The Exchanges

Q And then let's finish on today. The next five years for you and for, for Memrise, what does the future hold?

A For the next three years, um, we're just going to grow the product we're building now at Memrise, and I think we can, well, the target is a hundred million multi-active users. I want to have like a hundred million people learning languages super joyfully, and just importantly, actually having a kind of transformed outcome in terms of what they learn. And in doing that, I want to build a team who are like incredibly creative in learning tech, Incredibly creative in game design, and incredibly creative in content production. A kind of ideal mind, in a way. And I reckon maybe 250 people collaborating in this very diverse fashion across these departments. And then, you know, once we kind of killed it with Memrise, I then want to, like, um, build, like, a diverse family of incredibly awesome EdTech products. For instance, I want to turn Dawn Till Decadence into a virtual reality game, um, as, like, the ultimate way of teaching history There were tools I wanted to build for schools. There's lots of other stuff, but yeah, for first order of business, make a genuinely successful, powerful company with the app you have now, and then after that, use the team we construct in doing it to create, like, make it a thousand flowers bloom.

AI assessment note: “For the next three years, um, we're just going to grow the product”

Q Biggest mentor to you? Who's been kind of the guiding?

A Um, I guess I've had, like, two super key mentors. One, Bill Warner, who I met when we were doing Techstars in Boston in 2011. He invented the first video editing software, and he was someone who had this very profound belief that you should build a startup from intention. My intention being, like, give people learning superpowers, make learning joyful. Not Not always from the kind of, like, Paul Graham, like, test, test, test, test, test sort of way. Um, and so he, he had a profound impact on the kind of confidence with which I could approach the way I wanted to build the company. Um, and then more recently, Peter Reid, uh, in London, uh, has been, uh, an amazing mentor, uh, not least as he takes me to Burning Man.

AI assessment note: “I guess I've had, like, two super key mentors. One, Bill Warner”

Q And now you're teaching that at scale to, to millions of people with Memrise. So talk to me a bit about that. And as your LinkedIn profile says, as I said, millions of people learning languages joyfully with Memrise. So what's the essence of learning? And as you said, Josh in a year became a memory champion. So what is that? And how do you break down the process?

A So in one word, the essence of learning is emotion. Interest is at the very, very heart of, of all learning. And if you're, you know, if you're, if you're alive to something, if you, if you love it, if you're fascinated, then basically anybody can learn anything. And this is kind of true, obviously, to anybody who's ever learned something. You know, you learn most at school, not from your school, but from your friends. You know, when you find a topic which you absolutely love, anyone can learn really fast. So, um, so what that breaks down to in kind of, um, In fundamental terms is that the learning at its best is perceptual. It's like cognitively well organized. In other words, there's a good technique and it's social. Um, and, and that's where the emotion comes from, these three layers. Um, and like perceptually, you want to make learning which is, um, vibrant to the senses, which is full of color and entertainment and fun. Cognitively, you want to have a system which basically works with your brain to kind of install the information at the perfect rate and with the perfect amount of variety and so on and so forth. And socially, you need a reason to learn in the first place. Uh, so the way we've tied those into Memrise is we are creating, um, incredibly vibrant content. So that's at the perceptual level. So for instance, we, um, we've just, uh, sent a bus around Europe to go c…

AI assessment note: “So in one word, the essence of learning is emotion.”

Q Biggest mentor to you? Who's been kind of the guiding?

A Um, I guess I've had, like, two super key mentors. One, Bill Warner, who I met when we were doing Techstars in Boston in 2011. He invented the first video editing software, and he was someone who had this very profound belief that you should build a startup from intention. My intention being, like, give people learning superpowers, make learning joyful. Not Not always from the kind of, like, Paul Graham, like, test, test, test, test, test sort of way. Um, and so he, he had a profound impact on the kind of confidence with which I could approach the way I wanted to build the company. Um, and then more recently, Peter Reid, uh, in London, uh, has been, uh, an amazing mentor, uh, not least as he takes me to Burning Man.

