This issue's interview will talk about why Pyth chose the oracle track, the current problems with oracles, and the future relationship with competitors.

ZK, RWA, Intent… More and more projects are moving towards these hot directions. Some projects are thoughtful development choices, while many others are just chasing the latest trends. In less mentioned tracks, there are also many problems that need to be solved, such as oracles. On September 12, Google Cloud Oracle reached a cooperation with LayerZero and provided services to all LayerZero applications, bringing attention back to oracles.

At the recently concluded TOKEN 2049, Mike Cahill, CEO of Douro Labs (Pyth Network), shared his thoughts on Pyth's approach to the oracle track and the current issues. After the presentation, we interviewed Mike to discuss why Pyth chose the oracle track, the current problems with oracles, and the future relationship with competitors.

Pyth Network is an oracle designed for DeFi, holding a leading position in the high-throughput DeFi world. It currently ranks second in the number of supported protocols in the oracle industry (second only to ChainLink). Below is the full text of our interview.

Launching Pyth: Firmly Choosing the Oracle Track, Rediscovering the Missing Key Part of DeFi

TechFlow Deep Tide: It's great to have you for the interview. I also listened to your speech just now, which was very attractive and insightful. First, could you please introduce Pyth Network and your role to our readers, and what are your main responsibilities at the company?

Mike Cahill: Of course, I'm Mike, CEO of DouroLabs. Douro Labs is building a series of tools for Pyth and supporting related development work. Pyth is the largest first-hand oracle network, and its mission is to transmit all financial data to the chain, just like all the world's music flows through Spotify. We believe that in the future, all financial data in the world will be transmitted through Pyth.

Pyth belongs to the oracle track. Blockchain applications cannot access real-world data; they can only access the state of the blockchain. The work of bringing external data into the blockchain is done by oracles. Pyth is a permissionless oracle with extremely fast operation speed. Our goal is to become the fastest and most reliable oracle, providing price data covering all types of financial data, cryptocurrencies, stocks, forex, metals, and interest rates.

TechFlow Deep Tide: Thank you for your answer. Another question is, we all know that oracles belong to the Infra category. I'm curious why you chose the oracle track, is there any special reason?

Mike Cahill: This goes back to the very beginning. I was working at a trading company, and we did a lot of project investments, trying to understand the vision of DeFi. The vision that has always excited me is: to build a decentralized trading platform. Just like RobinHood in the United States or Interactive Brokers globally, but a decentralized trading platform would allow everyone to have the ability to access investable assets.

As I continued to research in this field, we realized that one bottleneck was the oracle problem. As I described earlier, you need to have reliable data. But when we carefully studied the solutions at that time, we realized that they were very slow. This is because of several reasons: oracles mainly fetch data from the internet, but the data on the internet is usually delayed. As I mentioned earlier, I worked at a trading company, and in a trading company, you would know that "delay" is very deadly. If your data is slow, you might lose money. We realized that this was a critical missing part in DeFi or infrastructure.

Since then, we started working on Pyth Network. We believe that without Pyth, DeFi would not have the ability to make further breakthroughs. We believe it can enable applications to compete with centralized versions in a very short time. Today, specific types of derivative trading applications have already done this, and there will be more global asset management solutions in the future that can achieve this.

TechFlow Deep Tide: As you said, price data is very important in DeFi, and we need very smooth and fast price data. Pyth also has many innovations in these aspects. Could you list some key features of Pyth and its impact on the oracle track?

Mike Cahill: The main innovations are threefold.

First, Pyth is a first-hand oracle network, with over 90 data publishers in the network. Each of them is a financial institution, either a trading company or an exchange, and they have their own data, and they are incentivized to publish this data directly. This allows Pyth to access different types of data as never before, not just free data or data that you can buy in packages, but more. For example, if we want to obtain some asset data in the United States, this is usually very expensive, but Pyth's data publishers can provide this data. This is the first innovation.

The second innovation is Pythnet, which is a blockchain based on the Solana virtual machine, with the same transparency as Solana. So you can see all the price data created on the chain, and you can also see the specific situation of these 90 different data publishers. This is a very important thing, because today we have 350 symbols (i.e., trading pairs), and this number is increasing every month. In January, there were about 200, and it will continue to grow. Each symbol updated every 400 milliseconds can be passed to over 30 blockchains through Wormhole. This is a huge innovation. We call it a complete oracle. Everything is published on Pythnet, and any of these symbols can be pulled by any of these 30 blockchains, such as Optimism, Solana, Sui, or Aptos. This is a breakthrough innovation.

