In this episode of Week in Edtech, Alex and Ben discuss:
1. AI
2. K12
3. Higher Ed
4. Workforce
Special Guests:
In this episode of Week in Edtech, Alex and Ben discuss:
1. AI
2. K12
3. Higher Ed
4. Workforce
Special Guests:
Welcome to Season Eight of Edtech Insiders where we speak to educators, founders, investors, thought leaders and the industry experts who are shaping the global education technology industry. Every week we bring you the week in edtech. important updates from the Edtech field, including news about core technologies and issues we know will influence the sector like artificial intelligence, extended reality, education, politics, and more. We also conduct in depth interviews with a wide variety of Edtech thought leaders and bring you insights and conversations from ed tech conferences all around the world. Remember to subscribe, follow and tell your ed tech friends about the podcast and to check out the Edtech Insiders substack newsletter. Thanks for being part of the Edtech Insiders community enjoy the show.
Ben Kornell:Everybody, it is Edtech Insiders, Ben and Alex coming at you with another week in edtech. So much going on in the world of education and technology. We've got news releases every single day coming at us from all sorts of angles. We're here to help make sense of it and help you figure out how to navigate your attack feature. All right, let's jump in Alex, what's going on on the pod we've got a ton of great content coming out on pot.
Alexander Sarlin:We do so we have just put out two really fun episodes. One was with Tom Sayer of Ello. He's the CEO and co founder of Ello. And Ello is just doing fascinating work in the AI space and the reading space. We have an episode with Andrew Sachs and Emine Naz Khan. And that is a company called Nobel learning does global peer to peer networks that really put students in the driver's seat. Amina is a Turkish student who's run all these student groups with Nobel and Andrew is the CEO. And coming up soon we're talking to Alejandro Gibes de Gac, who is the founder of Springboard collective and has a new tool called Paloma, which is all about parental involvement in education using AI and using really cutting edge mobile technology. That's a great conversation. So check those all out.
Ben Kornell:Yeah, on the event side, we have our happy hour in Redwood City here in the Bay Area coming right up. That will probably be before the podcast is released. And then we also have our big event at ASU GSB. Huge shout out to our sponsors who have helped us put that happy hour together. More information can be found on our LinkedIn page, sign up for the waitlist. We're laddering off the last, you know, 100 spots here, but we have an amazing group of folks who are going to come together for community collaboration, commiseration and inebriation. It'll be great fun. Speaking of inebriation, we're probably going to have some families in LAUSD that need a drink after they got the news today that the district is launching the first district wide student and parent AI Chatbot. That really is, I think, a bold move in the direction of AI embedded, teaching, learning and community engagement. It's called Ed and Alberto Carvalho, I was at an event last fall when he even talked about the concept. It felt bold at the time, and they've really gone through and delivered on it. My understanding is only available to 100 schools so far, and they're doing a gradual release in LA Unified and early returns are super easy chat interface. Multilingual, still mainly chat oriented, so not multimodal yet. And, you know, they've trained it on a repository of data and information from LAUSD. Alex, is this the future of K 12 education that districts are going to be launching and releasing their own products? Could you imagine universities doing the same thing? Yeah,
Alexander Sarlin:well, let's be clear. This is a partnership with an edtech company. So, you know, LAUSD is in partnership with all here, which is a really innovative chatbot based ad tech company, founder, Joanna Smith, Griffin, and they've been doing AI chatbots for schools for a number of years, actually, since before the generative AI revolution. So it's not that they're purely doing this completely on their own. They're actually partnering with an edtech company and then building a platform that incorporates lots of other ad tech pieces, features and connections to different systems within it. So is this the future? I kind of hope so. I mean, you know, they did a very big, very flashy launch last week for Ed and they're calling it ed as in you know, there's a character named Edie, it looks like a little sort of son, that Edie is the name of the character, but it's also obviously education and Department of Ed and all those great things. The idea here is that anyone in the LE USD, which is enormous will over time, at least as it rolls out have access to chatbots. This means that students will have access to ask about their own grades, their own records, they'll be able to find out when their breaks are through a chat interface, parents will have access to it, educators will have access to IT administrators will have access to it. The real innovation here as much as I understand it so far after watching the sort of launch event is that it's really meant to be one platform that combines all the information within a district that each stakeholder and constituency needs on a regular basis. I mean, it's one of these things where they it's a very flashy rollout, and we're gonna look back in a year and either be like, they were way ahead of the curve. And while everybody's doing this, and they're amazing, or be like, oh, you know, it was one of those. Wow, really high hopes really trying really hard but didn't quite catch on things like what I've had initiative in LAUSD, you know, many years ago, I'm betting on the former, I think that they're doing something really smart here. And they're being way ahead of the curve on districts, really embracing AI and embracing chatbots rather than fearing them or sort of tiptoeing into them, do you think work?