AI assessment note: “I've had, like, two super key mentors. One, Bill Warner”

Q And you're lighting up that desire with, with technology, as we said, with Memrise there. So what do you think are the prospects then for educational technology? What's the future? I mean, we have, we have Tim Berners-Lee saying that to some extent it hasn't fulfilled its potential. Do you agree?

A Yeah, I do agree that we are very, very early in the journey of ed tech, so in fact, there's not very much good stuff out there. I mean, I think the, the horizon is that we can create a system which can basically give everybody learning superpowers, um, and what do we have out there? You know, they work maybe, you know, pretty much like a textbook, maybe a textbook with a little bit of decoration, and I think that, like, over the next, like, five, 10, 20 years, what we'll see is, is less the kind of The current tech innovation, which is making it accessible, putting it on smartphones, and so on, and more something which can begin to drive, you know, hundreds of millions of people on a daily basis to learn just as part of their lives, and do so in a fashion which really lights up the motivation, which is cognitively incredibly powerful, and which is just entertaining and fun.

AI assessment note: “I do agree that we are very, very early in the journey of ed tech”

Q So far, who do you think has done this well? Obviously, looking at the landscape, is there one in particular, or a couple that you think actually they've integrated the kind of Enthusiasm and, um, kind of gamification or enjoyment that learning requires well.

A You know, I admire what the gaming industry's done for the ability to get people motivated and sustain them in activities as, like, bizarrely nonsensical as, like, swiping your finger across diamonds. You know, I think that that's a beautiful example of what can be done with tech if you have the right team and, um, and, um, and processes and narrative and so on. I'm, I'm a big admirer of, um, of what Quizlet are doing at Quizlet Live. So this is like a live classroom game, which actually kind of turns the classroom, turns that social space into a more focused learning environment. I think that's very creative and admirable, but you know, as I say, I think we're in the super early stages, and I think some of the stuff which will come in the next five, 10 years will make what we have now seem like almost embarrassingly faint and preliminary.

AI assessment note: “I'm a big admirer of, um, of what Quizlet are doing at Quizlet Live.”

Q I'm really intrigued, and this sounds like the most complex question. How has your learning on learning developed over the years, and how has your, kind of, uh, emotions towards it developed, and kind of the intricacies of it?

A Great question. I mean, I think, I, when I began, I thought of learning as purely cognitive. You know, with memory techniques and stuff, I was like, ok, I've learned to do this thing, I've taught other people to do it, It's just a purely cognitive thing, and the more I go through it, the more I have evolved into thinking of it as, even more importantly than the cognitive, a social thing. The meaning of learning is actually more important than how you do it. You know, you can learn with a blackboard and a piece of chalk, and if you're absolutely psyched about it, if you love it, if you're fascinated by the topic, if you respect your teacher, and if your teacher's at your level, then you can learn anything. And so, for instance, in schools, what I came to learn, because I used to do a lot of teaching in schools, is that, um, The most difficult bit is not like, you know, your classroom technique or technology you're using or whatever. The difficult bit is like actually, um, lighting up inside people the desire to learn and showing them why it's socially relevant to them.

AI assessment note: “evolved into thinking of it as, even more importantly than the cognitive, a social thing.”

Q And then if you weren't doing Memrise, what would you be doing? I wonder this sometimes.

A Um, if I was being ambitious and I wanted to sort of continue to try and change the world, I think I would work on, um, reinventing institutional structures to bring the same level of innovation into, um, institutions, university schools, groups of all kinds that you see just in tech within the narrow space of the products, um, because I think that most of society's ills are actually institutional rather than, like, specifically technological, and I think that can be incredible. If I wasn't religious, and I probably won't be after Memrise, I think, I think I'd have, like, a kind of philosophical commune on a Greek island. And then maybe we would sort of design artificial intelligences and just hang out and sit in the sun.

