[Messari AMA] Kain Warwick, Founder of Synthetix and Justin Moses, CTO

Messari hosted a public AMA with Synthetix's ($SNX) founder Kain Warwick and CTO Justin Moses in the Messari community group. This is a transcript of the conversation.

Messari: Hi all! I'm excited to have Kain Warwick and Justin Moses of Synthetix here this afternoon to speak about the work that's being done with one of the most exciting projects in the DeFi space. Thanks for being here Kain and Justin!

Kain Warwick: Thanks Jack! Thanks for joining everyone

Justin Moses: Hey everyone

Messari: We had a lot of great questions submitted but before we get started could you guys give us a quick rundown of your backgrounds and what brought you to Synthetix?

Warwick: I have been in crypto since 2014, I run a payment gateway that works with a lot of the exchanges in AU and now some overseas ones like Binance. I started working on Synthetix (when it was Havven) in late 2016 trying to find a way to solve arbitrage of crypto prices in smaller markets like AU and Korea when prices started to diverge.

Moses: Like Kain, I'm from Australia but moved over to NYC back in 2010. I've been working as an engineer in web applications since the early 2000s, most recently as Director of Engineering at MongoDB. I've been interested in the crypto space these past few years, and moved into Synthetix full time last year. The mission of Synthetix (well formerly Havven) - to create a decentralized stablecoin - was a appealing challenge. We've pivoted since then but the challenges have only escalated :) And like any engineer, I'm a sucker for challenges

Messari: Awesome, thanks! Going off the origin of Synthetix, can you talk a little bit more about the switch from a stablecoin to synthetic assets?

Warwick: Yeah for sure, back in 2016/2017 it looked pretty unlikely that regulated stablecoins would be a thing in the near term so the best option was going to be using a decentralised solution. Obviously that turned out to be a pretty wrong assumption when TUSD, Paxos, Gemini etc all launched. Obviously Maker hadn't launched DAI yet then either, so it looked like there was a large market opportunity (long term I still think there will be but even Libra is challenging to decentralised stablecoins) so we needed a way to deliver immediate utility to users. Our mechanism supported multiple assets and so in December 2018 we launch a multicurrency stablecoin. But that was still not sufficient and we were seeing activity from asssets like gold so we decided to lean into the trading side and list some more volatile assets like sBTC (synthetic Bitcoin)

Moses: The pairs we launched in Dec 2018 were AUD, EUR, JPY, XAU (gold) and KRW. (on top of USD which is our base stablecoin which we already had

Warwick: Trusty old KRW...

Messari: Got it so you guys made this pivot from a stablecoin to a synthetic asset platform, what do you see as the biggest use case for Synths?

Warwick: In the short term the biggest use case is trading. which is why we built synthetix.exchange

Messari: and long term?

Warwick: I think in the longer term things like synthetic equities will be very powerful, as they will provide access to modern financial assets to anyone who can access Ethereum (ideally everyone in the world) and because they don't require custody of the underlying asset they can be permissionless.

Messari: Definitely a huge addressable market! So we've received a few questions regarding the assets listed: How did you decide on adding TRX, XTZ and MKR and what is the decision making process behind assets to add? Are there plans on adding sets from Set Protocol?

Warwick: We have been looking into various cryptoassets that were not traded on exchanges like Bitmex for example that we thought users might want access to. So those three are just the latest, but we also have BTC, ETH and BNB. Also worth noting we have inverse indexes of these assets as well. Though TRX, MKR and XTZ are launching next week. We have spoken to the guys from Set a few times about this, but given the way the system works it would be really interesting to have a basket of tech equities as an example. So you could have a self balancing set of sTSLA, sAPPL etc.

Messari: So to get into the weeds a little could you tell us how Synthetix is able to represent real world assets in a decentralized manner?

Warwick: Sure. I assume most people here are somewhat familiar with Maker, which is the most well known example of crypto collateralised synthetic asset in Dai. We use a similar mechanism but we use SNX (synthetix network token) as the collateral rather than eth. So users lock SNX to issue debt into the system, and we use an oracle to maintain the price of that debt relative the other assets in the system.

Moses: So like DAI represents debt, ours is sUSD. This can be exchanged into any other synth using our DEX, which basically reprices that debt using a price from the outside world

Messari: Makes sense, so you mentioned synthetix.exchange can you talk about how that works and why you chose to make it P2C (peer to contract)

Warwick: Yeah it is sort of P2C by design, which just means that there is no order matching so you don't need to find someone who wants to buy sUSD for sBTC, the contract manages the conversion. This is just the way the mechanism works because you have different types of debt in the system which can be exchanged at the current exchange rate by the contract. So when you convert sUSD into sBTC the system burns the sUSD debt and creates a new debt (with the same nominal value) priced in sBTC.

