Putting video distribution on the blockchain
Welcome back to What’s NEXT, a podcast from Samsung NEXT exploring the future of technology. In this episode I speak with Sliver.TV and Theta Lab’s Mitch Liu and Wes Levitt about how their Theta token incentivizes users to share their bandwidth with others.
Ryan Lawler: Thanks for joining us guys.
Mitch Liu: Thanks for having us Ryan.
Ryan Lawler: I guess to start, maybe tell me about how you got start don’t worry Sliver.TV, and what it is? What the idea was behind it?
Mitch Liu: Sure. A little bit about my background. I was a techy nerd by training I guess. I went other MIT for my undergrad. I did research at MIT media lab working on all sorts of interesting problems like non-linear movies, interactive storytelling, human interactions I’ve been in the video space for a long time. Towards the last 2015, we started Sliver.TV at the heels of Amazon acquiring Twitch, and we saw an opportunity to really … We saw Twitch being a very static, sort of non-interactive one-way, one too many kinds of streamer to viewers in the ESports space. We saw an opportunity to be able to be a sort of next generation Twitch, which is much more interactive, and much more engaging.
Sliver started really at the intersection of virtual reality VR and e-sports in early 2016. Since then, we’ve launched the Sliver.TV watch and win platform that now has about 700 million monthly users, and the THETA token is our blockchain project that we started mid of last year. Mid of 2017.
Ryan Lawler: Okay. You have this sort of peer-to-peer way of transferring video, and that’s what THETA is a part. Can you talk a little bit more about that?
Wes Levitt: Yeah. THETA is a mesh streaming network for delivering both VR content, and standard 2D content. There’s a couple of different ways that can be implemented, but the primary one that we are beginning with utilizes WebRTC, which is an Open Source Google protocol, and the really interesting thing that allows us to do is establish browser to browser connections between users. There’s no client download, there’s no browser extension to download, and that gets us across one of the really issues with previous peer-to-peer networks is it’s very difficult to get a user, especially a mainstream user, to jump to that extra hoop. It needs to feel just as easy as they’re experience going on YouTube now. Very seamless.
That’s one of the innovations that lets THETA have a chance to get that mainstream adoption.
Ryan Lawler: Okay. Why did you see an opportunity in this space? What was it that you said, “This is missing from today’s streaming platforms,” that you really went to make available to users and those who are watching these streams?
Mitch Liu: Yeah. For instance, in the e-sports space, when you’re streaming popular e-sports games or titles like League of Legends, Counter Strike they’re all played on the PC screen. It’s a flat, 2D game. With the first phase of technology that we developed at Sliver, a number of patents that essentially rendered these PC games in a full 360 VR environment. You’re able to put on your VR headset and it’s sort of putting yourself right in the middle of the game. You can look around and see all the characters running around you, and it’s as if you’re inside a game.
That was a very unique technology, and we did livestream of those tournaments in real time, and that really sort of took that level of interactivity and engagement to the next level beyond what you would see on a regular 2D stream.
Ryan Lawler: Right. What does the mean from a bandwidth perspective, and just being able to render that in a 3D space as opposed to a 2D space, and render it for a VR environment as opposed to just watching it on a desktop screen?
Mitch Liu: Yeah. Huge bandwidth implications if you compare 720p or event in 1080 HD resolution, for a VR stream you’re looking at approximately 4K per I, that’s 10 to 15 times amount of data that needs to be pushed through for the equivalent amount of time. That was a lot of learnings back in 2016 and 17 that kind of led us towards THETA as a blockchain project. It’s really that particular is the fact that over time not only these videos of form factor increasing as a percentage of total internet data, but also that amount of data that’s being pushed through from 4K to VR content, and that’s only increasing.
Ryan Lawler: Yeah. I was going to ask about that because I have been covering the video space for a really long time, probably 12 years and remember there was this period of time in the 2006 to 2008 time frame where everybody recognized that video bandwidth was going to explode, and peer-to-peer seemed to be the answer, there were all these different peer-to-peer projects that popped up over the course of time, and they all disappeared by say 2010, 2011. What’s unique about the environment today? Is it just that WebRTC exists?