AI assessment note: “I think I would work on, um, reinventing institutional structures”

Q And you're lighting up that desire with, with technology, as we said, with Memrise there. So what do you think are the prospects then for educational technology? What's the future? I mean, we have, we have Tim Berners-Lee saying that to some extent it hasn't fulfilled its potential. Do you agree?

A Yeah, I do agree that we are very, very early in the journey of ed tech, so in fact, there's not very much good stuff out there. I mean, I think the, the horizon is that we can create a system which can basically give everybody learning superpowers, um, and what do we have out there? You know, they work maybe, you know, pretty much like a textbook, maybe a textbook with a little bit of decoration, and I think that, like, over the next, like, five, 10, 20 years, what we'll see is, is less the kind of The current tech innovation, which is making it accessible, putting it on smartphones, and so on, and more something which can begin to drive, you know, hundreds of millions of people on a daily basis to learn just as part of their lives, and do so in a fashion which really lights up the motivation, which is cognitively incredibly powerful, and which is just entertaining and fun.

AI assessment note: “Yeah, I do agree that we are very, very early in the journey of ed tech”

Q What have been the big challenges in growing such a, such a diverse, multifaceted, and kind of hungry for education, something that I haven't always been myself? Uh, yeah, the team.

A You know, one lesson you learn the hard way is that, um, You don't recruit people for their talent alone. You recruit them for their fit with the culture. And so everyone who comes has to amplify the team's mentality, culture, imagination. Another challenge is, um, is keeping the sense of, like, vibrant impact, which you have when you're super small, as, as the team begins to grow. And so, so we, we have this kind of philosophy of curated autonomy. So people love to have impact. They love to be autonomous. But how do you do that in a way which actually drives business forward and continues the growth and so on? Um, and we found a really nice methodology, which I actually got from the, a book called, um, Creativity Inc.

AI assessment note: “Another challenge is, um, is keeping the sense of, like, vibrant impact”

Q And then if you weren't doing Memrise, what would you be doing? I wonder this sometimes.

A Um, if I was being ambitious and I wanted to sort of continue to try and change the world, I think I would work on, um, reinventing institutional structures to bring the same level of innovation into, um, institutions, university schools, groups of all kinds that you see just in tech within the narrow space of the products, um, because I think that most of society's ills are actually institutional rather than, like, specifically technological, and I think that can be incredible. If I wasn't religious, and I probably won't be after Memrise, I think, I think I'd have, like, a kind of philosophical commune on a Greek island. And then maybe we would sort of design artificial intelligences and just hang out and sit in the sun.

AI assessment note: “I think I would work on, um, reinventing institutional structures”

Q And now you're teaching that at scale to, to millions of people with Memrise. So talk to me a bit about that. And as your LinkedIn profile says, as I said, millions of people learning languages joyfully with Memrise. So what's the essence of learning? And as you said, Josh in a year became a memory champion. So what is that? And how do you break down the process?

A So in one word, the essence of learning is emotion. Interest is at the very, very heart of, of all learning. And if you're, you know, if you're, if you're alive to something, if you, if you love it, if you're fascinated, then basically anybody can learn anything. And this is kind of true, obviously, to anybody who's ever learned something. You know, you learn most at school, not from your school, but from your friends. You know, when you find a topic which you absolutely love, anyone can learn really fast. So, um, so what that breaks down to in kind of, um, In fundamental terms is that the learning at its best is perceptual. It's like cognitively well organized. In other words, there's a good technique and it's social. Um, and, and that's where the emotion comes from, these three layers. Um, and like perceptually, you want to make learning which is, um, vibrant to the senses, which is full of color and entertainment and fun. Cognitively, you want to have a system which basically works with your brain to kind of install the information at the perfect rate and with the perfect amount of variety and so on and so forth. And socially, you need a reason to learn in the first place. Uh, so the way we've tied those into Memrise is we are creating, um, incredibly vibrant content. So that's at the perceptual level. So for instance, we, um, we've just, uh, sent a bus around Europe to go c…

AI assessment note: “So in one word, the essence of learning is emotion.”