Ben Kornell:I mean, I think it's too early to say whether it's going to work and you know, betting against things working in LAUSD would make you a lot of money. But that said, you know, Age of learning is a partner in this, you know, what I think they did was really smart in that they said, there's a unique call to action, there's a unique opportunity, how do we get philanthropic funding? And how do we get the players around the table, who know what they're doing to kind of work with us to build this, right. And I think they also have really landed on a very simple interface. It's basically Google AI search. But just for LAUSD, so you know, when you do an AI search on Google or on Bing, or something like that, instead of just getting the list of results, you get a summary of what the most likely result is. And on the Chatbot technology, that's not like five, or 10 leaps ahead, that's really one leap ahead. But the idea that they would consolidate all the information at the schools and across systems, that's really the innovation here, the interface, they're not trying to innovate, they're just trying to make a simple language interface. And then, you know, what they're able to do for a fraction of the class is make that accessible in multiple languages, and across platforms, because the AI is doing that lifting. So I think it's actually really interesting how they've designed this with the user needs in mind. The iPad was one of those cases where I think it was like, we have a technology that's game changing, let's bet on the technology. And we'll figure out the use cases. And in this case, they said, we've got this simple use case, like an Ask Me Anything bar, could we make that work for students and parents? Now, the reality is, there's going to be people with a political axe to grind, who are going to make it hallucinate, or they're going to make it say, incorrect or inaccurate things. And then the question is, what's the public tolerance for that, and the other area where I think they could fall into a trap, he just low user engagement overall. And so with a flashy launch, I think they're gonna get it a spike. But with most AI tools, right now, what we see is there's a big, you know, kind of burst of energy, and then it peters out much of the way like the lens camera release, like they went to millions of users, and then it all drops off, we're seeing that pattern play out on a bunch of apps. And I think, if this is sustained, I think it could be really, really powerful. Yeah,
Alexander Sarlin:and if there's training in place, and you know, a clever rollout, to actually teach parents, what it can do for them, and how to use it, and teach students what it can do for them and how to use it. You know, this is one of those things where just throwing the technology, even if it's amazing towards people doesn't mean that it's going to change their behavior or get them excited about it. But if it's actually integrated into the school year, if it's the, you know, for example, maybe it's the place where you schedule parent teacher conferences are the place where students get their report cards of the place where, you know, there's so many things that you can do, once this is really, you know, moving to, you know, force people quote, unquote, or at least, you know, really encourage people to get into a system like this. And that could be foot in the door for them using it for all the kinds of exciting things they're looking for. The other thing that I think is worth mentioning about it is it's extremely language friendly. This is something that AI does really well is translations. And this little Edie character talks in dozens of languages, I think 100 plus maybe and that obviously represents LA's unbelievably diverse community of families. They have Armenian in there as an example in the launch but includes lots of Asian languages, obviously Spanish and you know, many different things. So that's it. Another thing that's pretty interesting, sort of it's not the headline, necessarily of this technology. But that by itself could open doors for usage totally.
Ben Kornell:And all of this is against a backdrop of some pretty big and bold moves in the AI space. So you know, Alex, you and I, as we were prepping, we thought, let's do a quick around the world and AI and just hit on the big stories, a couple that came to my mind. One was Google signing a deal with Apple to basically have Gemini be powering the iPhone, we don't have full confirmation of all those things. But you know, many people were expecting a big apple release in the fall with some sort of unlock for their new AI capability. And I think what this shows is, Apple is behind when it comes to AI, but they're not about to align with open AI who's connected to Microsoft. So we now start seeing the battle lines being drawn. Apple and Google together, Microsoft and open AI together, Facebook are meta doing the open source. And then we also had grok, from x or Twitter released. So I think that's going to be an interesting thing to watch. One thing that it does make me think of is Google and Apple are the main authenticity platforms. So when you're kind of looking at a Google Store, or like Apple store, they have those marketplaces and the others don't. And so that really puts Microsoft at a potential disadvantage. The other big player here that we're going to have to figure out where they land is Amazon. And it appears that Amazon is pretty laser focused on b2b enterprise AI solutions through their cloud. So that's kind of my read on the landscape. A couple other great announcements, but I don't want to steal any of your thunder. What was on your radar? Well,
