Ocean AMA with Wolf Crypto
We held an AMA with Ocean in our public Telegram channel on October 26th.
Ocean Protocol is an ecosystem for sharing data and services. It provides a tokenized service layer that exposes data, storage, compute and algorithms for consumption with a set of deterministic proofs on availability and integrity that serve as verifiable service agreements. There is staking on services to signal quality, reputation and ward against Sybil Attacks.
Ocean Protocol helps to unlock data, particularly for AI. It is designed for scale and uses blockchain technology that allows data to be shared and sold in a safe, secure and transparent manner.
Ocean believes “big data” severely disadvantages users and network participants by sharing and monetizing their data while users have little-to-no control over the process. To fix this, the Ocean team aims to build a decentralized data exchange protocol that lets consumers, corporations, and everyone in between exchange data with symmetric control and transparency. Ocean also aims to activate massive amounts of dormant or un-used data by incentivizing participants through monetization. The ultimate goal of this protocol is to fuel continued artificial intelligence (AI) innovation, a technology that relies heavily on data inputs and a robust, reliable data market.
Ocean Protocol recently announced an update to their protocol, the long awaiting v3…You can read more about that here —
Ocean Protocol v3 launches on October 27th! You can join the Ocean Protocol v3 launch webinar here —
You can also visit the Ocean Telegram Group or the Ocean Website for further details.
Here’s what Ocean’s Founder Trent McConaghy had to say in the AMA.*
*This AMA has been edited for clarity.
WC (Wolf Crypto Telegram Member)
INTRODUCTION
WC
Welcome to yet another edition of Wolf Crypto AMA…the first for the week, after back to back to back AMA’s to end last week…Just bull market things I guess!
Joining us today is Trent McConaghy from Ocean!
TEAM & TEAM GOALS
WC
Trent, it’s our pleasure to have you here today. Before we get too far into things, how about a little background as to your history in the space…and any other relevant industry experience you might have?
TM
Hello! Great to be here.
Ocean is about AI, data, and blockchain. I started working professionally on AI in the mid 90s, and did two startups and an Engineering PhD related to it.
I’ve been working intensively in blockchain since 2013. First with ascribe — IP ownership, then with BigchainDB — databases, and on Ocean for several years now.
WC
Nice, now I know you have quite a large team…which is only dwarfed by the amount of advisors you have on the project!
If we spent time talking about them all, I think it would take up all the time we have for this AMA, so instead, how about a TLDR of the most relevant and shillable folks involved?
TM
The two founders are myself and Bruce Pon.
Bruce is an Engineer, and used to build banks for the likes of Daimler and Volkswagen. We’ve worked together in blockchain since 2013.
And yeah, we have a lot of amazing advisors, the full list is here —
WC
Now one of the things I very much enjoy with these AMA’s, is talking to projects who have been around for some time, as opposed to all the “launching soon” ones we see in a bull market like this…the one thing I didn’t enjoy about this one was the sheer amount of information you already have out there…and how long it took for me to read through and research the project…haha!
So with that in mind, how bout we take a trip down memory lane and talk me through how Ocean Protocol came to be and what’s gone on since your raise and token launch?
TM
In late 2016, I started thinking about how blockchain might help big data, and AI. All roads pointed to decentralized data exchanges as the linchpin technology. I wrote about it.
From that, Toyota Research reached out to us, and with them we built a prototype decentralized data exchange for self-driving cars, announced in May 2017.
We decided to dive deeper. So in the summer of 2017 we explored many token models; our approaches there helped to form the basis of the emerging field of Token Engineering.
We raised money in early 2018, and a smaller top-up round in early 2019. The OCEAN token went live in early 2019.
With the money, we grew the team and engaged with users. We shipped Ocean V1 in summer 2019, V2 in spring 2020, and V3 is coming this week. We’ve worked closely with startups and enterprises on data marketplaces, such as dexFreight and Daimler.
We’ve kept busy!
WC
It sure looks like it!
That’s QUITE the list of development and achievements…Let try and go through all that in a logical manner…firstly let’s unpack the underlying protocol that powers the platform itself…
Give me your elevator pitch on what Ocean Protocol is, what it does and why it’s needed?
TM
Here’s the (simple) what: Ocean is Tools for the Web3 Data Economy.
Here’s what it does // what you do with it:
You can use Ocean Market app to earn by selling data and curating / staking on data. Use Ocean Protocol libraries to build your own app for secure, privacy-preserving data exchange.
