During this year’s Consensus: Singapore 2018 conference, CryptoGlobe caught up with Dorjee Sun, project manager at Perlin, “the first practical, trustless and decentralized cloud computing marketplace,” that’s set to help up to one billion people around the globe have access to the crypto economy.
According to Sun, Perlin is set to help people earn up to $170 per year by simply using their mobile phones to power the supercomputer and make it globally accessible. During the interview he detailed how the organization is planning to reach its goals.
Among the details, Sun revealed Perlin is in talks with major organizations throughout the world, and that the cryptoasset’s blockchain is currently able to power “about 10,000 transactions per second.”
Perlin, according to its founder’s words, is set to offer cheap cloud computing services to clients, which could see it compete with e-commerce giant Amazon and its Amazon Web Services (AWS) offering. Below is the full interview, edited for clarity, consistency, and style.
Introducing the Project
Dorjee Sun: So we finished up our private sale, and our ledger. It's the first Avalanche consensus mechanism inspired by Avalanche, but it's now being changed in about 17-18 ways. That’s a big deal, because there are traditionally two fields of how to do these consensus mechanisms. And this is a new field that was put forward by the Cornell professor, Emin Gün Sirer.
If you look at the data, probably in two weeks/ by the end of this month, we'll have the finalized numbers. So we're about 10,000 transactions per second, which is still pretty fast.
CG: But the website says 1,600.
Dojee Sun: Yes, but we’ve really improved the performance. That's a big milestone that we've finished the visualizations of. We’re just really confirming the speeds, and the test should be done in about a month - it actually is mostly finished, but just, you know, fine cleaning, stabilizing, tightening it up.
Just to take a step back, our vision is that we will bring a billion plus people from lower per capita GDP to earn money by putting their phones and laptops into our distributed cloud, and then earning money.
Your compute power on this phone could actually be rented out when you're sleeping at night to do machine learning - AI computations - that would be taking money from say Amazon Web Services or Google Cloud. The idea behind this is, if you think about the bottom billion, they're on $1 a day. That phone would probably generate about $170 per year, if you were renting out the compute power for eight hours a night.
On top of that they would have web connectivity, they would have education they would have access to remittances through our token Perlin’s PERLs, [which] could be used as a store of value or transfer of value. That's really our vision. We are the first technically appropriate use case we see that can create earning. Earning, we think, is the killer app, because what do these one billion people need? They need money.
These are previously free distributed computing platforms where I can donate my computer at night, holding it at home, to solve cancer proteins, or [help] NASA Search for Extraterrestrial Intelligence. That's the vision. Coming back to the milestones, we've received funding from 200 strategic and institutional investors - Bitmain, FBG, a whole bunch of them. Also, big chat apps like a Kakao.
CG: So you would integrate with existing apps?
Dorjee Sun: Correct. We looked at the full mobile stack and laptops, and we thought ‘okay, so we’re meeting with HTC today, will we be able to sign? Who knows.’ Besides, they only have X number of devices. Then there’s browsers, there’s games, there’s chat apps, etc. All of them will have their pros and cons.
We're still trying to find access points. Another access point is we've signed an MOU [Memorandum of Understanding] with the Prime Minister's Office of India, and we're in negotiations with telecoms, Bharti Aitel and Geo Telecom.
We are hopefully signing this week with Telkomsel which has 180 million users in Indonesia. That's the state-based telecom. I was with the Minister for Bangladesh, ITC, last week, and also with the Myanmar Minister. We’re [also] in negotiations with the new telecom providers in Nepal and Turkey. Right now, our coverage is about 2 billion people – in India, Indonesia, Bangladesh, Nepal, Turkey, Myanmar, Singapore,
CG: Do you have plans to focus on China?
Dorjee Sun: We definitely have a lot of support there, we think it's a hot market to penetrate. Fundamentally, every blockchain company has to look at customer acquisition. I don't know if you've ever looked at app adoption, but you what's the cost of acquiring a user? Do they stay on? Do they leave? What's the virality coefficient? Did they tell other people, what's the lifetime value of that client?
We've done mathematics or where we think is the more accessible market, [and] China's just a tough one. We are talking to companies, but we haven't actually inked anything yet
CG: From a client perspective: let’s say I’m looking to perform some cloud computing, what kind of savings do you expect?
Dorjee Sun: We can do 50%-80% cheaper. Right now, you've got Jeff Bezos: he has to set up a server farm, buy all of the computers, network them together and have stuff operated. That's how he creates AWS [Amazon Web Services].
Whereas we go to the poorest people who have devices and it's up to them. It's like the equivalent of Airbnb or Uber, Uber X. We believe that the economic incentives that we can provide would mean that the marginal costs of bringing on people in the developing world is very low. So we believe we can offer 50 to 80% cheaper
CG: With the same efficiency, speed, and reliability?
Dorjee Sun: Right now we’re testing use cases. If I'm a high-frequency trader that's running, very time conscious, sensitive computations, I probably wouldn't be using necessarily what we're doing., We've got 15 partnerships confidentially, which we haven't published yet. For example, I met with big infrastructure companies: Exxon Mobil, Shell, guys like that.
They have huge computational requirements for predictive maintenance. Every hour their refinery is sending data, and their cycles are like one week to two weeks [apart]. Something like that is a no brainer. Plus they get to redirect the money to the local population, instead of sending the money to a multinational, trillion-dollar, billionaire-owned, Seattle-based company. So it's a CSR activity
CG: How flexible do you image DApps (decentralized applications) on Perlin being? Do you build anything you want, or are there limitations?