The third is Confidence Intervals, which no other oracle team has done. The key here is that to accurately represent the market, you need to provide a specific price, such as X, but there is an acceptable price range at any given time. Bitcoin does not have a unified price. Bitcoin has its own price on Binance, and it also has a price on Coinbase, Upbit, and sometimes these prices differ. Therefore, if you want to be a fast-growing oracle company, but your prices are divergent, it will be difficult to build trust. Confidence intervals allow us to develop very quickly because we provide more information. For example, the current price is X, but within an acceptable range, you may find that the trading price is positive or negative y.

These are the three core innovations that allow Pyth to maintain high-speed data transmission while remaining trustworthy.

TechFlow Deep Tide: Thank you. I thought you might also talk about the Push model and Pull model. How do you view the difference in this model? I think this is also a very important innovation.

TechFlow Note: Currently, oracles are divided into two main price feed models:

Push model: In this model, the oracle continuously pushes price updates to each blockchain and pays transaction fees for each update. This has become a fundamental issue limiting the scalability of oracles: the more quantity, the more blockchains, and the more frequent price feed updates, the exponentially increased cost.

Pull model: A demand-based price update model ("pull" model), this innovative design allows users (applications, clearing parties, arbitrageurs) to pull on-chain price data updates only when needed.

Mike Cahill: The Push model has various problems. But let's talk about its benefits first. The benefit is that as an application builder, when you use relevant data, you don't have to consider the operational mechanism behind the oracle, such as many external participants continuously updating data for you at a specified time. This is good, but the problem is that someone has to foot the bill for this.

Therefore, the business model becomes very blurry. So the typical operation of the Push model is that they charge the protocols or users around the protocols. They would say: we hope you pay for next year's data fees, we need to charge you some gas fees. If you do this, we will continue to publish and update this data. But now there is a contradiction between their profit maximization goal and the quality of the price data they provide. If they want to maximize profits, they will reduce the speed of data publication to the lowest level. If gas fees increase, their budget for maintaining these price updates may be insufficient, so this is unsustainable.

With the Pull model, the economic model becomes very transparent, and users can directly pay for the price data they use. In the Pull model, what you do is pull the price onto the chain, but you see all available prices, right? On Pythnet, we have 350 symbols, updated every 400 milliseconds, and increasing the number of trading pairs does not incur additional costs. If we want to maximize the profit potential of the entire oracle, we only need to increase the cost of pulling, and then users can decide whether they are willing to pay for it and whether it is a good deal. This is a fairer and more sustainable model because you can see all price updates, and the quantity will not decrease. So I think the Pull model has clearly demonstrated a sustainable business model.

TechFlow Deep Tide: So the best choice for DeFi projects is the Pull Model, right?

Mike Cahill: Yes, that's right. I think the current challenge with the Pull Model is more integration work and following the market to integrate data into more networks. Synthetix is a good example; it has abstracted from end-users. If you trade on Synthetix, as part of the trading process, there is information called the Pyth price, but the user doesn't know because it's embedded in the backend code. Synthetix developers used to have to do more work, but now they have released more growth potential through Pyth.

TechFlow Deep Tide: The next question is about Douro Labs. We know that Douro Labs is actually a newly established independent organization. Could you explain why this independent organization was created and how it will impact the future development and ecosystem construction of Pyth?

Mike Cahill: This is an important milestone in Pyth's growth. Before the establishment of Douro Labs, contributors from various entities were working for Pyth, and they were not necessarily full-time. Douro Labs provided Pyth with a full-time development team and some truly talented individuals dedicated to Pyth's development work and protocol business expansion. Douro Labs has 19 people, many of whom come from Jump Tradings, and core members also have professional backgrounds from Goldman Sachs, BNP Paribas, AWS, and Chorus One. This is a very talented team, now focused on developing Pyth. Perhaps in the future, our team will focus on things other than Pyth, but for now, Pyth is still our core focus.

TechFlow Deep Tide: Okay, thank you. As mentioned on social media, you will be pushing for token governance deployment. So, do you have any plans for the future of the Pyth Token that you can share with us?

Mike Cahill: The whitepaper mentions that there will be a token-dominated governance mechanism at some point. We are developing the corresponding tools. A specific timeframe has not been announced yet. However, token-dominated governance means that users of the network can decide how they want their network to operate. So this is important, such as fee standards, or whether to charge fees, such as new data publishers or new symbols. If you consider the current projects associated with Pyth, there are actually over 200. This usually requires many different symbols, which can be managed by the Pyth token holder community in the future. For us, a truly decentralized Pyth oracle network owned and operated by the community would be a surprising thing.