In Ocean Protocol, each data service gets its own datatoken. This enables data wallets, data exchanges, and data co-ops by directly leveraging crypto wallets, exchanges, and more.
The “why it’s needed” has a bit more context…
The data economy is already huge — $377B+ in Europe alone. But there are two problems.
First, only a few companies benefit from the current siloed structure and because of this, the status quo represents a shadow data economy.
Second, every day, individuals, companies, cities and even nations struggle with who can see their data, and who controls their data. There’s widespread surveillance, hacks, and more. These are issues of data privacy and control, or more generally, data sovereignty.
Why is this? Two things. the lack of open infrastructure to enable secure, privacy-preserving sharing & monetization of data among data owners and data consumers;
And The lack of an easy-to-use, scalable approach to turn data into data assets: to establish a price for data that the market accepts along with supporting financial tools.
With that ‘why’ context, here’s a more elaborate ‘what’ for Ocean:
Ocean brings all the capabilities of data sharing in a web3 native architecture that (a) makes it easier for people to publish their data, auto-price it, and sell it, and (b) bridges the gap between data services and DeFi, a set of tools for financialization of data assets, including lending, exchanges, and derivatives.
WC
Wow, that’s a LOT to take in…
Before we get too far into things, one thing I also like to question “Web 3” projects on, is what their definition of that actually is, as it seems to vary wildly, depending on who you ask…so…what’s yours?
TM
I think it’s great if people have different framings for a field or movement. AI is like this, blockchain is like this, and Web3 is like this. We can still collaborate even if we have different definitions:)
So here’s one Web3 framing — Web3 is a movement and collection of tools that combines the self-sovereign goals of Web1 with the UX & scale of Web2.
WC
You mention monetizing “data” as one of your main USP’s…contextualise that for me…what type/s of data are we talking about here?
TM
It’s mostly about data of value yet has not been unlocked yet, for fear of loss of privacy & control.
This includes individuals’ personal data, or organizations’ data that is sitting latent.
Ocean gives a way to unlock its value — to monetize — while preserving privacy and control.
The verticals can be logistics, mobility, health, whatever.
WC
Amazing, actually know a fair bit about this one having funded Decentr along the way…but in the context of Ocean, I’m interested to know…
You have “data providers”, “data consumers” and “data curators” as participants in your ecosystem, break down those roles for me as to whom they might be and perhaps a real world example of each?
TM
Data providers are people and businesses who have the data I’d mentioned above, interested to monetize it.
Data consumers are people who know how to create value from more data. Most typical is data scientists & AI practitioners who use data to improve model accuracy, for applications like scheduling logistics better, building self-driving cars, predicting weather better, etc.
Data curators are people who stake OCEAN on particular datasets. They earn transaction fees proportional to sales volume of those datasets. Stake is a good signal for the quality of a dataset, since people are incentivized to earn more. (And helps filter out spam then too.)
As a real world example:
In dexFreight data marketplace: data providers are trucking companies. Data consumers are logistics optimizaiton companies and wall street. Curators are people in the industry who get the datasets, but also more broadly data scientists and any OCEAN holder, really.
WC
I’m interested to know, how does one go about selling personal data while maintaining privacy? Isn’t that somewhat of an oxymoron?
TM
I find it hilarious that it sounds like an oxymoron at first glance. To me that’s a sign we’re onto something :)
The trick is to sell *access* to the data, for “AI eyes only”.
You bring compute *to the data*, to train the AI model on-premise next to the data.
Then the updated AI model can be used for business outcomes like mentioned earlier. It can be simpler than AI too, it could be as simple as computing an average next to the data (business intelligence or BI).
So people never see the data that’s sold. And the AI training algorithm only sees it locally, just long enough to get value.
WC
So one of the things that I always find interesting when it comes to blockchain and data, is that there is a large amount of scope for abuse, and now with the European Union introducing GDPR regulations, a large amount of compliance necessary if you want to operate in Europe.
So in a very long winded way, how do you ensure Ocean can’t be used in a nefarious manner…say Child Porn as an example, and how do you also make sure you’re compliant with GDPR regulations while maintaining decentralisation?
TM
Re filtering content:
Ocean is on a permissionless blockchain; Ocean can’t censor at that level. But the last mile — the webapp — you can.
This is similar to how TCP/IP + WWW work, eg Google filters out the bad stuff in its search results, despite there being not-so-nice webpages.
It’s also similar to how other Web3 apps work, such as Uniswap (permissionless base layer, but the webapp geoblocks).
