Subscribe
Share
Share
Embed
Learning a new language can be challenging. Brett chats to Bryan Riester from Univoice about how they have gamified the learning experience. Univoice provides users with a library of music along with lyrics and translation.
What does it take to found a globally important company in these times? We’re interested in what happens before universally-acknowledged success.
Join Brett Kistler as he engages in deep conversations with business leaders from emerging markets, being vulnerable about their experience in the early- to median-stage moments of their founding journey.
Intro: There are always growing pains. There are always things you kind of need to learn rapidly how to do, and when you are the executive of the company, your job is just to do it. Sometimes you are not going to find somebody. You are not going to be able to delegate. Your job is to do that, but on the days when it can't happen, it can't happen. You are the one who needs to stand up.
Brett: So everybody, welcome back to the ClearView podcast. Today I am speaking with Brian Riester of Univoice. How are you doing today, Brian?
Brian: Excellent. Riester, Easter with an R. Fantastic weekend. I have had an absolute blast. Getting a lot done with Univoice. We are expanding our library like a rocket, and I would love to tell you about it.
Brett: Let's start with what Univoice is. What do you guys do?
Brian: Have you ever tried to learn a language?
Brett: I have.
Brian: Tell me how it went. What language?
Brett: I started by learning Spanish. I was about 19 or 20, and I was about to move to Mexico. I went down there for my brother´s wedding in 2007, and I was just like woah, I could work remotely from here and live on the beach instead of Cleveland, Ohio. I was like all right, I am going to move to Mexico. Then a couple of months before that date I had set for myself, I was like okay, time to start learning Spanish. I downloaded Rosetta Stone and started doing Rosetta Stone.
But I knew that wasn't going to be enough. I just wanted to immerse myself in it, so I started listening to a lot of Spanish music. I would listen to Mana, Shakira, Juanes, whatever I could find, and then I would look up the lyrics and then basically try to learn all of the lyrics to one song and all of the grammar to one song so I could hear one song at the speed they were singing it and understand it fully.
Brian: Amazing. How did that work out? How did it work to use music in your learning?
Brett: I mean it worked really well because it includes everything. It wasn´t just learning the vocabulary. It wasn´t just learning things that are useful for travelers, but it was also getting some of the soul and essence of the language and at the cadence and rhythm people might speak it at.
Brian: Something real, right?
Brett: Including some of the idioms and poetic ways that people speak in other languages that don´t map directly on to your own.
Brian: I feel like that´s the most critical part of language. I feel like the majority of our communication is unspoken, and it is the interplay between those words, the idioms, the turns of phrase is really where we create meaning. To answer your original question, Univision is the company you could have found back in that day that would just give you a library of music along with its lyrics and its translation so that you can just sing some karaoke while you are learning. Just let us know what language you are learning, what kind of music you like, and rock out.
Brett: Oh wow, so that's basically what you guys do. Language learning through song.
Brian: Yeah.
Brett: How have you gamified it?
Brian: We are in the middle of launching that system right now. We are coding it up. We are not going to do that without a big library to use. Initially, when we are just putting in songs, which is pretty difficult because we do need it translated into all of the languages that we offer because this is an omnidirectional product, any language to any other language, so it does take a lot of leg work upfront. While we are doing that, we are perfecting our gamification system in the background, and that is primarily about helping people access more content and stay motivated over time.
It is about making sure that this is a product that´is accessible to the whole world. We don't want to limit ourselves like most people in the industry to only the people who are going to pay ahead of time. We really want to make an impact. Gamification is something that is relatively new that is allowing all of these different companies, not just pure game companies, but education companies and other people like us to create products that are accessible to a ramp of unpaid to paid users. It is much more attuned to the modern philosophy where some people can be patrons of the arts, and there are some users who can support us and support that vision in order to serve everybody who needs to be served.
Brett: Tell me how that ramp works. How does a user who just wants to start playing with this gamified musical language learning application, how does their journey go before they get to the point where they are paying?
Brian: They can pay at any point or they can never. That's really up to the user. If you are going to maximize your experience, you need to do the things that are conducive to actual language learning. Here's what I mean by that, and this is where I think Univoice as a company is one of the best places I have ever worked.
In the normal industry of just pure gaming, the idea is to hook people and to maximize engagement and to pull as much as out of pockets as you can. We have gotten really good with the psychological sciences at targeting specific levels of motivation, some different types of motivation. We can use that to a productive end. It doesn’t need to be predatory.