Wes Levitt: It’s personally that in that WebRTC is a relatively new development, but that’s actually a good segway into the second component that THETA brings to the table is the fact that it integrates a native blockchain that we’re building into it. Blockchain by itself doesn’t help video delivery be any more efficient, but what it does do is solve two of the key problems that those previous peer-to-peer networks you were talking about had. The first being that it’s very hard to get users to actually share out of their goodness of their own heart on the peer-to-peer network. You need some way to incentivize them to do so, and being able to introduce a token that shares that value created with the users that share their bandwidth allows you to incentivize that.
And the second is that video platforms, content creators, they all were left with a very bad taste in their mouth from the first generation peer-to-peer networks in that they were used for 90% for piracy. Digital rights management is a really important issue for these platforms, and being able to track each video segment that’s shared through our peer-to-peer network via this blockchain ledger allows us to introduce guarantees to video platforms to adopt THETA that their content will not be shared over our network with someone who doesn’t actually have the authorized use of it.
Ryan Lawler: What was the driving force behind adopting the blockchain technology? When did you say, “This is an application for blockchain that makes sense?”
Mitch Liu: Yes. This happened last spring right after raising our financing, our lead investor, looked at operating expenses and they saw that 30 to 50% of our monthly cost was spending in video delivery. This is CDN. Content delivery networks like Akamai and Amazon Cloud Front, and a bunch of networks that we used to deliver particularly our live streams to users around the world in developing countries like Brazil, and Eastern Europe, and Russia is really hard too and very costly to deliver those streams in real time.
They asked me, “Mitch, have you thought about solving this problem using blockchain?” They were early investors in Ethereum, and Rave, and definitely a bunch of blockchain projects. That was really solving our own problem if you will as we did the why paper and the project, we saw this actually a huge problem. It’s not just us, but just about every media, and content provider, and video platform out there has similar issues. And the main guiding factor for us is THETA is an infrastructure. The go-to-market strategy is to partner with an existing content provider. We’re not trying to be the next YouTube. We’re not trying to be the next HBO or Netflix. In fact, we’re looking to partner with Twitch, YouTube, HBO.
If you like Game of Thrones, and you Game of Thrones through HBO’s app or website or what not, and you share that with the community whether it’s your friends or other people in the network you earn the incentive for doing so. It’s really a very different shade of model.
Ryan Lawler: Right. If I’m a Samsung VR user/viewer tuned into some of those content whether it’s on my desktop or driven through my desktop, or whether it’s driven through your VR headset, is this something I have to opt into to be able to get those streams? And what does that mean from the user perspective?
Mitch Liu: Yeah. That’s a good question. The tests channeled out, we’re in the process of building, is very, very clear to the end user that you are opting into parts of spending, as part of the community, and parts of sharing this particular content. If you don’t want to share, right click here and you’ll be disabled, and you won’t be doing so. That’s really important. That’s really part of the user experience, and if they do choose to be a part of the sharing network, if you will, they know that they’re earning THETA tokens as a result of that.
Again, it’s sort of that messaging that’s really important because we want to be really clear, really transparent. It’s sort of like the sharing economy. With Uber, you’re sharing a car. With Airbnb, you’re sharing your apartment or your condo. With THETA, you’re sharing your excess bandwidth.
Ryan Lawler: Right. Let’s talk about those incentives and how they come in, and what it means to be a part of the THETA network, and what it means to get tokens. What was the thinking behind that even just issuing a token?
Mitch Liu: Sure. The problem space that we started looking at last spring, last summer is why the past peer-to-peer networks failed or it didn’t get the kind of traction that they had expected, and it really sort of came down to the incentive structure. The crypto-economics of that. Blockchain came in as a technology that solves that particularly because you need to make micropayments, and micropayments, it doesn’t make sense to pay someone 1/1000th of a penny. How does that even track in the real world? But with blockchain and cryptocurrency you do in fact. You can have one millionth of a penny. Whatever makes sense.
We can uniquely identify that two or five or seven second video segment and say, “There’s a resource ID associated what that, and there is a micropayment associated with that.” Now, the harder problem is how do you scale that? At any given time if you have a million concurrent viewers of a video stream, you’re talking about maybe 10x or 50x in terms of the number of transactions needed to process all that. That’s a lot of the hard problems that we’re working on no.