Alexander Sarlin:I found the NVIDIA launch really interesting. Yeah. And Vidya has seen, I think it might be the fastest growing stock price in the history of markets and stocks, I believe. It's something unbelievable, I saw a story this week that said that, if it continues to grow at this pace, it will outpace the global economy, which is, you know, probably won't continue at this pace. But it's incredible. And in the midst of these heights, it's not slowing down at all. And they've released their new AI graphics processing unit, they call it Blackwell this week at their developer conference. And you know, Nvidia has already left the field. And you know, there are lots of people trying to catch up with them in terms of AI hardware, but the fact that they're not sitting on their hunches at all, and they're continuing to innovate and move it forward, it just means it'd be maybe quite a while before anyone else can catch up with them. But I do want to come back to this apple Google thing, because, you know, Apple also announced that, you know, without a ton of fanfare, that they launched their first multimodal AI model, their first large language model, basically, but it's also multimodal, which includes images and text. And they launched that this weekend, it's supposed to it will compliment you know, Siri and some of the other things they do. But to your point, I think Apple also knows that they're behind in this race. And they don't want to limit their ability to put AI into their hardware on their own ability to create really top notch models. So this partnership with Google is a big deal. And I couldn't agree more with what you're saying about the integration piece being key to this. So you know, I've been playing with Gemini a lot over the last couple of weeks. And some of the killer features which seem really obvious when you think about them is you can export from Gemini directly to a Google Doc, with one click, you can export from Gemini directly into a Gmail draft. With one click, you can use Google inside the Gemini interface to double check the answers. So basically, if you have something that comes out that is a, you know, trying to be factual, and you're like, Well, I'm not sure if it's factual, because large language models aren't always you can use Google to check if it's factual with one click. So these kinds of integrations are amazing. And if you add the fact that, you know, what we all do all day, many of us is use Gmail and G Suite and drive on our iPhones and Macbooks, then the combination of being able to authenticate as you're saying, get into the AI features through a Google or an apple login, and then get out I think this is maybe even a bigger deal. Take the results of any AI search, whether it's, you know, text or multimodal, and actually have that go into your, you know, iPhone photos app or having it go into your Google Maps application. Like this is the way that AI is actually going to become part of our daily lives outside of the having to search it out and go to a chat GBT or go find Gemini or go find an interface. It's going to be baked into the systems we use every day, every hour. have every day. And that is how it's gonna happen. And the idea that Apple and Google are teaming up so that this will work with both of their systems is pretty amazing. So I'm really excited about this partnership. I hadn't heard about it until relatively recently. And I think that it's a very smart business decision. And it's kind of an unusual one for Apple, where instead of trying to verticalized and own the whole stack, they're partnering with an industry leader. I will also toot my own horn briefly here. I've been saying for quite a while that Google is going to overtake open AI. And I feel like the news this week is starting to end that this might be possible for all those reasons, right? Google is integrated. They also have incredibly large teams, we saw Microsoft hire the co founder of DeepMind, you know, which Google bought as their head of consumer AI. So even Microsoft is trying to sort of build on Google's models in some ways as their main competitor. It's a really amazing moment. I will stop there. Are there other things? Let's keep going around the world? Are there other things in the AI world that you are finding interesting this week?
Ben Kornell:Yeah. I mean, two other main takeaways this week one is the the robotics combination with AI, there's a demo video of basically, a robot version of chat, GBT. That, of course, has everybody. Very, very excited. I think the reality is that we've seen spectacular demos that still feel like far away from reality. But I think it's reasonable to say within the next 10 years, it's feasible that there would be domestic robot assistants that are voice activated. The voice activation part, you know, we've seen with Amazon Alexa, we're used to voice activation in the home. And now you've got machines that basically can take voice commands, turn it into the coding that would be required to automate. And so then the question is like, is it better to have an all in one robot that does everything and is expensive and large? Or is it better to have, basically, your Internet of Things, devices now all have, like, level up to have some level of robotics in it. The other thing I would say is, there is a growing consensus in the investor community that the winds of the AI revolution are going to the cloud and Chip providers, not to the models. Reid Hoffman just sold inflection, his big mega AI project, for a bag of beans to Microsoft, it was definitely one of those where they had raised hundreds and hundreds of millions of dollars at billion dollar valuations. And it became clear that they were losing the race, that they had an incredible team. And if they, you know, were to let it fall apart, you know, all the value would be destroyed. And so you're seeing the field clearing a little bit. But you're also seeing that this race to build the biggest baddest next model is actually a race to nowhere, because the costs are going down. The open models are always just one step behind breathing down your neck. Yep, will GPT five be that much better than four, perhaps. But once you get to 6789, you start reaching acid total value. And at that point, it's actually about like, very specific to task training datasets. So I think overall venture space is starting to get really, really concerned with these hyped up AI companies that have not really shown the economics that would lead to them being, you know, billion or trillion dollar companies. And meanwhile, it's the old blocking and tackling in Silicon Valley. Let's make chips. That's what's stealing the show. So that's kind of what I'm hearing and watching. Yeah,