Q I'm really intrigued, and this sounds like the most complex question. How has your learning on learning developed over the years, and how has your, kind of, uh, emotions towards it developed, and kind of the intricacies of it?

A Great question. I mean, I think, I, when I began, I thought of learning as purely cognitive. You know, with memory techniques and stuff, I was like, ok, I've learned to do this thing, I've taught other people to do it, It's just a purely cognitive thing, and the more I go through it, the more I have evolved into thinking of it as, even more importantly than the cognitive, a social thing. The meaning of learning is actually more important than how you do it. You know, you can learn with a blackboard and a piece of chalk, and if you're absolutely psyched about it, if you love it, if you're fascinated by the topic, if you respect your teacher, and if your teacher's at your level, then you can learn anything. And so, for instance, in schools, what I came to learn, because I used to do a lot of teaching in schools, is that, um, The most difficult bit is not like, you know, your classroom technique or technology you're using or whatever. The difficult bit is like actually, um, lighting up inside people the desire to learn and showing them why it's socially relevant to them.

AI assessment note: “I have evolved into thinking of it as, even more importantly than the cognitive, a social thing.”

Q So far, who do you think has done this well? Obviously, looking at the landscape, is there one in particular, or a couple that you think actually they've integrated the kind of Enthusiasm and, um, kind of gamification or enjoyment that learning requires well.

A You know, I admire what the gaming industry's done for the ability to get people motivated and sustain them in activities as, like, bizarrely nonsensical as, like, swiping your finger across diamonds. You know, I think that that's a beautiful example of what can be done with tech if you have the right team and, um, and, um, and processes and narrative and so on. I'm, I'm a big admirer of, um, of what Quizlet are doing at Quizlet Live. So this is like a live classroom game, which actually kind of turns the classroom, turns that social space into a more focused learning environment. I think that's very creative and admirable, but you know, as I say, I think we're in the super early stages, and I think some of the stuff which will come in the next five, 10 years will make what we have now seem like almost embarrassingly faint and preliminary.

AI assessment note: “I'm a big admirer of, um, of what Quizlet are doing at Quizlet Live.”

Q What was the speed at which you taught people?

A Well, um, one interesting example is this journalist called Josh Furr, who I met at the, the American memory championships, which forgive me for mentioning, I won. Afterwards, he was like, um, yeah, so you're a genius. And I was like, no, no, no, I'm not a genius. It's all the technique. It's all how you handle your imagination, your mind. And he said, well, that's nonsense. And I was like, okay, I'll train you. And so, um, I ended up training him. I trained him for a year. And he actually won the American memory championships the, the year after. Um, and out of that came this book he wrote, Moonwalking with Einstein, which tells that story a little bit along with a bunch of the history of memory techniques.

AI assessment note: “I trained him for a year”

Q What was the speed at which you taught people?

A Well, um, one interesting example is this journalist called Josh Furr, who I met at the, the American memory championships, which forgive me for mentioning, I won. Afterwards, he was like, um, yeah, so you're a genius. And I was like, no, no, no, I'm not a genius. It's all the technique. It's all how you handle your imagination, your mind. And he said, well, that's nonsense. And I was like, okay, I'll train you. And so, um, I ended up training him. I trained him for a year. And he actually won the American memory championships the, the year after. Um, and out of that came this book he wrote, Moonwalking with Einstein, which tells that story a little bit along with a bunch of the history of memory techniques.

AI assessment note: “I trained him for a year. And he actually won the American memory championships”

Q What have been the big challenges in growing such a, such a diverse, multifaceted, and kind of hungry for education, something that I haven't always been myself? Uh, yeah, the team.

A You know, one lesson you learn the hard way is that, um, You don't recruit people for their talent alone. You recruit them for their fit with the culture. And so everyone who comes has to amplify the team's mentality, culture, imagination. Another challenge is, um, is keeping the sense of, like, vibrant impact, which you have when you're super small, as, as the team begins to grow. And so, so we, we have this kind of philosophy of curated autonomy. So people love to have impact. They love to be autonomous. But how do you do that in a way which actually drives business forward and continues the growth and so on? Um, and we found a really nice methodology, which I actually got from the, a book called, um, Creativity Inc.