Re GDPR: Ocean doesn’t host data. In fact, Ocean helps GDPR. GDPR doesn’t like when data leaves certain jurisdictions. Ocean compute-to-data makes this easy to follow.
The latter point makes Ocean nicely attractive to data regulators. They’ve had a tough time enforcing GDPR. Ocean helps avoid the scenarios to even need to enforce.
WC
Past the data component of Ocean, in researching for the AMA I came across a lot of information regarding Ocean and “AI”…It seems your end goal is for AI to have accessing to your platform and the data contained within…for…something?
Now you’ll have to consider me a little bit of a skeptic when it comes to blockchain and AI…as in my time in the space, I’ve seen a hundred AI style projects come and go, so from you end, I’d like to know what importance AI has in relation to Ocean Protocol and specifically your USP of data as value?
TM
I get that people are skeptical about AI. It’s important to understand that there are excellent AI use cases.
I’ve had personal success in building practical AI systems: my previous two startups used AI for designing computer chips. Both were acquired. The products from that work is used by the likes of Apple, Sony, TSMC and Qualcomm to design their chips. The computer you’re typing into right now was probably developed with the help of that tech.
In Ocean, we talk about AI because AI tools know how to extract value from data. And, we know AI:)
BTW one of my favourite framing of AI is simply “a set of tools to do things that traditional numerical algorithms could not”
The main point is: don’t let skepticism deter you from using a good tool!
WC
I believe you’ve gone through a few iterations of the protocol so far…
Can you step me through the initial protocol, what it was designed for and how that’s matured along the way?
TM
Ocean v1 did blockchain-based access control, but in a customized way.
Ocean v2 brought in compute-to-data ,a major feature for privacy as discussed above.
v3 does blockchain-based access control but tokenizes it, which makes it simpler and more interoperable.
In Ocean v3, every single data service gets an ERC20 datatoken.
WC
We’ll get to v3 in a bit, but firstly I wanted to understand a bit more about your network infrastructure…You make a claim as to decentralisation, but exactly how decentralised are you?
…I think we all know by now, decentralisation is somewhat of a spectrum!
TM
Yes decentralization is a spectrum, and, decentralization is a journey.
Decentralize everything too soon and you can’t adapt to achieve product market fit, to build something people want
But ultimately you want something that acts like a public utility…Hence, a journey.
Here’s Ocean status for v3:
Network deployment. In line with the roadmap, Ocean V3.0 will be deployed to Ethereum mainnet, which is censorship-resistant and permissionless-decentralized.
Smart Contracts. The aim is to move away from a small handful of gatekeepers (like v1/v2 was). Here’s how: Ocean v3.0 contracts do not have upgradeability built in. Therefore the only way to upgrade contracts is by community consensus to use a new set of smart contracts. Sometimes, less is more.
Finally, OCEAN token control will change soon as well.
WC
Community consensus sounds a lot like a DAO?
TM
Community consensus in this sense is:
If you really want to upgrade, you deploy a new version of the smart contract and you ask the community to point to that.
It’s like a hard fork, but at the smart contract level. If people don’t want to come along, they don’t have to, they could use the old version.
There’s still an aspect that is not decentralized: funding the Ocean community to help long-term sustainability. The decentralization journey for this begins in Ocean v4 “OceanDAO” project. We plan to deploy the first version of that in Q4. It will be a multi year journey.
WC
The token would remain the same as part of this process? Or be backwards compatible etc?
TM
It would remain the same.
WC
Explain to me the role of nodes…or “Keepers” as you refer to them as part of this network infrastructure and what their incentivisation is to participate in the network?
As I assume they play a pretty critical role in all you’ve described above?
TM
Keepers are an artifact of v1 and v2, when we had our own POA network. In v3, we are deploying to Ethereum mainnet. The concept of keepers goes away. (As part of decentralization process I described above)
But we do have incentives.
Blockchains are incentive machines. They’re great at getting people to do stuff to add value to the network
So the questions to ask is: what do you want people to do?
In the case of Ocean, we want people to supply data, to consume data, and to curate it. Suppliers are naturally incentivized : if they get paid then they’ve got good data. Buyer are incentivized naturally too: if they find the data useful they will buy it
Curation takes a bit more. You want an authentic signal for quality
We solved that when we solved the challenge of “how do you price data?”
The answer is: create an automated market maker (AMM) pool of OCEAN and datatoken.
In the pool, the price is auto discovered. Boom.
And curation = amount of stake in the pool. Because people get tx fees on data that’s sold. Useless data won’t get sold.