Alexander Sarlin:I think that's a great take. And you know, just speaking of the open source model, the last thing that I think is worth mentioning in this space, there's obviously a million things that are happening is we saw a year ago, I think that you know, meta basically made a business choice early on to say, well, we have our we have a great AI team. And we have some great large language models. But instead of trying to get in direct competition with the likes of open AI and Google, we're going to put our model out, make it open source and allow people to build on top of it and as a way to sort of disrupt the field. And you know, I don't know exactly, I mean as a way to sort of, rather than getting into a head to head competition, use their technology for it to sort of grow the field and change the way it develops. And this week, you know, you mentioned x ai X, formerly Twitter, Elon Musk's company, just put out the biggest open source model that we've seen in a long time and I saw a laminae which is a company I follow in the space Sharon Zhao, a really, really amazing entrepreneur. comment about this that Really the grok, the open source model that sai put out, is really quite amazing. And it's going to by itself, you know, accelerate quickly the ability for people who are building on the open source models to catch up and to stay, you know, one step behind, as you say, or maybe even almost dead, even with these incredibly fast moving companies. But that said, I still think the difference is going to be how they're used to, you know, your question about the Internet of Things versus a single robot is a fantastic question. But the chances of us using AI in our MacBooks in our, in our applications that we already use, that's like a sure thing, the chances of having a robot or are things connected, you know? Yeah, it's sort of like one of these is going to happen, period. The other is, hey, if things roll the right way it might happen. But you know, I think back to like the Microsoft Surface, or some of these things that are like really the roll up screens, right, that, you know, the idea that I think Samsung came out with, and like these things are really science fiction, and they're amazing. But if they don't have an obvious use case, that like fits into people's lives, they may not grow that fast, whereas AI fitting into all the tools, and we're constantly using, it's just like, done deal. So I think Google is still going to keep the advantage because
Ben Kornell:this wife said, like he who owns the hardware actually has real leverage. And your point about the Google integration with Google suite, and Gmail, turns out, those were great investments. You know, 20 years ago, when they launched Gmail, it almost feels like going back in time with our next topic. It's like, we're talking about the future where everything is going. And then our next topic is higher education, is going back to the LSAT, they really couldn't find anything better. And they're like, please help us because we now have no affirmative action, we have no way of assessing or determining, you know, objectively, anything, basically, in a student's application, AI is providing maximum opportunity to cheat and write your essays, right. So please give us this old school LSAT. I've been fascinated by it. Because there's almost a rewriting of the narrative, that LSAT is actually more equitable than not having it. And that was the whole reason we got rid of it in the first place. You're the higher ed expert, what's your take on the SATs, making a comeback in higher ed.
Alexander Sarlin:I mean, I think you're sort of zooming out and looking at it in the context of schools are probably at a record low in terms of meaningful signal and trying to figure out I mean, elite universities, very selective universities are getting more applications than ever, they're in some of the lowest percentage, you know, acceptance rates, yet, some of the signals they've had for a long time are going away. And you saw Dartmouth brown Gale, you know, recently started to say, hey, let's, let's go back to the good old LSAT, at least it gives us a number that we can compare to another number and, you know, gives us something that we can work with. I mean, I've always been a little bit confused. You know, I hope this isn't a horrible thing to say, but by the the narrative of the standardized tests being super racially charged and sort of creating inequality, just because I mean, not that it's not true. But just because, you know, the whole point of a standardized test is to do your best and put all of this money and effort and research into trying to a tests that are truly standardized. And, yes, they might not be perfect, but if you take them away and go back to other signals, well, those are a whole lot more stratified. If you're looking at what sports people do, you know, that's a lot more stratified by if you go to a private school, you're taking lacrosse, you know, if you go you go to a public school, you're more likely to be on a basketball team. It's like those singles are a lot blunter than the difference it is sad. So I've always been a sort of apologist for standardized tests. So this is not exactly bad news, in my personal worldview, at the same time, it is really interesting to see the landscape changing in this way we one entrepreneur I've been following I really like and we interviewed her for the podcast is coming out soon is a woman named Julia Dixon, and she has a company called Sai. And it's all about how to tell your story in an application process, especially for a college application process using AI to support you now it's not about AI writing the essays for you. It's about AI helping you draw out some core pieces of your unique narrative that you can then leverage. And one of the things that she says that I think is really important to note is that, you know, yes, affirmative action is off the table, but it is also constitutionally protected to talk about your personal life lived experience in any other kind of application, you know, material including essays, but you can imagine from an admission standpoint, you are that's not that easy to do, right? You're getting so many you can't compare one essay to another in a very