AI assessment note: “Another challenge is, um, is keeping the sense of, like, vibrant impact”

Q I'm intrigued about the personalization of learning and kind of, you know, different people learn in very different ways. How does that affect kind of your work with memorizing? Like, does learning need structure as well? We said about kind of...

A Yeah, I mean, I, I, I think that the personalization of learning is very, very interesting, but actually there's a much more fundamental level, which you've got to get right, which is basically the kind of the fundamental technique of it. And I kind of like to give the analogy of a bicycle. So like a bicycle can like take anybody and make them go three or four times faster than their body could. All by itself. And the design of the bicycle took ages. You know, it's only in the late-nineteh century that we finally worked out how to amplify the human body in this way and make it go three or four times faster. And then later on, you can talk about the personalization of bicycles. Very interesting. But actually, the bicycle is the thing you've got to invent first, and that's kind of what we're doing with our learning tech. It's quite simple, actually, but very powerful when you combine together repetition, active recall, you know, all the kind of temporal and spatial elements of Of, uh, of a learning system, which deposits learning in the brain, basically at the perfect, rather accelerated rate.

AI assessment note: “personalization of learning is very, very interesting, but actually there's a much more fundamental level”

Q up in the next six months, and thinking even more better than that, and running a startup, you as founder and CEO, how do you approach kind of running a scalable, but So you're a slow growth startup that's not a Snapchat, all up and to the right, as we love in our graphs and VC pitches. So how do you approach that kind of slow, scalable, and continuous growth?

A So, uh, for me, the, the essence of building a startup is, is constructing the team. And, um, what are you constructing a team for? You're, you're, you're constructing a team to, like, express an idea and pursue a truth. So I know, my contribution is I know that anyone can have learning superpowers. And the process of growing a startup is to, like, Find and arrange the people who have all the chance, which I obviously lack. And by the way, doing a startup has been a startling experience of my own limitations. Um, but like, growing this team, and it took us, it took me a long time to fully understand this, but since about, ah, early 2015, when we were about 10 people, we're, we're now at 45, um, but you have to express in how you build the team the values you want your product to espouse. And so, for instance, our team is like more than 50% women. It's incredibly diverse. There are 26 nationalities in a team of 42, We have, like, constantly, we have people coming to the office to deliver talks on random things. Next week, we've got a sommelier friend coming in to teach them on wine tasting. And so the whole atmosphere and architecture of the team has to be in love with learning in order for us to collectively create the ultimate learning product.

AI assessment note: “for me, the, the essence of building a startup is, is constructing the team”

Q I'm intrigued about the personalization of learning and kind of, you know, different people learn in very different ways. How does that affect kind of your work with memorizing? Like, does learning need structure as well? We said about kind of...

A Yeah, I mean, I, I, I think that the personalization of learning is very, very interesting, but actually there's a much more fundamental level, which you've got to get right, which is basically the kind of the fundamental technique of it. And I kind of like to give the analogy of a bicycle. So like a bicycle can like take anybody and make them go three or four times faster than their body could. All by itself. And the design of the bicycle took ages. You know, it's only in the late-nineteh century that we finally worked out how to amplify the human body in this way and make it go three or four times faster. And then later on, you can talk about the personalization of bicycles. Very interesting. But actually, the bicycle is the thing you've got to invent first, and that's kind of what we're doing with our learning tech. It's quite simple, actually, but very powerful when you combine together repetition, active recall, you know, all the kind of temporal and spatial elements of Of, uh, of a learning system, which deposits learning in the brain, basically at the perfect, rather accelerated rate.

