Dr. Prakash Kamaraj tells us about the rapidly evolving tech landscape in India and how this is impacting the AI and Web3 industries.
Dr. Prakash is leading the Indian Ocean Outreach initiative – https://indian-ocean.gitbook.io/indian-ocean-outreach-program/
Indian Ocean Hackathon details – https://indian-ocean.gitbook.io/indian-ocean-outreach-program/indian-ocean-hackathon
Ocean Missions Hackathon Bounty (Creative use of Etherscan API) – https://indian-ocean.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F8bVk2qqeYgIMe1pocDxe%2Fuploads%2FDGoHxmhMyMyVoRzhkmA7%2FOcean%20Missions.pdf?alt=media&token=d76848e5-47e3-4ed5-a78b-64be75027a35
Ocean Academy (Ocean 101) – https://oceanacademy.io/
The following is a rough transcript which has not been revised by Ocean Missions. Please check with us before using any quotations from this transcript. Thank you.
[00:00:00] Scott: So today we’re going to be speaking with Prakash from the Indian ocean outreach program. Welcome to the podcast per cache.
[00:00:06] Prakash: Oh, thanks buddy. Yeah, pleasure to be here.
[00:00:10] Scott: I was wondering if you could just tell us a bit about your, your background a bit about yourself and what it was that led you to Ocean Protocol
[00:00:18] Prakash: yeah, I think I started off with a long story, short version, I think my switching careers or switching fights in search of learning is not new to me. So my first switch began in 2015 where I actually copied it in medicine, but I wanted to be a technologist, more of a one to many. Then one-to-one.
So I did my engineering matters as well, where I had a chance to do it in an Institute of technology, one of the leading institutes in India. But essentially it was like a really good master’s program that gave us exposure to a lot of eating degrees as well, as well as project experience. And we had the Liberty to choose a lot of new stuff that we wanted.
And I kind of majored in more of a engineering tech in terms of AI products and innovations. So that was the beginning of my data science experience, or even held their experience if you would say so. And immediately after that, I think my industrial experience gave me a lot of bandwidth to learn a lot of stuff, led to health tech and engineering and even aspect to product management and project management.
So that gave me the. Overview of what’s possible in the industry and why a lot of doctors are actually missing out on all of this stuff. And that’s how the Genesis started. Initially. You were a point what kind of, kind of, you would say stop me a bit was a lot of solutions that we build. It’s always stretched.
Let’s do it in silos. Like you, I develop a solution. It’s going to be held like a pittance. Yeah. I’m rarely do I find a lot of cross border and seamless. Innovations happening. I think that’s, that’s the major drawback within the web to ecosystems. And once the pandemic hit I kind of researched more on private AI
I think, I believe private AI I had a lot of scope in developing the scope of projects to be involved in a lot of applications. Right. It’s not just, you develop a solution it’s only for one sector or one geography. It’s very little. So that’s when I started reading a little more good private AI when it came to doing a bit of research with the private Dave in terms of vaccination passport, but that was just the beginning.
So after that naturally my interest towards AI and blockchain kind of increased and it didn’t really hit, it hit the impact of cryptocurrencies in general. Didn’t hit me until middle of the pandemic like that or somewhere. Middle of 2020, and that’s what made learning, learning curve increased. And then I knew, okay, what’s actually been possible is it is obviously the first gateway for a lot of people who are, who try to get beneath the investing level.
So once I got into that, then I think that began my search. I always found like DeFi ecosystem is very mature, but what about AI? What about data science? Is the technology not robust enough to handle these operations or. What is the limiting factor is what’s my big question. And I kept searching for a lot of similar solutions.