Moses: It was how we evolved really - we didn't come at it planning to write a DEX, but that's basically where we came to our synths included foreign currencies, commodities and then crypto prices.

Messari: This design works by having oracles maintain the price of the debt in the system. Earlier this week there was an incident that occurred with the oracle which you issued a succinct response (https://blog.synthetix.io/response-to-oracle-incident/) to but for those unfamiliar could you give a breakdown of what occurred?

Warwick: We use a number of oracles for different prices, forex etc. One of these went down which happens more often that we would like tbh since they are commercial API's (but that is a different story) then a second one started sending a price for KRW of 1000x more than the actual rate. This caused our Oracle to change the rate for KRW several times over an hour to a rate that was significantly incorrect.

Messari: This oracle problem is a well documented one in crypto, do you think its one that can be solved?

Warwick: Solved is hard, I think it can be made more robust for sure. The issue is ensuring that the feeds are not centralised, which can't really be solved but can be improved significantly. The second is ensuring that the data can't be manipulated or served incorrectly, which is also a hard problem. Like most crypto mechanisms you can really just have assurances up to a point, that the cost of attacking is higher than the pay off from success.

Messari: Exactly, so that's probably a good segway to your partnership with Chainlink can you speak to that and what it will look like?

Moses: Sure. We've put together a PoC of using chainlink in correlation with our existing oracle so we can start to migrate over. However we were waiting for them to add some functionality that we need to get us up to par with our current centralized one. We've also just been trying to priotize the work we have on our plate, so now that they now have most of the features we need on mainnet, it's on us to get in integrated. As you can imagine, it's fairly high priority right now.

Messari: Definitely. So we have a few more miscellaneous questions here and then we'll wrap up. With more smart contract platforms recently going live or in the near future, have you considered building on other blockchains besides Ethereum?

Moses: We actually spent a fair bit of time working with EOS (myself in particular) - I even spent some time trying to get a good CI (continuous integration) workflow to match something like truffle & ganache in Ethereum. We ultimately decided to postpone it as a) I personally felt the technical community support in EOS lacking and b) after the pivot to synthetic asset trading it really didn't make sense to try to maintain support to multiple blockchains at once

Warwick: Also Ethereum is really the only viable platform right now for DeFi. so that is a big factor in us putting all of our efforts there.

Messari: What do you guys see as the biggest barrier preventing DeFi/synthetic assets in particular from gaining widespread adoption?

Warwick: Right now the people using these systems are almost all crypto native. So we need better methods to onboard non-crypto people into the space. I think there are also some primitives like ID systems and custody that need more work, but lots of projects are making progress here.

Moses: I also feel that finance and trading can get heavy quickly (how many people do you know who could explain how shorting works?) and that in and of itself is a bit of a turn off. But of course if there's a financial incentive, people will spend the time learning. Though the chasm between those who understand and those who don't widens very quickly - it can very be daunting to anyone outside looking in.

Messari: Other than tokenized exposure to real world assets, what are some other areas of DeFi or crypto in general that interest you guys?

Warwick: I think a lot about identity and how to manage it in a decentralised context, so that is something I follow closely. I think another aspect that gets a lot of people interested is governance, which is not just a DeFi thing but just a more general crypto issue that needs more experimentation.

Moses: Definitely agree with identity. Also if there's a financial incentivize to onsell one's purchasing decisions, it's a better avenue that what we have now with the big tech companies

Messari: If someone figures it out that would open up a lot more interesting use cases forsure

Warwick: Yeah Vitalik actually presented on this at ETHNY and it was an interesting approach, again I think there are mechanisms that can work but it is about raising the costs of Sybil attacks vs friction for users.

Messari: But last questions here, what are you guys most excited about thats coming up the pipeline for Synthetix? And what is the best way for people to keep up with Synthetix developments?

Warwick: The best way is our discord or blog. I am a recovering crypto trader, so the most exciting thing for me is probably the synthetic positions mechanism we are building. Which will allow users to trade with leverage and take long and short positions on a lot more cryptoassets that are not available today.

Moses: Easily the decentralized oracle for me.

Messari: Awesome, well thank you guys so much for joining its been very insightful. I'll unmute the chat now so feel free to stick around if you'd like!