Wes Levitt: To add to that, one of the critical reasons why we thought the best way to do this was to issue a token, and use that as a way to transfer value within the system is that while this begins as a protocol development by THETA labs and with a couple of key partners, the end game is that we essentially get this network to a robust consistent state, and then it’s operating at a decentralized manner where THETA labs no longer has control of it. There’s different stakeholders that all have varying control whether it’s THETA, Samsung VR, future video partners, the viewers themselves, content creators. But we don’t want the end game to be centralized THETA labs doling out rewards to everyone who contributes to the network. We want it to be done via protocol in a trustless way, and for that to happen you need to have something to carry value throughout the network, and that’s the THETA token.
Ryan Lawler: Gotcha. You’ve recently sort of a test net of how this all works. Talk me through that, and what that actually means. What is the THETA test net?
Wes Levitt: Sure. By test net, that’s kind of the parlay in blockchain, but you can think of it as a beta release essentially. A public beta that users can start testing on. What that meant specifically for THETA and for SLIVERtv was the test net THETA blockchain is live. There’s a block explorer, which is essentially a visual interface where you can see all the blocks being produced on the THETA blockchain, and the transactions on that, and it’s incorporated into SLIVERtv in the form of a THETA test channel. It logs onto watch the ESport streams on that channel if they opt in. It immediately begins sharing bandwidth just like we talked about without any kind of download just by clicking onto that channel.
There’s some instructions as to what’s going on in the background for them, they can click on to their specific THETA wallet and follow the transactions through the THETA blockchain and see exactly who they’re sharing with, and the transactions associated with that. It’s kind of a teaching tool in some ways for how it works. It’s our testing grounds essentially as we move forward to main net launch in Q4. As of the first month, we had 300,000 unique users that had visited from 149 different countries sharing their bandwidth. It’s been invaluable as data for our engineering team to see where the blockchain could be further optimized, where their streaming protocol could be optimized. Essentially get the kinks out of it before we move to the full production launch in Q4.
Ryan Lawler: Any big learnings from that beta launch or things that maybe surprised you once you put this out in the world and test it out?
Mitch Liu: Yeah. We’re working through a bunch of issues, and that’s sort of the goal and the purpose of the test and the beta. One of them being that the messaging to the end user and the end user experience, I guess in blockchain lingo going to the guts of it, we are moving towards a two-token system. THETA is a staking token with something called GAMMA, which is a functional or utility token. Explaining that to the end user is change because it’s hard to … For adoption purposes, the end user when they hit that particular player or video stream, the just want to know what they’re watching the content they love, and that they’re sharing that content with someone that is in the network.
They don’t necessarily need to know that there is a two-token system, and they don’t necessarily need to know the guts of it. Just saying when you use a Visa card, you don’t have to understand exactly what happens behind the scenes. Messaging that and explaining that to the end users is something that we’re continuing to improve, and that’s really important because as we launch with our future partners, other areas I think our blockchain explorer, which is a tool if you will, run into scalability issues. As you can imagine, over 300,000 users over a course of three, four weeks with very large session times. 3x average session times compared to our regular channel on Sliver. That was really pushing the boundaries of the number of transactions per second, and so the tool scaling that is something that we’re working on now as well.
Ryan Lawler: Just curious, what do you see as driving the session times? Why are they so dramatically longer? Is it just that people are really engaged in the content or is it that maybe because they see some value exchange, they’re more willing or more interested in capturing some of that?
Mitch Liu: Yeah. That’s a really good question. It’s something that we were a bit surprised of it as well because we had expected it to be similar to our 10, 20% variance from our other channels. What we found was that the feedback from the initial side of users that they love the fact that they were seeing this little graph in terms of how much they were sharing with others network. It almost became a bit of an obsession in terms of how much tokens they were earning. It became sort of this leaderboard kind of a thing that you wanted to. In chat, you were saying, “I shared 23 tokens worth with the community. How much did you share of?” Being recognized. That recognition is kind of interesting in itself.
Secondarily, I think the incentive, the fact that you can actually do something with those tokens that have certain value as you described. A value exchange made it really. That they had to come back tomorrow because they needed to get to that 20 threshold in order to be able to redeem for sort of like that credit card points or airline mileage kind of concept. It’s a very, very interesting sort of social dynamic.
Ryan Lawler: Do you have to worry about users giving the system because that’s a way for them to sort of rack up some sort of currency?