Competition: Strengthening the Entire Ecosystem, We Will All Succeed

TechFlow Deep Tide: Okay, let's move on to the second part, about competition. The first question is that there are many oracles in the market, and in terms of the number of supported protocols, Pyth is currently in second place. So how does Pyth view its relationship with other oracles, and from your perspective, in which aspects will Pyth stand out?

Mike Cahill: The three innovations of Pyth that I mentioned earlier I think are very unique, and we are proud of them. I think we are in a low-latency oracle flood that can expand from crypto assets to traditional assets. In today's 350 symbols, 25% are real-world assets. This is a major difference for Pyth, as most other oracles only focus on crypto assets.

I think our relationship with other oracle companies is friendly. We appreciate many things that other oracles do that Pyth does not. Pyth is very focused on obtaining the best financial data on-chain in the fastest time. Pyth does not have random products, such as random number generators, and does not deal with sports and weather-related data. So other oracles need to do these things, and we admire them for being willing to do so.

Some oracles have also developed different products, such as bridges: Chainlink has introduced their CCIP suite using Wormhole as a bridging mechanism, but in the future, it may use a different bridge. I think the bridge field and the oracle field are very different in a vertical direction. We are still in the early stages, and we must support our industry partners because if this ecosystem eventually grows and strengthens, we will all succeed.

TechFlow Deep Tide: Currently, Pyth mainly supports DeFi protocols. Will you venture into more areas in the future?

Mike Cahill: It's too early to draw conclusions now. There are many definitions of success at the moment.

For Pyth, because DeFi uses financial data, we want to be the best product in this vertical field in the world and then move on to other areas. I can imagine many things, such as using other types of data, there will be many exciting and interesting innovations, but there are currently no plans to enter other types of data.

TechFlow Deep Tide: Okay, thank you. Another question is about the recent collaboration between Google Cloud and Layer Zero in the oracle field. Could you share some thoughts and potential impacts of this collaboration?

Mike Cahill: Because this was just announced, I don't know the specific details either. Layer Zero is a bridge, and Google Cloud is a cloud service provider. So if I were to speculate on what they are doing, I would say that in the Pyth model, Pyth uses Wormhole as its bridge. But the way Pyth sources data is very unique; we have 90 data publishers. Google Cloud doesn't give me the impression of being a data source, and of course, it doesn't have a competitive advantage to incentivize protected financial market data to be brought onto the chain. Therefore, it gives me the impression that it may focus on non-financial data and find a more suitable domain. For example, open-source data, such as what is the current temperature in Singapore? I think something in the public domain being brought onto the chain through Layer Zero would be very suitable. I find it hard to believe that Google Cloud can obtain data for the financial market without a specific business model.

Future Plans: All financial data in the world will be transmitted through Pyth

TechFlow Deep Tide: I have a personal question, could you please introduce which DeFi protocol you think will stand out more in the next 10 years?

Mike Cahill: This is also difficult to predict. But Synthetix has already proven to have enduring strength. They are one of the earliest derivative protocols and one of the largest on-chain derivative protocols. Recently, they are deploying V3, and their team has gone through so many cycles, so experienced, so it's hard for them to fail.

TechFlow Deep Tide: Another question is, there are many oracles now, and in the long run, the overall market landscape will definitely undergo some changes. How much market share do you think Pyth will occupy in the future oracle competition?

Mike Cahill: This question is difficult to predict. But from the perspective of financial data, we believe that all financial data in the world will be transmitted through Pyth, just like I mentioned, all the music in the world will be transmitted through Spotify. This is not unreasonable. In our vertical field, in DeFi or financial data, I expect 100% of the data to be transmitted through us.

TechFlow Deep Tide: Cool, the last question is, perhaps many users will be interested in the whitepaper you mentioned just now, because there are many practical references to Pyth tokens and mechanisms. Other oracles have already launched their tokens, for Pyth, can you share the unique approach and advantages of Pyth tokens? Or any future plans?

Mike Cahill: This token is a governance token. It will allow users to decide the direction they want Pyth to take. Similar to other governance tokens, it will also be used for fee payment, rewards for publishers, and may be used for publisher staking to enhance network security.

There is currently no announcement about when the token will be issued. But as I mentioned, we are working to establish relevant applications so that we can have a token-dominated governance model. Some tools are being built, and we are excited to see how the community will take Pyth to a higher level.

TechFlow Deep Tide: So this token will be a community-driven token?

Mike Cahill: Yes, it will be a community-driven project, serving as a community governance token, that's the goal.

Deep Tide Note: To learn more about the principles of Pyth, feel free to read: In-depth Study of Pyth V2: Key Transformation of Price Feed Model, Explosive Productivity after Cross-Chain and Multi-Scenario Support

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。