clear, clean way, they're not standardized. So you can have somebody write really beautifully about something that may or may not be aI generated, you can have people write about things that are related to their race and ethnicity. And you have to decide what to do with that it is a messy, I would not want to be an admissions officer right now is I guess what I'm trying to say. And I think maybe them throwing out some of these schools throwing up their hands and saying, Oh, can we please just go back to the time when we could compare, you know, an 1800 to a 1900? Or I don't even know it scored anymore, is it back to 1600. But like, you know, I get why they're bringing it back, even though it's maybe a little bit sad. And this whole test optional movement, I've been excited about it. But at the same time, like, you know, I don't think they had a good replacement, as a very clear numeric signal for the LSAT or AC t. And therefore, that's why they're sort of finding themselves in this position of having to regress, that's my read, there's sort of no winning in a situation like this. college admissions is such a disaster. In my mind, I just think it's so awful that we have a system where these small number of elite schools have to choose between unbelievably talented, incredible students from all over the country and world. There's no winning and admissions, but I guess, you know, all things being equal. I'm not as sad as maybe others might be about the sad coming back into the mix. Yeah,
Ben Kornell:I think we've made a lot about the EdTech winter, and how hard it is to be an edtech company. But I think we should also acknowledge how hard it is to be an administrator and take well systems and in higher ed systems right now, because of all these conflating components, and the real challenge of getting meaningful signal about what kids or students need. And you know, what the right fit is, are set of solutions. I think speaking of that tech winner, I think we're also seeing the m&a landscape, just laying bare what we've known for a while, which is that valuations are meaningfully down, and that the kind of golden age that we had during the pandemic abruptly came to an end. And that's with Udacity being acquired for a fraction of the valuation previously, I'm also hearing rumors that epic from buy juice is either going to be sold or is going to be wound down. I can't confirm that. But, you know, word on the street is that, you know, we did this big expos a in the newsletter, word on the street is that the team has that has been keeping it together admirably is finally you know, falling apart and the lack of investment for a period of time because of by juicers, the parent companies need for cash has really started showing. So you know, Udacity, huge, huge player, market defining player in the space for many, many years is now you know, I believe it was an $80 million acquisition. So now really a shell of itself. And it makes you wonder about all the other, you know, quote unquote, unicorns and where where they stack up. This is a little bit of your beat. Alex, what's your take on the Udacity? Exit?
Alexander Sarlin:I don't have a lot to add here, you know, per TechCrunch, they basically were saying that there were there were rumors about Udacity being in talks with upgrade in India at an $80 million price tag. And even though Accenture that bought them for part of their learning suite, didn't name the price, you know, it's probably safe ish to assume that it's around that, and this was, you know, Udacity. So in my back in my Coursera days, you know, it was the thinking at the time was it was Coursera, edX and Udacity in this three way, you know, head to head race of who could you know, own the MOOC, so to speak market, the online course market and Udacity was the first one to get get that unicorn status, that $1 billion valuation, it was back in 2015. And it happened right after Sebastian Thrun had decided to sort of turn his back in a lot of ways on universities and instead embrace companies and corporate entities as the main use of course creator community for Udacity and at the time that felt like a pretty bold, somewhat innovative somewhat, maybe shooting yourself in the foot choice. We were all sort of trying to figure it out. You know, colleges have their ups and downs. They aren't always the fastest movers but they are very, very high quality. Very, very smart people. There are benefits and amazing brands right and benefits to working with them. So if you just flash forward from 2015 there to now almost 10 years later, you have Coursera which is IPOs but since its IPO its stock price has gone to a small fraction of what it is I think it's down like 80% you see hopefully they'll back up in my opinion but you also see edX which was bought out by to you for seven I think 100 million dollars, but then you know, has caused a lot of chaos and to you is arguably on the verge of bankruptcy, we don't know what's going to happen next, even though it's now called edX. And then you see Udacity, which if this is right, that is sold for $80 million, even though it were had a $1 billion valuation 10 years ago, I think Udacity is content is amazing, and has been for a really long time. And they were very smart and what they were building, but it just shows that, yeah, the three big players in that, you know, early 2010s, movement, the 2013, through 2015 movement, none of them have quite blown the lid off the top of the world in the way that we all expected, especially those within within that world at the time. And, you know, I do think that there has been, it's an ad tech winter, and it is a little bit sad that Udacity is, you know, being sold for a very small fraction of what it was valued at. At the same time, if you look at the actual numbers of users of certificates of partners, like these movements have continued to grow. And I still think that they're a meaningful part of the ad tech landscape and a meaningful part of the the educational options for people all over the world. But from a business standpoint, none of them have quite broken through to become the kind of, you know, household name, business success that I think they were all hoping for back then.