Mitch Liu: Sure. It’s an interesting question in itself. In some ways, if a user is telling their friends that, “Holy smokes, you can watch this piece of content, which by the way you might like, and at the same time earn something as a result of that.” I think that’s pretty interesting in itself because it essentially increases the user … It’s a user acquisition tool in some sense. If you have a free site anyway, wouldn’t someone like a YouTube or Twitch or Samsung VR want to double their viewership for their content, that’s pretty interesting. Those questions become really interesting in themselves because it benefits the network, but I think the even more interesting sort of second half of that is what do the users do with the tokens?
Mitch Liu: This is probably the most exciting part in our work with our Samsung VR team is is there an opportunity to start monetizing to build a new revenue model on sort of this blockchain approach? And what if you’ve earned those 30 or 50 THETA tokens, and now there’s a popup that says, “Hey, would you like to redeem that for a premium? For episodes two, three or four? Would you want to use that disable video ads? Right? No one really likes ads all that much. Or if you’re in Netflix, would you want to use that to offset or to pay for your next month’s subscription?” Those are new revenue models is where it gets really interesting.
Ryan Lawler: What do you see as the five or 10 year vision of the company? Where does this go?
Mitch Liu: Yeah. We always look to blockchain as a three to five year vision in terms of the technology adoption. I think we’re very early. It really feels like mobile 2010 or internet 1997. From our perspective, we think that THETA as sort of this infrastructure play where we’re looking to partner with. We’d love to get to a day where you’re watching your favorite YouTuber, and you are rebroadcasting that particular content, and you’re earning tokens, and you then head over to HBO, and then you’re able to use those tokens to unlock the next episode of some premium content that you love, and then accumulate enough tokens the next day to be able to pay for your Netflix subscription for free, and that’s a really … And to be able to do that seamlessly.
Your digital wallet if you will just travels with you, and you don’t necessarily have to think about it. It’s almost the concept of single sign-in that those tokens are just with you. All these different sites. That’s really popular, and that’s really interesting for us. But it’s much more than just video. Our vision is not just video infrastructure network in five or 10 years. We want to move into all sorts of content deliver. We envision apps, mobile apps. Any kind of app or even PC based apps. There’s a lot of app updates that happens all the time, and especially in mobile games. These are hundreds of megabytes or larger files. What if you could in fact instead of each person pulling that big file from a central server, which is what happens today at huge cost to game developers or platforms, what if you could pull those files because an update is an update, right?
Pulling bits and pieces of those files from other mobile users around you makes a lot of sense. What about enterprise video conferencing? What about file sharing? What about static website data, which comprises most of data out there today anyway as far as CDNS are concerned. There’s all sorts of decentralized was of sharing content, not just videos as a form factor.
Ryan Lawler: We’re recording this in mid-August, and at least in the crypto-exchange is a lot of those tokens have taken a dive over the last several months. I’m curious. What does that mean to you in general and what do you think it means for the broader idea of the blockchain industry for lack of a better term?
Mitch Liu: Yeah. I think from an investor standpoint there’s a lot of this sort of concern over the day-to-day or even the hour to hour prices, and when you have a token for a project, and it’s being traded it’s almost feels like that you’re running a public company instead of being a stock market being open from 9:00 to 5:00 or whatever the case may be. This is like 24 hours a day, seven days a week. It gets kind of crazy. From our perspective, by nature our DNA we’re technologists. We are developing a technology leveraging blockchain, and it’s not so much a day-to-day.
Sometimes we look at the prices, and to be honest we don’t really look at it that much because in our view it’s about the real world adoption. It’s about make sure our technology scales, and it launches in December this year to make sure that we have those first three, five, 10 launch partners, which are large platforms. Consumer platforms, and making sure that it scales appropriately. But there’s a lot of distraction, and to be honest I think this whole industry is sort of in the early days feels like a wild, wild west if you will. But I think over time I think you’ll see a lot of changes in the next six to 12 months. I think you’ll see a lot of great projects sort of floating up to the top, and we hope THETA will be one of them.
But I think at the end of the day it’s really about end user adoption that really makes all the difference.