And finally I did notice Chainlink and the graph, but there was still something missing. I think that’s kind of somehow led me to ocean protocol. When I was looking at this friend actually pointed me from an investment perspective, then I dig deeper. Then I saw the look at the old activities behind the team and all the activities of the ocean DAO
and I think that’s what this spurred me on to it because. This is the one I had been waiting for. Maybe he could say that because it’s a lot of future attached to it. And the tech is already general enough to handle a lot of, lot of use cases. And whether it’s health tech, or generally AI products, I think that’s what I wanted to be built out of sorts.
And that actually kind of brought me towards ocean so to speak. And that’s probably why I’m still here waiting to build a lot of stuff.
[00:04:02] Scott: Great. And would you mind just defining private AI?
[00:04:08] Prakash: Yes. That’s a good question. I think it would be a lot of useful for listeners. So imagine when, so that is this question began to me. When we were discussing vaccine current years, a lot of people were hesitant in sharing a lot of details, but when it comes to a public good, you tend to see a lot of things.
And even if it tends, there are good intentions behind the data can always be misused and retrospectively, you know, for a lot of the terms and conditions always change. Right? So that is something that always keeps changing and you can never predict what happened to you. There is no proper audit and so much missing out, but in the Web2
I noticed that the private from in terms of private day means it’s essentially placing a gating mechanism so that there is enough privacy for the user themselves. And even with the privacy, the shadow privacy on a spectrum particular data or an identity, the stakeholder can still make use of the information.
So we were exploring while working on. Vaccine passports initiated with MIT, but check labs. So this was a research group that was specifically researching on private AI, like differential, privacy, federated, federated learning, spit learning. That’s another technique that’s really curious that they will learn about.
So these techniques were essentially a bit of a. Data science and machine learning techniques, if you sort of speak so where it allows for people to perform computations and analysis on data assets in the traditional world. Even if you add a lot of privacy layers on top of it, because there’s always a Pareto frontier, you could call it like you had to adjust for how much information can you actually reveal and how much do you have to consider?
Because if you’re going to consider 90% of the data. Identity, it will not be useful for research states. So at least for research of public good, you still need to provide a decent parameter preparation. So I think, I just think for that ratio research has had began quite a lot. It has been quite for awhile as well.
All these techniques like differential privacy are very important in medical college. And federated learning, which was a propaganda propagated by Google. It shows that the data doesn’t leave your mobile phones or laptop devices. So that’s one way of safeguarding data, privacy of SAR. So that essentially entire gamut, that effort is referred to as private day and a lot of, and I thought I was wondering, why is this being hugely invested on groups?
Like. You didn’t not this Facebook. He has, there are so many folks coming up, Facebook, Amazon, they would actually investing in private AI research as well. So obviously I think the answer to that is almost obvious because Facebook, Google, they’re known to harvest data. They’re known to have so much trust issues with the governments, even with the running, with the GDPR and lead.
Governance bodies. Right? So I think so they took a step back and recognize, okay, data privacy is important. I think we need to, we need to be the first people to do that. That’s what I did, but there was still something off with it. Why should I trust a data privacy mechanism by Facebook or by Google?
That is always something off with that. So I was actually looking for decentralised way of doing. And I think that’s when I came across a lot of protocols, but none of them were actually talking about it. And when I looked at ocean, I had his first discussion with Trent. Like you have so much traditional methods in the real life.
So how is it compute to data actually effective? And he gave actually a very convincing answer saying that these techniques have been around for quite awhile, but they’re not really blockchain effective. Right. So. You need to ensure you’re never uploading any sensitive data on the blockchain. You still need to maintain privacy by the data.
Doesn’t leave your sites so compute to data makes sense in that regard. So as soon as I understood more of the concept, that kind of made sense, okay, this is why compute to data is necessary. And that’s. It’s very effective. And I thought it was a proof of concept, but when I looked at the tech stack, it was been quite an issue for quite a while.
I think that’s what kind of captivated me to look at the space in more detail,
[00:08:20] Scott: private AI. Is then excuse the layman’s terms, basically this idea of accessing a lot of data from a lot of different sources and ensuring that that data doesn’t necessarily leave the device that it’s on. So whether that’s the person’s phone or laptop or so on.