Ben Kornell:Yeah, maybe were riddled with good mission outcomes, and not great business outcomes, that that would be like definitional of like a not a great investment sector. But I do think this is also a really good chance to say, is this moment different with AI, with what we're seeing in virtual reality, with what we're seeing in terms of this new technology, and its intersection with teaching and learning? Can we do teaching and learning in new and better and more engaging ways, you're going to have to count me in the optimistic camp on that. And I think the earlier days were about new methods of delivery and new channels. And now we've got new methods on top of these new channels. So I'm very excited. And with that, I think we should transition to our two guests who will tell us a little bit about that future. We have Pat Yongpradit, who is Code.org Chief Academic Officer, but also a TeachAI founder. It's a coalition that he's leading. And then Quinn Taber from Immerse, we are super excited to bring them here. So we'll move over to our guests. Hi, everybody, we have a special guest today on the pod. I am super excited to welcome Pat Yongpradit, the Chief Academic Officer of Code.org, and also the lead of TeachAI. Welcome to Edtech Insiders,
Pat Yongpradit:Ben thank you for having me, Alex, thanks for having me.
Ben Kornell:Pat, you wore kind of two hats on this interview. One is the CAO of Code.org. And the other is leading this collaborative around TeachAI. Before we dive too far into that, tell us a little bit about your journey and how you came to both these roles.
Pat Yongpradit:Wow, I don't know how far to go back here, Ben. Let's see if I can go back to the starting point. And then be fast about it. It was fifth grade where I was on a apple to EA and I learned how to program and logo. That was an experience that was very special. I wouldn't say it changed my life. But it definitely created a foundation for me and computers that extended into high school where I went to a special magnet program in Maryland for math, computer science and science. I took computer science didn't like it, took it a little bit later on. And it was about modeling and simulation, and I loved it. Unfortunately, no one told me that there were careers in computers. So it could be there anything at that time. So I studied neurobiology in college, volunteered in a special education program realized that was my calling. And then I taught for 13 years Special Education science and computer science, middle and high school. And then now I've been Ecuador for 10 and a half years leading a variety of things from curriculum development, professional learning, dabbling in government affairs, international development, and now spending the last at least a year focused on this initiative called Teach AI. That's a great background
Alexander Sarlin:and you're certified in many different kinds of teaching. You've been in the classroom, and then you've done all of this very scaled education for coding and now AI. One question I have for you because you have such a diverse set of experiences in this space is we're at this crazy moment where AI can code. People can learn to code and then make AI code and AI really have this funny relationship with each other. And as somebody who thinks a lot about computer science, education and AI education, I'd love to hear you talk out, you know how they go together and how people are predicting that the future might combine these two really interesting interrelated fields.
Pat Yongpradit:Yeah. Okay. So let's start off with just the AI side of things. We know that most people out there, I mean, the podcast is called ed tech insiders think of these things like AI is an educational technology tool, right? And it is totally, I mean, we have to use it safely and responsibly understand the limitations and applications, the ethical social impacts. But then there's this other side of AI, learning how it works. And some people think learning how it works. That's just for like, super technical people. But actually, we learn about how our bodies work, we learn about how the world works, we even learn about earth signs and things like that, and chemistry and all that. So, I mean, our world is super digital AI is a big thing. Let's understand how AI works. And that's where computer science comes in. And then in terms of AI and computer science education, specifically, teach AI is actually working on guidance on the future of CS education in an age of AI. And we are exploring things like what it can do for broadening participation in diversity in computer science, really lowering that coding bar, and making it a more creative tool for everyone. What it can do for developing higher order thinking skills like computational thinking, and just in general, like helping people understand how these things work at a very fundamental level, so that they can make better decisions about them.
Ben Kornell:Yeah, one thing that we've talked about offline, and I'd love to give you some space to talk about it here. We're also excited about the potential of AI and the possibilities that it can unlock, but also some of the challenges that it presents. At the same time in this hype cycle, there's a bunch of misconceptions about what AI is what AI does, what it can be. How have you been thinking about that? And how are you educating your constituents about these misconceptions?
Pat Yongpradit:So my constituents are, you know, executive branch, folks like governor's and prime ministers, state superintendents, district, superintendents, administrators, principals, teachers, and then even a little bit of parents and students as well. TeachAI's mostly works is focused on folks outside of the classroom, because there are a lot of partners who work inside the classroom. And here are the some of the misconceptions that we're actually addressing and a policy resource we're going to be releasing pretty soon. But the idea that AI is a black box, or generative AI or deep learning, you know, creates these black boxes. And you know, black box is a term that's not exactly accurate, because it's not really a black box. It's like an opaque box. And when we think of it as a black box, we think, oh, there's nothing. We don't know. We can't know how it works. It's magic.