Wes Levitt: Yeah. It depends a lot on who you ask because there’s a lot of projects where at the core their token being worth a lot of money, and very quickly is at the heart of whether the project succeeds or not. We have a very different perspective for a couple of different reasons. One being the people that participated in the tokens or our credit investors and Venture Capital funds, our media conglomerates like BDMI and such, and the majority of the tokens where sold to the existing equity investors of Sliver.TV. There’s very much aligned incentives, and no one is looking at this as a day trading token play.
And because there’s so many of the corporate partners that we’re speaking to and we’re looking to partner with that Mitch referred to are not looking at the daily token prices. In fact, it’s something that we’ve talked to and we’re at various stages of program with, they couldn’t tell you what the price of any of these tokens are. They’re focused on the technology and what it can do for its end users. We don’t feel the pressure that I think some of these projects do to be subjected to the day-to-day swings of these markets.
Ryan Lawler: All right. One question I’d like to as everybody is what’s one controversial opinion you have that is really strongly held?
Mitch Liu: I think one area related to what we do is decentralization. If you think about blockchain as about empowering the end users. It’s a lack of a centralized authority, and I think there’s some controversy there. You have a lot of organizations or central government bodies that see that as a threat, and is that necessarily a good thing or a bad thing. Even from a monetary standpoint you see a lot of governments and countries looking at whether Bitcoin or Ethereum or what not as a threat to their centralized currency. That’s interesting in my view. Honestly, I think that in the end when it becomes truly decentralized and is driven by end users, and that power is distributed, and it’s governed by this body that’s sort of highly decentralized in the end, and you’re in a much better place than a centralized authority.
I think the path of getting there is the challenging part. It’s not a lot of obstacles and pain along the way, but once you get there, it’s something that I think will be sort of the next platform, if you will, of building technology over the next 30, 50 years.
Ryan Lawler: I want to press on that a little bit because I’m kind of curious. There’s a lot of optimism in the tech space about this idea of decentralized infrastructure, and decentralized applications in many ways sort of returning back to the early democratization of the internet before it got sort of taken over by these huge platforms. How do we make sure we don’t screw it up again because the idea of a decentralized internet or place where anyone can exchange their ideas or share applications or whatever is great. But at the end of the day, that platform is now monopolized by Facebook, and Google, and some of these other companies.
Mitch Liu: Yeah. I think the way we want to get there is not necessarily to say we’re going to be able to next decentralize YouTube or Amazon or Facebook or Google, and we’re going to kill Google. Look, no one is going to kill Google anytime soon. You can come and try. The best way to get there, and I think that’s our view is sort of as a building block we want to work with the existing incumbents being the Googles, and the Facebooks, the Amazons or the world, and you use that as a path towards decentralization because I think in the long term, the Amazons, the Facebooks will see the value of their own services and products and whatnot being decentralized, and bi powered by, being democratized.
It’s almost a path forward, if you will, that is not necessarily saying that we’re going to kill Amazon, and that’s the part we’re excited about, and I think that’s the path to get there in 10, 20 years.I think the Amazons and the Facebooks will look very different in 20 years.
Ryan Lawler: I think the other interesting part of this is just this idea that in 10 years no one is going to know that an app is decentralized. In the same way no one thinks about something is being mobile or on the internet.
Mitch Liu: I think that’s kind of what we’re kind of trying get to. We think media entertainment is a really easy, onboarding towards leveraging cryptocurrency and blockchain. It’s something you do, something you consume, but you necessarily have to think about the fact that those tokens are powered by this cryptocurrency. All you know is that those tokens have a certain value.
Ryan Lawler: Right. Last question, if you weren’t doing this, if you weren’t doing THETA and Sliver.TV, what would you be working on? What’s an area of interest that falls outside of video streaming and blockchain?
Mitch Liu: AI. Artificial intelligence. I think there’s so many exciting things that’s happening in that whole domain. When I was back in school doing movies, interactive cinema types of things, and at that point looking at human interactions, but an element of that was sort of the early incarnations of AI. When you are thinking about how do you build a non-linear path, a scene which is constructed in real time based on human input, there is a lot of elements there in terms of how you think about building that, and I think AI has huge implications in a bunch of different areas, a bunch of different industries.
Ryan Lawler: All right. Cool. Thanks a lot of being here guys. I really appreciate for you taking the time to be part of our podcast.
Mitch Liu: Yeah. Thank you.
Wes Levitt: Yeah. Thanks for having us Ryan.