[00:08:41] Prakash: Yeah, exactly. So there are two kinds of private AIs so one, you have to ensure that it doesn’t leave the device, but you still can transmit a lot of models. Right? So the complex AI models that are being built are they, you still have to send them to the centralized service. Like there’s no escaping that.
So, but at least that’s still acceptable. You don’t send your data, but you still send a bit of a model. So the baffled company. Can actually make sensible information out of it. The biggest example of that I can point out from techniques like federated learning is your Google board orient mobile, right?
So when you search for recommendations for the, towards the mirror space on your G board, the data as an athlete, each school will. It’s only the model that keeps being sent to Google so they can provide you with really good recommendations. So that is one context of it. The other way of private AI is considering a lot of information.
Like you can see just enough so that your privacy is protected, that you are, your product is GDPR regulated. Right? So just making sure there’s a really good gatekeeping mechanisms so that you don’t reveal sensitive information.
[00:09:47] Scott: Well, awesome. And then computer data obviously is one of the key features enabled by ocean protocol and, and that basically enables data publishers to approve models, to run on their data sets.
And it also ensures that the data of the publisher is never leaving its premise, which is a nice thing for, for GDPR. And so.
[00:10:11] Prakash: Exactly. I, I think we, I think I noticed ocean or ocean stack also being respected in that regard. Right? So given the mature ecosystem in Europe, I think they are very much on the way to do that. If any build product is build out a fortune, I think it will be complaint a lot of, because when you’re building a lot of top donor purchase, like the GAIA-X mechanism, it’s going to have to be GDPR compliant.
There’s no. So I think that gives us some proof of concept, at least even from an outside perspective that, okay, what they’re actually built is will be complained to a lot of data privacy issues as.
[00:10:45] Scott: So you’ve given us a really good sort of overview of, of your background and also an overview of, of ocean protocol.
And you came to ocean protocol or the ocean DAO with the proposal which was all about. Outreaching to the Indian ecosystem. So I thought it would be nice to just explore a bit more you know, from your perspective, the state of the tech ecosystem and India, as it is today you know, how that has kind of matured and changed over the last 10, 15 or 20 years,
it would be nice just to give like a, a really kind of, just a bit of a brief overview of, of how the Tech eco system is kind of emerged and changed over the years in India.
[00:11:32] Prakash: Yeah, just circling back. So when, just right about when I was actually finishing my medicine or even mainly masters there is a lot of. You still have to look at what the research trends are in terms of industry trends, right? So that’s when the AI naturally caught my attention.
And there was a lot of popular courses on beating basic data science, almost everyone was going to have to learn about it. And that interest basically kind of permeated towards all the institutions in India. I think at least every single, every second person in the cream of Institute. As to learn a lot of the AI techniques, even if it is not part of the stream, they end up learning it purely out of interest.
And B what the echoes from the material markets. I right, because even the hiring signals from the big corporates or even the big MNCs, they were actually giving a lot of signals and value to. Our division until the intended analysts. So these techniques were well-respected and there was a lot of values in having that as a skillset when people get placed in the industry.
So this was not just echoed across the industry. I think even the Indian startup ecosystem was valuing a lot of the AI projects as well, not just in town for projects that people submit to the government. You also had a lot of. Startups coming in that field and this also permeated towards a lot of initiatives like AI startups for the the top 30 startups, pretty startups not distant in health sector is I noticed that it was boom, boom, you’d have to all the sectors that I could find.
And that kind of it made sense. It’s not just something that only the corporates are doing. It’s also something the government is taking notice of. That’s something we kind of. And adores the end of 2018. I would say I also started. Absorbing, what are the other space people are looking at then a lot of interests was there in blockchain as well.
So when it comes to blockchain, they’re obviously not very concerned or the private cryptocurrencies aspect, but they were not against the tech. That’s what they, that’s the chance they maintain even to date. So even though there are a lot of regulatory concerns, the blockchain aspects are always well respected.