Ben Kornell:Yeah, it's it makes us the victim of whatever comes out of it rather than empowering. That's right.
Pat Yongpradit:And it's just, it's just opaque. And yeah, explainability is hard. Transparency is hard. But when we use better terms, we alleviate and address certain misconceptions. Another one is hallucination. Now, again, that was a term that folks picked over the idea of like maybe confabulation, or error or whatever. And hallucination became the idea. But not only is it co opting language from like, mental health, which by the way, leads to like the anthropomorphize ation of AI, right. But again, it just doesn't really tell you exactly what's going on. And it leads to misconceptions that AI can think thus, it can hallucinate, right? And it can have dreams, I guess, in visions, it's just not true. It's just literally hallucinations are literally just how general AI just works. And it shouldn't be called hallucinations. It's just it's a feature, we actually want it in some cases, and we don't want it in other cases.
Alexander Sarlin:And it's really interesting. I love hearing you talk about that. And I'm going to use the word hallucinations in a very different way, or maybe not use it at all in the future, because I like the way you're thinking about that. You know, one of the things that I find very interesting about TeachAI don't know if our listeners know a lot about it is it's a really very collaborative approach between some very big names in ed tech. Can you tell us a little bit about how this coalition has sort of come together? And what your overarching goals are to bring all of these sort of heavy hitters in education to help influence AI policy? Yeah,
Pat Yongpradit:I'd love to tell the story. So first, there are five organizations that steer TeachAI Code.org, ETS ISTE Khan Academy and the World Economic Forum. These are folks that long had positions of thought leadership in the world of tech literacy and technology in ed tech. And independently, these folks were working on different things in different aspects of the AI and education puzzle. And Code.org when they decided, You know what I mean, basically, we were responding to lots of inbound requests from our Are constituents people who use Code.org People who work with Code.org education agencies, ministries of education as to like what to do per AI, and we decided that we needed to bring folks together, we couldn't handle these things on our own, we shouldn't handle these things on our own. And so we brought these groups together, along with 90 Plus Advisory Committee organizations, including Microsoft, open AI, Google Mehta, Amazon, etc, etc. policy organizations like CCSSO, as a NAS B NSBA. I could go on and on. If you know any of those acronyms, then good for you. But basically, just think about it as there are a lot of these policy organizations, nonprofits, big tech companies, you know, policy, advocacy organizations, research organizations, think tanks, etc, that are all interested in this space. We brought them together to actually do work, and not just be a names on a website, but actually split up form workgroups around key projects and actually inform the projects.
Ben Kornell:Can you tell me a little bit about the vision of TeachAI and what the world looks like if you achieve the vision? And, you know, I think weaving in where you think AI and AI literacy is going would also be helpful for our listeners to understand because they're, they're certainly linked. Yeah.
Pat Yongpradit:So TeachAI's mission is to guide education leaders and policymakers around how to rethink education in an age of AI. Now, that's a term that we hear often rethink education blankety blank, and then especially in an age of AI, but that really alludes to one of the key issues with the key things with Ed Tech that we just do the same thing, but in a faster or cooler way, or whatever. Whereas with AI, this is yet another window of opportunity to make real change in education. So that's the vision of teach AI, real change in education, which is Ben's vision, Alex's vision, everyone else's vision as well. We just want to use this AI moment to do that. Now, it's a big moment, because it's a, it's a moment of professional learning for everyone. It's a super powerful tool. But they're also you know, ethical considerations that everyone needs to understand. Another aspect of that vision other than transformation is that per AI literacy, people understand that it's not just prompt engineering and knowing how to use all these different apps that are popping up. But it's also understanding how AI works at a fundamental level, because people still don't even know how computing works at a fundamental right. So that it's another crack, and making sure that digital literacy is not just consuming and using, but it's also creating an understanding at a deep level. That's the vision of TeachAI.
Alexander Sarlin:It sounds like some of the TeachAI philosophy is really informed by some of the mistakes we have made as a society around computer science education over the last two or three decades. Can you talk to us a little bit about that, and how you're trying to avoid recreating those mistakes with AI?
