Bo Bennett On Building Science Based Learning

HOST: It's good to have you back, Bo. Thirty-one episodes in. Last time, you told me you and your wife sit down every couple of days to do a section of your memoir on MemoirMaker, and I've been thinking about that, because today's topic is another personal one—ScienceBasedLearning.com—which sounds like it started with you actually trying to learn a language and finding that nothing out there was good enough. So let's start there: what were you trying to learn, and what did you try first? GUEST: So yes, like most of the websites, it was developed as something that I ultimately needed and wanted. I was trying to learn Spanish. And I was finding that all of the solutions out there just were not good enough. They were very much based on hype. And a lot of the solutions were really they the primary goal was to get users to subscribe to their app and not really teach them or teach them just enough to kind of drag them on as long as possible to get as much revenue from them as possible. And I found that incredibly annoying. As a scientist, a social scientist, I wanted to know what exactly is the best way to learn a new language, the right way to learn, and incorporate those techniques in some kind of application that allowed people to learn as effectively as possible. Without wasting time. Uh like I said, I built this for me. And it works beautifully and I I love it. I use it every day. I've been using it every day for the last eight months or so. And I believe that I've been learning quite a bit. HOST: Eight months of daily use, that's impressive. Now, you mentioned existing options focused more on hype and subscriptions than actual learning. When you say "best way to learn from a scientific perspective," what's the actual mechanism and has it replicated? What specifically are those foundational principles that you were looking for? GUEST: I was looking for how the brain retains information, especially new information doing having to do with language learning. And because that's a little bit different than other information. It's not the brain doesn't process all information the same way. So focusing on language learning was kind of new to me. I'm a social scientist with dabbling in in cognitive science, but language learning is really a very specific branch of cognitive science, and that's the part I really had to dig into to find out what the tricks were, or if you want to call them tricks, they're really techniques that um that have been proven time after time in study after study to be the most effective. And specifically, one of the most effective ways to to learn a language is by uh having a word or term presented to you, and once you check it off the list, it doesn't just go away forever. It'll come back. It'll come back in 15 minutes. And then you check that off if you get it, and then it comes back in an hour. And then it comes back in a day, and then three days, and then a week, and then a month. So it keeps on extending that time to make sure that it's remembered in your short-term memory and your long-term as well. And here's the secret. When you forget it, let's say after three days, I'm like, ah, I don't remember that word. I don't remember what it means, and you you check it off as no, I don't know this one. It goes back a step. So now you're learning it in the next hour instead of the next three days. So it'll bring you down the scale, and then it'll bring you back up. And this technique really allows you to master the terminology and remember it long time long term. So it's not 100% effective in that if I do go through a term when I remember it once and then and then I remember it for the hour and three days and seven days and then eventually 30 days and it goes away, then a year from now, the chances are probably anywhere between 90 and 95% that I will remember that term for the rest of my life. And that is really good. You think, ah, well, that's you're still forgetting 10% of the terms. Yeah, but compared to what people normally forget, uh, learning the classic way, then this is it's completely different. If it's it's a whole another ball game. So this is called spaced repetition, and it's proven, and it's incredibly effective. But yet, still, that's only one of the techniques that's in this app, just one of the many. HOST: Spaced repetition, right—really leveraging that forgetting curve and memory consolidation. Now, you mentioned infinite AI content as another key difference. So you have this system for retention, but how do you make sure the content itself is calibrated to the learner's level—you mentioned CEFR levels. Does the app automatically move you along, or do users self-report? And how do you prevent AI from getting repetitive in its descriptions so it stays engaging? GUEST: Yeah, you got plenty of questions there. Let's start with the AI content and uh and how it creates the content, how it calibrates the content to the user's learning level. And that is through the CERF levels, which is a a classification for language learning that goes from A1, A2, B1, B2, C1, C2. C2 being like the most advanced, A1 being the you're a very beginner. So, basically, it's self-reporting. As a user, you self-report. Most users who, well, if you never started, then it's just A1. You start there. But if you have been learning language for a while, you pretty much know where you are, and then you could choose to start on that level. If you find that you're really struggling and you're having difficulty with all these terminology the terminology and uh getting through reading an article or or whatever other task that you're working on, then you could lower the level. If you find that it's just too easy, it's not challenging enough for you, then you raise it a level. And you could do this as you go along, too. So, it it really does have to do with self-reporting. Now, your second question has to do with the uh getting repetitive in descriptions. Um AI is very good at this. We have it set up so there are literally thousands of possible uh permutations of the way that uh AI will present a word or a term. And uh it it just it there's no repetition where there's not supposed to be. So, for the uh for the terminology, for example, when you're going through it, it is repetition. It's the same term over and over again. Uh the same definition, that doesn't change. What does change is the articles you read, the stories you read. Uh those are completely different. Those will never be the same. Um when you're conversing with AI kind of in a freestyle way, that's always going to change because your response dictates the response of the AI, and that response changes, and it changes your response. So, no two conversations are ever the same. So, there's plenty of variation where there needs to be. HOST: Got it, so user-driven self-calibration across those levels, and generating truly dynamic content for reading and freeform conversation. Now, every tool we've talked about across thirty episodes has been aimed at authors, marketers, or businesses. ScienceBasedLearning is for anyone trying to learn French or Japanese on their lunch break. That's a fundamentally different customer. What surprised you about building for a mass consumer audience versus the author vertical you know so well? GUEST: Well, it's again a different market and I am not very well versed in marketing to the end user, the consumer. And that's the part that I got to get better at because this really is an amazing application and I feel like not enough people know about it or taking advantage of it. So that's the part I have to um to get better at. Whereas yeah, when it comes to marketing to authors or businesses, that's what I've been doing. So I have I have a better grip on that market.Well, it's again a different market and I am not very well versed in marketing to the end user, the consumer. And that's the part that I got to get better at because this really is an amazing application and I feel like not enough people know about it or taking advantage of it. So that's the part I have to um to get better at. Whereas yeah, when it comes to marketing to authors or businesses, that's what I've been doing. So I have I have a better grip on that market. HOST: Yeah, that direct-to-consumer marketing push is a different challenge for sure. Now, you also built something called Professor Mode—it's like an instant grammar explanation you can pull up from any screen. That sounds like it's replacing the moment when a learner might have to stop and search somewhere else. How did that feature come about, and is that the piece that most existing apps are missing? GUEST: to AI but it works extremely well. Is it the piece that people been the one piece that's been missing? I'm not sure about that but it is an important piece of the puzzle that brings everything together nicely. HOST: It does sound like a critical piece. Before we go — for listeners who want to follow up on all of this, where can they find you and the work you're doing? GUEST: You can find me at archieboy.com. That's my main website and also sciencespeechlearning.com. And this is this is also this is actually an app that's available on the app store for both the iPhone and the Android system. HOST: Great. Thanks so much, Bo. It was fascinating to hear about how your background in psychology came together with building this tool for yourself. And thank you for spending part of your day with us. We'll have to get you back on soon to show me how it's actually working for you. Until next time — that's a wrap.