And I think they believe. It’s not just a buzzword people regard that it’s a good, really good skill for people or companies to adopt as well. Whenever there is a really good use case. So I think that’s when I knew, okay, like this is not just a buzzword, it was actually being inclusion in a lot of investment aspects or even government initiatives as well.
But the only drawback is that they’re only focusing on private blockchain, private blockchain, as in there is no currencies in wall. It’s only for. Operating within a siloed ecosystem. So that again is not useful for anyone. Right? I think at least we in the crypto ecosystem kind of recognize that.
So, but it’s always good to have both both spaces being I studied with intent because we are still early. Like it’s a decade old and we still have so much innovation that’s happening and it just takes a short steep go for anything to kick off. So it is kicking off. I think the fundamentals are really strong for the crypto ecosystem.
And I think with a lot of really good products and use cases being established, and I think the trust in the ecosystem will also grow. And yeah, I think to summarize. In the, as interest in AI has already been established and the blockchain, our curiosity is always there. And people, if you look at the Indian ecosystem right now, the Web3 space.
The Indians are a lot, lot more working on it. And with the regulatory concerns, it doesn’t matter. All they have to do is they just registered the company out of Singapore, Dubai. And that’s how you’ve got to have polygon, right? So they may have entities in India, but the actual Istation, illegal legislation, there’s always a workaround because if you’re doing something useful, nobody’s going to come and stop you.
So it’s only when you start engaging in nefarious activities, that’s when the common crackdowns happen. Right. So in terms of that, so in have been bullish on the emergent. So the AI space has been a lot more mature compared to the crypto space, but I think it will be. I so the best thing about something like ocean protocol is it covers the best of both worlds.
And even to code on ocean, you don’t need to know the blockchain specific language, but it definitely helps if you want to score on smart contracts. So these are the difficulties. I feel a lot These are the panels that are observed, that I feel I should really connect. And that was my first thesis available, a summary of the proposal.
And I still believe that observation holds true as you know, like almost three months into the proposal and having gathered some good fundamentals and feedback from the folks in India.
[00:16:15] Scott: Yeah. And I think it’s an important point. You know, I’ve sort of been following the blockchain ecosystem pretty closely for about five or six years now.
And I know in 2017 there was a lot of ICO activity and a lot of people, you know, raising a lot of, you know, an astronomical amount of funds for some idea or whitepaper for some copy-paste protocol. But it was, it was pretty wild and you still needed to know. You know, a pretty comprehensive understanding of, of, of the blockchain and, and solidity and all those things.
And, you know, in the same way that NFTs have recently enabled use cases for artists and graphic designers and creative people all around the world to, to get involved with the stuff. I kind of feel that ocean protocol is doing. For data, scientists and developers who work with data. There’s just so many opportunities to start to get involved and, and create valuable data assets, and then leverage the, the ocean market to sell them.
So, yeah, I definitely think it’s, it’s, it’s still just such an early proposition for the market, right? Relatively high learning curve, but once you get through that learning curve, it really starts to open up. And I mean, use, you mentioned that it was only not even two years ago that you started to dive deep into, into this world of, of sort of cryptocurrencies and, and you know, already you’ve started to, to fully grasp the concepts of, of ocean protocol and all those things in your you’re highly active.
So. It’s a pretty, pretty impressive achievement. So the Indian outreach program, as you mentioned, has been active for around three months now, what are some of the things that you’ve been doing to, to introduce ocean protocol, to, to the ECOS?
[00:18:10] Prakash: So the one thing that I have observed that so when I first made the proposal, I did realize that there’s a lot of ambitions in that because there is that I have to mind with it to actually go for something that’s actually scalable or do I apply for something. But I believe I actually wanted to learn a lot about the ocean proposal, why they working on the grant?