AI assessment note: “personalization of learning is very, very interesting, but actually there's a much more fundamental level”

Q up in the next six months, and thinking even more better than that, and running a startup, you as founder and CEO, how do you approach kind of running a scalable, but So you're a slow growth startup that's not a Snapchat, all up and to the right, as we love in our graphs and VC pitches. So how do you approach that kind of slow, scalable, and continuous growth?

A So, uh, for me, the, the essence of building a startup is, is constructing the team. And, um, what are you constructing a team for? You're, you're, you're constructing a team to, like, express an idea and pursue a truth. So I know, my contribution is I know that anyone can have learning superpowers. And the process of growing a startup is to, like, Find and arrange the people who have all the chance, which I obviously lack. And by the way, doing a startup has been a startling experience of my own limitations. Um, but like, growing this team, and it took us, it took me a long time to fully understand this, but since about, ah, early 2015, when we were about 10 people, we're, we're now at 45, um, but you have to express in how you build the team the values you want your product to espouse. And so, for instance, our team is like more than 50% women. It's incredibly diverse. There are 26 nationalities in a team of 42, We have, like, constantly, we have people coming to the office to deliver talks on random things. Next week, we've got a sommelier friend coming in to teach them on wine tasting. And so the whole atmosphere and architecture of the team has to be in love with learning in order for us to collectively create the ultimate learning product.

AI assessment note: “the essence of building a startup is, is constructing the team.”

Q And a super nerdy question, slightly controversial, you mentioned, you mentioned the integration of AI into learning processes there. A lot of people think AI will actually kind of Diminish learning for humans and create us apathetic beings just kind of waiting while machines take us everywhere and do things for us. What's your take on kind of AI's introduction into education?

A AI generally is, um, I think nobody actually understands. I think everyone's kind of got the wrong end of the hook. I think, you know, people's question with AI is like, oh, is it going to become superhuman and then like take over? I think that the more fundamental question is like, what kinds of intelligence can we create? Intelligence isn't a single, single thing, um, and I think the challenge for people designing AIs about the quality, personality, unique charisma, perspective on the world of the intelligence that they're inventing. I actually think of AI as continuous with text, so text actually captures people's intelligence, and so you can read a book, and then you can hear the intelligence of someone from, like, a hundred years ago, and then you have, like, text can capture a Beethoven symphony in note form, and then you have a Beethoven performance, which can actually deliver that symphony live and actually record that. AI, I think, is the next step Whereas instead of just having this particular Beethoven symphony, we designed the Beethoven artificial intelligence, which can create the symphony live, and we can collaborate with you. And so for me, AI is about, um, the challenges to design beautiful artificial intelligences, and the possibility is mainly in terms of collaboration.

AI assessment note: “I think that the more fundamental question is like, what kinds of intelligence can we create?”

Q And a super nerdy question, slightly controversial, you mentioned, you mentioned the integration of AI into learning processes there. A lot of people think AI will actually kind of Diminish learning for humans and create us apathetic beings just kind of waiting while machines take us everywhere and do things for us. What's your take on kind of AI's introduction into education?

A AI generally is, um, I think nobody actually understands. I think everyone's kind of got the wrong end of the hook. I think, you know, people's question with AI is like, oh, is it going to become superhuman and then like take over? I think that the more fundamental question is like, what kinds of intelligence can we create? Intelligence isn't a single, single thing, um, and I think the challenge for people designing AIs about the quality, personality, unique charisma, perspective on the world of the intelligence that they're inventing. I actually think of AI as continuous with text, so text actually captures people's intelligence, and so you can read a book, and then you can hear the intelligence of someone from, like, a hundred years ago, and then you have, like, text can capture a Beethoven symphony in note form, and then you have a Beethoven performance, which can actually deliver that symphony live and actually record that. AI, I think, is the next step Whereas instead of just having this particular Beethoven symphony, we designed the Beethoven artificial intelligence, which can create the symphony live, and we can collaborate with you. And so for me, AI is about, um, the challenges to design beautiful artificial intelligences, and the possibility is mainly in terms of collaboration.

AI assessment note: “I think that the more fundamental question is like, what kinds of intelligence can we”