I'm not going to talk about price here, but there was the whole like, uh, Pyth unlock, f or some parts, and actually on May 19th they are like big events historically. It was. What did we find? I think we found that the first, the Star Wars, the Phantom Menace, was released on May 19th. It was plenty of weird kind of what happened on that day, and it's just how you recall stuff forever. So we need more esoteric stories so people can bond over.

Lukas Seel: 1:00:17

Yeah, I agree, such an interesting aspect that this memory kind of hook is something that the storyline, the narrative is, something that humans connect to. Marc, I want to thank you so much for coming on and spending time with us, looking forward to what's to come for Pyth and especially excited about the entire governance part. I think that's a really interesting and cool step. Yeah, thank you so much for being here.

Marc Tillement: 1:00:44

Thanks for having me and I guess I'll see you on the, I don't use much MultiversX, but I'm an Injective, I'm a ninja, so I guess I'll see you in the Helix order book.

Lukas Seel: 1:00:56

Amazing. Looking forward to this. Thank you everyone for tuning in. This was helios horizons, episode 7, with Marc from Pyth. Looking forward to seeing and hearing you all uh next week and uh take care, celebrate bitcoin pizza day, and talk soon. Bye-bye.