So I think that was my intention behind going quarter, really, at least a mid based of sort of a grant. Even at that time. I think that even the first proposals that had a cap of 20 K. But I took the chance and I think I went for the really moderately ambitious claim. I mean, granted, so that, that it gives enough meat enough time to work on it.
I mean, work on, not just a proposal to understand so much more about the ecosystem and what it is all about and what are the products that are being built. So the non-technical. Proposal kind of helped me in that regard. So, and then I started to attend a lot of working groups and to increase my knowledge of what ocean protocol is all about, what the community is all about, because community is also equally important because just with high-tech, it’s not just scale any blockchain network.
I think you have to have a community approach and you have to have a financially sound approach as well. So these things were important and I had a feeling that all three. Boxes with being able to predict an aid. I could recognize that from the ocean proposals, that is not just technical outreach programs, you also focus on comedy arteries and you also have the ambassador programs as well.
So there’s actually the ambassador working group that I joined first to learn more about the ecosystem and they actually increase Laura and especially encouraged me to apply if I really wanted to lead the initiative. And that kind of started me and I had some good guidance from Robin as well, or to be the best way to tackle this.
And so I outlined a few, it does actually lend the proposal if you actually go back and see it was a much lengthier proposal. But in terms of deliverables, they were just at least it could be reduced to at least 6, 7 lines and what I wanted. And I knew that that could be an irritation in terms of physical events, but I still wanted to get some feedback from the community on what the events would be all about.
And so that’s how I kind of got in touch with the community as well in dumps of how the acceptance in terms of discussion of. And not just any network, not just ocean protocol, any network, how the acceptances and the awareness was actually a lot better than I expected. So the, my first one or two months was focusing on acceptable online activity while I go the social media presence.
That was one and the. I also was writing a lot of activities while I continued to set the base for the events aspect as well. So that’s what I kind of contemplated initially started on fundamentals first in terms of how the acceptance could be, and then actually started to focus on the events, sprint that I outlined for him, the end date of Jen.
Right. So even though you physically went to not possible right now, I have kind of three strategies towards community calls with a lot of really interesting communities, have a lot more to impact as compared to physical events. So that’s what I had started. So even now I think it’s very on track to complete by January.
And so that was my initial thesis on having a productive on an activity as well as the. Even something as simple as including a grant hackathon, which was outlined in the proposal, which is what we had proposed to conduct by the end of January.
[00:21:39] Scott: Do you want to maybe just tell us a bit more about the Indian ocean hack and just a few more, few more details.
We’re going to be publishing this around about the time. The Indian ocean hack goes live. So people listening you know, just maybe giving them a bit of an overview of what what’s involved and what they could expect.
[00:22:01] Prakash: Awesome. Yeah. The events that. It has an organic UDaB. It definitely leads towards the one single grantee when it’s called the Indian ocean hack. So the details of it will go, I think it should go live in parallel to this podcast. And I think we have done a good job and having a lot of call to actions from the mini events and workshops that we conducted that it will be in lead up to the final ocean hackathon.
Where it will not just be about our tech deliverables, because a lot of hackathons that is developer heavy, and there’s not a lot of community-based corrections, right? So we have done something really well in dumps off having really well measured activities and bounties to start with, because if you want to start off on a protocol that is not as popular SLA, once a losing layer, two solution, you still have to create.
So that’s the idea behind this hackathon. There’s a lot of achievable quest out there. And in terms of anyone can deep dive, not just from the Indian ecosystem, but it’s also good measure of from the ocean ecosystem as well. I also place it’s not just restricted to Indian it’s from anywhere in the world.
It’s also, I think I have, we have placed some specialty categories for the best overseas steam as well. So I have a good, great hopes for engagement from the ocean DAO community. So that’s the idea. And on that regard, I think I really glad to say that the first apartment bounty should who’s coming from ocean missions.