Dorjee Sun: We have a 10-year roadmap, and it’s always about being in the right place at the right time. We're quite patient in terms of when we think the supply and demand are going to mature. We are very confident because the demand for AI compute is doubling every 3.5 months, according to the CTO of Open AI.
CG: So I’ve taken a look at the whitepaper, has Avalanche been tested and peer-reviewed?
Dorjee Sun: I believe it was peer-reviewed, but you’d have to check with professor Gün Sirer. It was announced in such a spontaneous fashion that maybe it wasn’t, I'm not 100% sure whether it was peer-reviewed. But actually, our ledger is not 100% Avalanche. It is inspired by Avalanche. We’re currently undergoing the security audits and the reviews.
CG: OK. So how do you see the economics of PERLs playing out, with users generating a specific amount of PERLs and an incentive to liquidate to a local currency? From an investor’s perspective, do you see growth in value of these tokens?
Dorjee Sun: Absolutely. With all of these fixed supply tokens, the question is really like: will the demand for those tokens outstrip the supply? The way we look at it is, the addressable market for the cloud space is $300 to $500 billion by 2021, 2022 and 2023. So let's just say we’re able to get 1% to go decentralized. That would equate to $3 billion to $5 billion in demand. You could achieve a price per tokens that’s of a few dollars.
If you think about Bangladeshi workers in Singapore, they currently pay Western Union 15% for remittance. Actually, I think PERL tokens could end up being a store of value, or at least a transfer of value. If that's the case, we think that the token velocity will be much lower than Bitcoin, which is currently being high-frequency traded. Let's just say the token velocity is one or two, then you still going to have a token price of a roughly $1.5 to $3, something like that.
CG: From a security perspective, is all data encrypted?
Dorjee Sun: Yep. There's been a few academic papers, I'll get my CTO to send them to you. It's on the latest cryptography. Last year, there were a few papers in crypto in Asia that we based our privacy on, so in some ways think that it's safer than centralized cloud services which have been compromised by government agencies.
CG: So you can execute smart contracts, you can create DApps, it should be more scalable and faster than Ethereum – do you see the ICO market moving towards Perlin?
Dorjee Sun: It's a really interesting question. To be honest, we have not been actively promoting that, but it's a really good idea. I think that what most of the readers for CryptoGlobe would be understanding of is the wealth accumulation opportunity.
One of our pilots: if you think about me being a farmer in Indonesia and taking 10% of my income in PERLs, and let's just say PERLs appreciate significantly because of network usage, and then I decide to invest in other ICOs, suddenly a farmer has the ability to effectively participate in a growing asset class, which is the greatest form of social mobility, that currently is just not an offer for them.
If you're a farming Indonesia you have two choices, either plant something or cut down something and sell There's no ability to invest into Facebook shares or other forms of wealth accumulation. So the short answer is we have not actively pursued ICOs it's just because I think that regulation is going to change things.
When you raise it, I think it would be very powerful as a proposition if developing countries could also participate in wealth accumulation. Obviously, they’re not very sophisticated so you need to make sure they don’t get scammed. I mean just imaged that: ICOs for the billion. That’s pretty cool.
Wealth Accumulation Opportunities
CG: I think the key point, which I didn't really appreciate from checking out the website is the fact that you're essentially allowing something like universal basic income.
Dorjee Sun: But just think about where we see it. And this is also like, when Uber or Instagram nails the iPhone as being the device, which would then save their photos/GPS, location-based apps into stratospheric use cases. we're predicting ubiquity of web, you know renewable energy, ubiquity of energy, and then cell phones.
The greatest incentive, I think, for people who are poor is: children, health, education, you know, betterment. We did a call with the head of Telenor in Bangladesh today. What if you could provide them with free tele health services, right? So they've got health, some income access, then what about education? What if they could get educational components?
It's really “how do you create that type of platform” as opposed to just: "we have this side chain, blah, blah, blah interoperable, yada, yada, yada," which, you know, really could use adoption. But if I can say to everyone, hey, earn $170 a year and get a free phone, and all you have to do is plug it in every night. I think that that's a really rapid adoption.
CG: As for the mainnet launch, do you have an estimated launch day?
Dorjee Sun: February of next year is what we say externally, but we hope to hit testing in January. What we've inked with most of our partners is a 12-month adoption. The idea of actually have devices earnings in 12 months, that’s our goal.
CG: So you aim to be embedded in the devices? You wouldn’t download an app?
Dorjee Sun: We're trying to do devices, we're looking at browsers, we're looking at apps. My line in Indonesia has 40 million users. We talked to them about trying a pilot - we just looked at any access point, but obviously the easiest is direct download.
CG: So for people looking to invest, do you when the public sale is?
Dorjee Sun: Most likely it's reliant on the market. We're not strapped for cash. We are really just omniscient of, I think, right now it's such a negative environment, and that's why we haven't announced a lot of our successes. It’s kind of like shouting off a cliff, like throwing eggs against a brick wall.
We think that unless there's some stabilizing, mild upward trajectory towards the end of the year, for a public sale we're probably thinking two months but we're not in a hurry to rush it out because all our investors understand that it's a long hold.
CG: Final question: Can you summarize where you hope to see Perlin in 10 years?
Dorjee Sun: In 10 years time, Charlie, mark my words: I'll be 51 and we'll have one billion plus people generating some income and hopefully utilizing that tool to provide education and opportunity. Low cost remittances, store value transfers across the most underprivileged nations. Effectively also hopefully providing the younger market with devices as well so a billion users into this ecosystem. It's gonna happen.