With Univoice, we are really deliberate about creating our gamification systems around what´s the optimal amount of time per day that somebody needs to practice in order to learn. At what point is it just absolutely overkill? We don't want anybody to burn out. We want them to actually be able to learn the language. At what point, are you maintaining your engagement? We can´t let somebody be bored anymore than we can overchallenge somebody. Maintaining a good ramp inside the flow state of your psychology is a part of the gamification system.
There is a difference between mastery and exploration. You do want to maximize for your vocabulary, but at the same time you want to practice, practice and practice so you can perfect your pronunciation. We are very careful about measuring and defining these paths. Maximizing your points means doing the things that optimize your learning.
Brett: What are some of those things that are unique to Univoice?
Brian: It would be something like the amount of time that is spent. Most research is pointing between half an hour and 45 minutes is on a daily cadence something that´s going to be very, very productive in comparison to two hours, where you have a very rapidly diminishing return on the time spent.
Brett: Interesting. I am curious about that diminishment of the time thing. Because if you move to a foreign country, you are probably going to run into people speaking it much more than 45 minutes a day.
Brian: Absolutely. We are not really talking about trying to absolutely maximize the speed at which you can learn the language. We are really trying to maximize the chance that you actually will if you understand my meaning as well as retention over time. You don't want to lose the language. If somebody wants to pick up a language as quickly as they possibly can, then do what you did. Just throw yourself into it. Sink or swim. That's always going to be the fastest way to learn something, but that's not the use case that we necessarily want to focus on. That's probably the fewest number of people that are actually going to get through it.
Brett: But in the middle of that bell curve, you are going to have people who are willing to spend 45 minutes a day, and if they do two hours a day for a week, they are going to burn out.
Brian: Absolutely, but if we can take somebody who does not have necessarily the right habits or the right intrinsic motivation to do that half hour a day, then we can kind of push them over the edge and just keep that carrot on the stick. That's what they are signing up for. That´s explicitly what somebody wants out of us, so we are proud to provide it.
Brett: So tell me a little bit more about the methodology behind this gamification. What are the things that you measure? Are you measuring users while they are using the app and then their overall engagement and then kind of like A/B testing various techniques?
Brian: I guess it comes down to our patented speech engine. We are uniquely able to help people understand the efficacy of their pronunciation. In context of actual speech, different words are dramatically different than one another or they are homonyms, and there is some space in between. The sort of phonetic adjacency is what lets the native speakers understand somebody who has a deep drawl or a trill, or some accent that would otherwise make them very difficult to understand. If I am talking to you about footwear, then that context is going to tell you that when I say boat, I probably mean boot. I am probably not talking about a sailing ship. That phonetic adjacency is something that computers usually have a great deal of difficulty with, and the Univoice speech engine does not.
Brett: What do you account to that?
Brian: I cannot talk too deeply into the technicals because I am not the primary engineer, but mostly it is about context awareness and training the AI to understand the overall topic that is being discussed as well as what it is expecting.
Brett: I was discussing this recently on another episode with an AI company. It was fascinating. So tell me a little bit more about what got you into this, what made you want to build this.
Brian: It was years, years and years ago that I created the seed of the speech engine. Univoice is not my idea. My partner, Sammy, he is the best language teacher in the world. There´s a reason for it because he has a very specific technique of using music in his lessons, and as a tutor, he was just so madly effective. He said I am going to turn this into an app. I am going to make this thing, and here we had Univoice.
When he met me and I already had the technical background, the entrepreneurship, etc., this is what I do as a career is create software businesses like this, I was so excited because years, years and years ago I was so dead sick of my phone´s autocorrect ducking all my shift every time I tried to use a word it didn´t have in its dictionary. I just needed to go make my own. I made a keyboard that would be more context aware and have a lesser propensity to bring things out of context and censor people. It turned out to be so effective that I would have to correct my autocorrect maybe once every three to six month as opposed to four to five times per sentence. We were just natural partners, and it could not have been a more smooth first year.
Brett: Tell me more about that first year. How did that go from the idea to the seed stage to growth?
Brian: That is, thankfully, where my partner and I have been absolutely complementary, so people come to me when they have a business that they know they can scale. My job is to take whatever arbitrary business that is, and turn it into something that can be a national or a global institution. My expertise is in software, business, economics and game theory. Sammy is hands down the best CEO that I have ever met. He has been able to raise every bit of money that we have needed, make all of the relationships we have needed, bring the contracts for the content that is going to make this app an absolute global powerhouse. I have just been free to run operations. It's been a dream.