It’s really nice to see that. And I really loved the quest that they outlined in temps to the best use of iterate, either scan APA. Right? So I think that’s a lot of very accessible for a lot of new developers as well. And it’s really well constructed. And I believe that it’s going to be really exciting to work on that as well.
[00:23:45] Scott: Yeah. Also, I’m very much looking forward to I think it’s going to be, be an exciting opportunity to, to, as you say rally the community building efforts that you’ve been putting in and, and focus all of that energy to. This, this hackathon. And yeah, I mean as, as you sort of touched on the ether scan and API is, has got a wealth of information hiding in there and you know, it’d be great to sort of unleash.
Some skills into to trying to uncover some interesting and useful insights there. So if people are interested in learning more about this we’re gonna add links to it in the show notes. But for someone who’s listening and they wanting to get, you know, more involved with ocean protocol or the Indian outreach program we can get.
[00:24:34] Prakash: I think I think along this podcast description, we should have an article information to learn about ocean protocol in general. Right? So I think ocean academy was my gateway to learning. So even if you don’t intend to finish the course right away, right. I think the, you can still scroll to the materials to understand what it’s all about.
So I think the ocean 1 0 1 is a best way to start just to go through it, what it’s all about. And you can clearly understand, even from a very non-tech perspective, I’m sure you heard the con to that. So ocean one-on-one is a very good place to start so that you don’t get overwhelmed on what’s going on in the GitHub site, on the website, on the marketplace, that it could be.
So I think that academy and ocean Owen is the best way to start on ocean protocol and it dumps with the Indian ocean outreach program. I think. Yeah, we figured that we need not to need to have a website per se, but we just started with a few notions we just to start with, but I believe there’s a really good way to sync stage.
A lot of initiatives we’ve been doing through a GitHub. I mean, sorry, get book that we had just outlet launched sometime back and it gets updated every two, three days. Right? So that’s a really good way to keep track of what we are also doing. And we also had to have a really good discord server.
Absolutely telegram group for builders. Right? So that is also the. So these are the right channels to put to us. And Twitter following has also reached a really good number. In fact, it is gone to a level of outreach that was much more than what we promised in any metrics. So that’s a good way to look at things.
We, we we had sufficient partners to help with that. We also had gone my labs from the cutdown on community. They also had an even even though they may not relate to ocean protocol, but they were really helpful in having a lot of community activities online and offline as well. So that kind of helped us really well and having a really good local outreach and yeah, I’ve had some good partners within an ecosystem as well.
So, but yeah, it’s too much details I mentioned and is going to be present in a deliverable checklist next month or something it’s going to be lot to grasp. So for that purpose, I think from a project management perspective, we are sticking to the GitHub get book process where all you need to do is just scroll to the separate tabs for really interesting.
Decorating. And so the good book is the way to understand what’s going on and then the ocean outreach, and to get more detail, they can help on do associates at any point of time, telegram discuss. These are the best way to reach out. And yeah, I think for the listeners feel free to hop onto the social links and get into, see what’s.
[00:27:11] Scott: Great. And yeah, I, I doubled down on the the ocean academy being such a useful and also reasonable. I know that when I was first trying to wrap my head around ocean going through through the academy was a great way to grasp. The basic concept and there was another ocean DAO initiative.
So really cool to see the community coming together and building some interesting and useful resources there. Hey, well, thank you so much for taking the time to speak with us. It was great to learn about the work that you’ve been doing and, and. And learn about how much activity is happening in India at the moment, and the focus that’s being put on AI and specifically, and then now, you know how that sort of morphing over into the web three and blockchain space.
So yeah, just massive. Thank you for coming and discussing and yeah. Wishing you all the best and looking forward to that.
[00:28:04] Prakash: Cheryl Scotty, thanks a art. You had an amazing post and get me as at ease with the questions that were really, really difficult for anyone to listen to. I think you had a lot to discuss, but we also had, it was nice to see that we had the breakdown from your perspective as well. They do really good to have the session and looking forward to more as well.