And that’s the summary of Ocean incentives (at the marketplace level). We also have broader incentives for funding. That’s part of v4 / OceanDAO. But I will stop there!
WC
Amazing…you’re segueing me into my next question almost perfectly…cause setting a price on value is one thing, but you need a venue to buy or sell such value on…so…TLDR for me what the Ocean Market is, and what’s the intended purpose for it?
TM
Ocean Market is a community-oriented data marketplace, powered by Ocean tech
In our v1 and v2, we had tech for people to build marketplaces. But it was hard for people to imagine what could be done. And there was no place for the community to buy & sell, unless going to a vertical-specific 3rd party marketplace.
Ocean Market addresses both those issues. People can use Ocean Market to: publish data, stake on data (curate), and buy data.
WC
So what’s the logic behind using AMM technology to power the market instead of a more standard order book?
TM
Order books and AMMs (and auctions) provide price discovery, when we compare those…
-Order books. For a sale to occur, order books require bids and asks to match up in real time, aka a “double coincidence of wants”. This is not feasible for newly created long tail assets like datatokens. (However, they are useful once a datatoken gets enough liquidity and traders).
-Auctions occur over a time interval, such as an hour or a day. Auctions are useful for an initial pricing, but after that we still want automated price discovery for the rest of the lifetime of the assets. Auctions do not provide this.
-AMMs provide automated price discovery without the disadvantages of order books or auctions listed above. AMMs work for an initial asset offering and throughout the asset’s lifetime. AMMs don’t require a double coincidence of wants; they can be thought of as robots that are always ready to buy or sell.
Conclusion: Ocean’s datatoken framing allows for any of the above. And Ocean Market focuses on the most promising: AMM pools.
WC
So as we know, not all AMM’s are born equal, and with the success of Uniswap, we’re seeing a bunch of different types of AMM’s pop up…
So for Ocean, what’s the best fit and why have you chosen it?
TM
So we’ve got AMMs under the hood. We worked closely with the Balancer team to provide it. We chose Balancer because of its weighting scheme.
Specifically:
The publisher has lots of datatokens. If we make the weight 90% datatoken and 10% OCEAN, then they only need to provide 1/5 as much OCEAN liquidity as they would have done compared to weights of 50% — 50%.
That’s why we chose Balancer for under the hood. We are excited to see datatokens used with many other AMMs in the wild however!
There’s a lot of great innovation happening from Uniswap, Balancer, Bancor and more.
WC
So you semi answered this before with the curators…but, how does Ocean, or any other party, ensure the quality of the data on the platform, more specifically…
I.e How do you ensure I can’t come along and sell a bunch of fake or copied data on the market?
TM
If you start selling crappy data, few or no people will buy it, and if you don’t believe in it then you won’t stake on it, nor will others, so it won’t be very prominent in search results.
When people post data, they are making a declaration that they have the IP rights. If they don’t, then the standard laws apply.
If people have really valuable data, then there’s another way to guarantee exclusivity:
Use compute-to-data. Then their data never leaves the premises. No one else can copy.
WC
Is this a read only access kinda thing?
TM
So compute-to-data is not just about privacy and control. It actually changes the economics of data.
In compute-to-data, the seller’s data never leaves his premises. The ‘buyer’ has an algorithm run next to the data, and returns only the compute results.
“Read only access” does not quite capture that because people could still download.
WC
It’s my understanding that although you’re kicking things off with Ocean Market, you’ve developed the infrastructure to be able to support multiple markets, that anyone can spin up…
What’s the reasoning for doing so and how can I spin up a market and use Ocean to power it…and why would I?
TM
Correct, there’s an infrastructure. Ocean’s core infra is: smart contracts + libraries that make it easy to publish ERC20 datatokens, and consume them. And publish AMM pools, and use them.
Then we have Marketplace code on top. And React hooks etc.
In Ocean, we’ve always seen that we want to catalyze a whole open data economy. And an economy is not just one market, but 1000's of them. We’ve set up the incentives such that each marketplace gives a tiny % of revenue back to the Ocean community. Marketplace runners can specialize. Different verticals, pricing strategies, go to market strategies, etc
How to spin up a market? 3 ways:
Fork Ocean Market code. We will open source it soon.
Or, build up using our React based tutorial
Or build more directly using ocean.js etc. You could even integrate into existing apps.
Marketplaces will come in all sorts of shapes and sizes. E.g. Metamask now has a DEX built in. Well if you have datatokens in your Metamask, your datatokens are in one of the places that Metamask points to, then you can trade those directly in Metamask.