Brett: Wonderful. Brian mentioned you have social built into your app. Tell me a little bit about how that works.
Brian: There are two aspects of social motivation that are diametrically opposed. We are very competitive, and we love to have partners. We try to address both of these things. Our first system is challenge a friend. That's pretty elemental, and this is totally asynchronous. That makes it very easy for us on the programming side. That's just a matter of you have done a song, and you send it over to me because I am a friend. I do the same song, and we can compare scores. We can address the competitive aspect here as well as the cooperative aspect here by letting it be an iterated game. What we are putting together here is an ongoing language learning buddy relationship. Every day you can challenge me. We are going to get a little bonus for that challenge, but only the winner is going to get the lion's share of that bonus.
If we are participating in the system, we are both better off, but one of us is going to be more better off, and let´s us find that settle point between us being language learning buddies. We are going to do this every day because we are incentivized, and it gives us an extra little competitive kick for fun.
Brett: I am starting to get a better picture of this. You are using Univoice, and it is teaching you language through singing. You are actually singing, and it is recording you. There is some kind of scoring based on an analysis of your singing. Then you can upload that socially, and share it with friends. Then you get scored against them, and then you guys get to compete. Everybody wins for participating, but also you win more if you win the match.
Brian: Absolutely. We should back up a little bit because the challenge a friend is not necessarily that everybody needs to participate in. It's just a little extra feature that if you want. The core of Univoice is just sing along. We will let you know how you are doing. We are going to find as many ways to do that to make it as compelling as we can, but that's the core of the technology because when it comes down to it, there are two major things you need to learn a language. Number one is that you need to train your semantics, your vocabulary, your active vernacular, and that is mostly flashcards. You just need exposure.
The other side is your syntax. That's your ability to actually string sentences together to understand conjugation, how word meaning changes in the context of a sentence. What's really interesting neurologically is that semantics are ultimately attached to your muscle memory. Those are neurons that are activated when you are speaking, and so if you fail to create speech, if you only listen, then you are not going to be nearly as able to create memories and create knowledge as you would be if you did the exact same thing but just said it out loud.
Brett: Absolutely, or even speaking silently but just letting your lips mouth is better than just doing it in your head.
Brian: It is amazing what just basic neural activation does to learning.
Brett: Yeah, or neurons connected to your muscles connected to the rest of your body´s experience and sensory milieu.
Brian: Absolutely. It has been really fascinating to see the linguistic research as I have done. I was very interested ahead of time, but I never really had reason to really deep dive until Univoice. That has been a real pleasure.
Brett: So tell me a little bit more about your personal journey with this startup, or you have mentioned that you have done several others or that this is kind of something you have been doing. What is the biggest personal challenge that you have had growing this company?
Brian: That's A good question. I think that the biggest personal challenge that I have had is learning fundraising. That has been kind of the key sticking point for my life personally. It is hard to just throw that out there because my strategy to overcome that is just to meet partners like Sammy and just focus on other things.
Brett: Right.
Brian: I don't want to throw you something where my response is just to ignore it, but there is always growing pains. There are always things that you need to learn rapidly how to do, and when you are the executive of the company, your job is just to do it. Sometimes you are not going to find somebody. You are not going to be able to delegate. Your job is to do that, but on the days where it can't happen, it can't happen. You are the one who needs to stand up.
Brett: What's the tricky thing about fundraising for you?
Brian: Oh gosh. I don´t know. That's a good question. If I knew, I suppose I would solve it immediately.
Brett: I guess for me I haven't had to raise funds because that´s just not the kind of company that I have, but the thing that scares me about the idea of going and fundraising. A lot of my friends who are founders, and they talk about their fundraising. I am just like you mean I am going to try to take all of my vision, put it into a deck and then go talk to people I have never met before, and have like 30 seconds to sell them on something and then take money from them, not knowing them, and then be beholden them. That whole thing just sounds horrifying to me. I don´t know how anybody does it, but people do it.
Brian: My first couple ventures, fundraising wasn´t necessary because I am working with partners that have an existing business or it is something relatively simple. When I wanted to make money in college, I just did the fast hash. That´s a joke about marijuana, so kids are going to love it. Potatoes are cheap, so I can make some hash browns. Everybody wants breakfast. Done.