WC
Amazing, when can we Ocean Protocol v3 to go live?
TM
This week. And specific info will be coming very soon!
WC
Ok so I know we’re running real low on time and I have to ask about v3…
You ticked off a few answers to this before, but to summarise, Ocean v3…what is it and why the upgrade…something to do with DeFi?
TM
Ocean v3 = datatokens + Ocean Market.
Datatokens are the new way we do access control. Simpler architecture yet more interoperable. They’re ERC20 so they play well with DeFi.
Ocean Market is the community marketplace. Fixed pricing + auto pricing. Auto via Balancer AMM.
The other big thing about Ocean v3 is that it gives new potential for people to earn in the Ocean ecosystem. You can earn by (1) staking OCEAN on datasets in Ocean Market, (2) publishing and selling your data in Ocean Market, and (3) building & launching your own market.
(1) and (2) are wholly new, and (3) is a lot easier.
WC
ERC20 datatokens…they can be connected up to various other DeFi ecosystems?
TM
Correct, think of it this way:
Just like REN and others wrap BTC so that BTC can be an asset in ETH DeFi,, Ocean wraps data so that data can be an asset in ETH DeFi.
Bitcoin has >$100B in AUM… on Bitcoin chain. So pulling it into ETH DeFi is a big deal. $1B and counting. The data industry in Europe alone is $400B. We see Ocean as a big fat bridge from the data industry to DeFi.
It will take time of course. But this is the aim.
WC
Contextualise for me, with this newfound ability and being able to hook it into DeFi, who do you expect to start creating datatokens now…and what for…past your existing use cases…
I can datatoken my AMA’s and sell them for mad cash on Ocean?
TM
Lol if people will pay you mad cash for your AMAs, then yes! But for sure you can sell them on Ocean Market if you want to.
The users of Ocean Market who want to buy & sell data , which we discussed before, are using datatokens. Datatokens made it possible to leverage the AMM for auto pricing.
We see that datatokens are exciting because they unlock many more use cases. ERC20 wallets become data custody and data management tools.
DEXes and CEXes become data marketplaces. DAOs become data DAOs. For data coops etc, and you can even do loans and other crazier DeFi stuff… on data.
This won’t be overnight. But the ERC20 nature makes it much easier to do than before.
WC
I love the idea of that, loans secured by personal data! Default on your loan and we release your nudes!
Social reputation is one of the things Aave were looking into, though I don’t think they’ve tokenised it.
TM
Yep! There’s a precedent, in Bowie Bonds.
The idea is that if you can establish a value for your asset, then you can borrow against it.
How to establish value for an asset? Auto-price it with an AMM!
Reputation is a weird thing. The best approaches to reputation are extremely specific. Focusing on a narrow use case. Like Reddit karma, or an Uber rating.
WC
And if I were to do that, I could do it via an…IDO?
TM
Exactly, IDO = Initial Data Offering. Someone in our community coined this, I love it.
WC
So I know we’re at the times up stage, can you drop some of your Medium resources on me for future readying relating to these latest updates?
TM
This is a listing of all v3 related posts:
Here’s the main post on Ocean Market:
Here’s the main article on datatokens:
FYI Ocean Market is the place to do staking from. (Staking = adding liquidity to an OCEAN-datatoken AMM). Here’s an article that drills into it:
And we’ve announced the Ocean Data Farming program, to help incentivize people to stake more. It doesn’t come out with V3, it will be after.
Actually.. you’ve had great question’s, So perhaps…hmmm.
How about we do something special right now…Ready?
Ocean v3 will be released tomorrow. (Tue Oct 27).
We’re doing a webinar tomorrow on Ocean v3. I’ll be a part of it, as well as a couple excellent engineering colleagues.
Here’s the link:
CLOSING THOUGHTS
WC
Ok I think that’s a pretty good note to end on…You heard it here first people!
TM
OK, I really do have to run now, thanks again very much for the great questions!
Ciao!
Ocean Protocol Resources
Website: https://oceanprotocol.com
Telegram: https://t.me/OceanProtocol_Community
Twitter: https://twitter.com/oceanprotocol
Medium: https://blog.oceanprotocol.com
Whitepaper: https://oceanprotocol.com/tech-whitepaper.pdf
Wolf Crypto Resources
Public Group: https://t.me/WolfCryptoPub
News Channel: https://t.me/WolfCryptoAnnounce
Twitter: https://twitter.com/WolfCryptoGroup