That doesn´t take startup capital. I had some folks who were going to lose the farm. They were photographers. They were the most amazing people, absolute masters of their craft, but moms with a camera were doing every wedding, no senior pictures. The industry just absolutely flopped over the course of only a couple years. This was right after I had gotten out of college and had finished my career in the military. We got together. We identified where the majority of their income was coming from on a day-to-day basis, and created a new business model around that. We spent a few years together. We had an absolute blast, and there was no reason to go raise capitals since they could just make sales. We just went and made sales.
Brett: What was the first time that you had to raise funds? What was that like?
Brian: That was for my nascent business Euphrates. The idea is to make freight easier.
Brett: Euphrates.
Brian: It was a great system. We did a lot to push forward DLT technology, in particular with acyclic graphs.
Brett: Something like Iota.
Brian: Yeah, or Phantom. What we realized is that. Sorry, I am saying we instead of me when I was just sitting on the side of a hill for a couple years pondering the world and wondering what to do with myself.
When you have a ledger, it doesn't matter where data enters at all. The ordering matters. Fair ordering is critical, but that's a temporal thing. It's not a spatial thing, so the idea of a direct acyclic graph, I feel like it was Hash Graph and Hedera have the patent on it, but it was convergent evolution. There is no way there weren't 100 people exactly like me coming to the same conclusion on the same day.
Brett: Right, and they weren't actually around first. I don´t think there is a viable pattern on a DAG structure.
Brian: I don´t think it should be viable to patent math.
Brett: There are others using it in production right now.
Brian: Yeah. As it concerns freight and on demand technologies, what we realized is here´s a great way we can create a marketplace with fair ordering that actually has the volume to do things that Uber can´t, that Lift can´t because they are so limited with their technology stack and the number of transactions per second they can actually accommodate.
Brett: What is this? What was this that Euphrates was going to?
Brian: Being able to put contract chains is something that's unique to DLT, and so the idea here is that freight can be made easier through more effective drayage, more effective bills of lading, and other things that usually are taking a vast amount of time or money. You being able to use cryptocontracts to mediate things via AI rather than having a human do drayage, that alone would be huge.
I had a friend up in Seattle working at a maritime shipping firm. They have 30,000 containers in port at any given time, and a couple hundred thousand at sea. All of those containers at port are being charged coastal rent, and that is a crazy amount of money, especially in places like Seattle, LA or San Diego. If you just move that 100 miles inland, you can still store it for the same amount of time, but saves tens of millions of dollars every year per firm. We were just going to start there, but as the technology expands, then you are looking at things like what happens when you arrive. It would be nice to be able to spread the gravel instead of just dump it out, little things.
From there, we could edge into the more generalized service on demand, so ultimately we were creating a service on demand on demand platform. Ultimately, we would want to use this to provide a gradual nice inclined plane to the fourth industrial revolution because it is a platform that could implement for humans or for machines and give us a nice gradual on ramp there rather than expecting people to use something like Uber and then turn around and use something that's completely different for calling up a machine.
Brett: Yeah, so how you do tackle an industry like that? Like supply chain where you have so many different supplies connected to so many different vendors all the way through the chain, how would you get them all to use? I mean how many of them do you have to bring together to use one platform for them to any get value from it.
Brian: That is dependent on the game theory. It really depends on whether there is a strong incentive for centralization. With Euphrates, the game theory is pure centralization, and so what I told myself is if I ever meet anybody that is ultimately going to accomplish this goal, then I should just go all in with them. If they have already snowballed beyond what I could expect of myself, I would be really silly to try to compete. I know we are going to snowball in the same direction and combine, and so that was the strategy. Around 2017 is when I heard about Chainlink and what Sergio Nassar was doing with that was absolutely phenomenal. He has Ari Jewels on the team from day one, and Ari has been publishing some of the best math regarding cryptography I have ever read since the 90s. The Chainlink team has been wonderful to follow. It has been really incredible.
Brett: So it sounds like you have kind of stepped aside from this Euphrates concept into Univoice. Was there a reason for that or is the industry not quite ready or did you end up snowballing into the other company and let them run with it?
Brian: I figured I am going to need a lot of money to do this, so I went fundraising. To answer your question, how do you solve that problem, by creating a business.
Brett: Okay, so part of Univoice is to be generating the funds for you to do this other one.
Brian: Yeah, it is way easier to do something like create Univoice and launch it than it is to walk around begging. Create value, much easier than beg for money.
Brett: Somehow people manage to do both. I haven't figured that out.
Brian: Very successful CEOs, yes.
Brett: Thank you, Brian. This has been a really great conversation.
Brian: A lot of fun. Thank you so much for the call, Brett. You have a good one.