Pat Yongpradit:Yeah, so mistakes around computer science education, there's actually been a lot of successes over the last decade or two as well. And I think the mistakes really are reflected in the idea that, you know, we could efficiently use computers without understanding how they actually work. And that would be okay. And we don't do that for most anything else. And what I mean by that is, like, even with a car, where a lot of people are like, I don't know how an engine works, and all that you still understand, like the fundamental things behind it, though, right? Like physics, we study physics, right? Yet with computing, like, our world is super duper, duper, duper digital. And most kids still don't really understand a lot of the technical concepts behind these things, just the fundamental concepts. And yeah, that's a problem because they don't have a framework for making decisions around these digital technologies. I'm not saying that the issues that we have with social media are because kids don't know how the internet works, or other things like that. But frankly, there are a lot of things that you can learn about the internet that would help guide your social media use in the same way with AI, like kids dealing with misinformation, honestly, just understanding how general AI works will help them understand that no, this is not like a fact based machine or a fact proxy like the like a search engine, right? It's just math, statistics, probabilities, and, you know, pattern recognition, and knowing that will help them kind of reframe their mental model for engaging with AI chatbots, whether socially or educationally. So the you know, that's a huge mistake that we've made in the digital world, and that we should learn from when we're engaging with AI. So
Ben Kornell:practically speaking, you know, I'm on my local school board. And you know, also there's a bunch of folks who are trying to sit at this intersection of AI AI and the future of education. And it seems simultaneously so big of an opportunity, but also so big of a shift that it's almost paralyzing. If there's one thing that our listeners could do, or if you had a magic wand, like one step that people could take from where they are today, what would that be?
Pat Yongpradit:I'll give two steps. First step, if you're an education system that you you said, you're on a school board call for the development of an AI, an Education Task Force, that I have parents, practicing teachers, other staff paraprofessionals, all that jazz. And along with that polled listening sessions, and then I'll add one more. And then when you're ready to do something, an actual external something, provide just instructional guidance, just principles and ethics statement, you know, double down on academic integrity and be clear like, when and when not, when AI should not be used, things like that. Just like, you don't have to develop big P policies or anything like that. Right now. You don't have to tackle tech regulation, stuff at the district level or anything. There's just stuff that's happening right now that's increasing the AI Digital Divide are increasing the digital divide, but now through AI, and just some light guidance would help alleviate some of those.
Ben Kornell:Wonderful, well, we're going to end it there. Pat, young, pretty, Chief Academic Officer of code.org and lead of teach AI. Thanks so much for joining a tech insiders today, everybody
Alexander Sarlin:should check it out with each AI website. It's incredible. There's so many amazing people and organizations and ministries of education around the world and state ministries involved in this movement. So please check it out. It feels like one of the most exciting things happening in AI education.
Pat Yongpradit:And we're really seeing a policy resource on April 16. So look out for that much.
Ben Kornell:Awesome. We'll see you at ASU GSV then for that lunch. Thanks.
Alexander Sarlin:Hello. We have a terrific deep dive this week with Quinn Taber, CEO and founder of Immerse, Immerse does a really interesting VR based language education and immerse has really been in the news in some really exciting ways over the last few weeks. So we're gonna get into that today. Welcome, Quinn. Hey, thanks, guys.
Quinn Taber:Great to be here.
Alexander Sarlin:That's great to have you here. So first off, you had a little bit of a coup when it comes to Ed Tech publicity a few weeks ago when you appeared on The View. Yes, the view a lot. I think I've met a representative talking about VR and education. That's amazing. Can you tell us a little bit about what that was like? And what it's like sort of having ed tech on that kind of stage?
Quinn Taber:Oh, yes, no, it was sure a highlight. So I've been running a merge for seven years, we started partnering with meta, right around 2020, to really kind of prove out a new use case for being language learning. And fast forward after three or four years of working together on research and really trying to like build out a large audience of VR language learners. We had this opportunity land on our desk, to fly to New York, and be interviewed and represent kind of both the quest threes launch, as well as some of the kind of new features that we've been launching as one of metas top holiday TV placements. If you aren't familiar with the view, I definitely recognize it. But it is the number one daytime talk show. It's five ladies on a panel, the one that I definitely knew is Whoopi Goldberg. So I kind of geeked out when I met her. But I think there's something like four to 5 million daily active viewers of the view, no pun intended. It was. Yeah,
Alexander Sarlin:amazing. And I think you know, one thing that's been so interesting about watching your work was immerse, as you've said, it is a very concrete and exciting use case for VR, that I think people can immediately get their heads around and there aren't frankly, aren't as many of them as you might expect. Can you tell us a little bit about what it's been like being sort of one of the poster children? You know, for VR education? Generally, there's you've raised around recently, you've gotten a lot of publicity. I know you were just touring through Africa. What does that all look like? Yep,
Quinn Taber:I think it really goes back to probably like a lot of founders the origin story. In my case, I was working for a large philanthropist internationally. So I was I had traveled to 50 countries. We were doing a lot of education, philanthropy and investing. And at the time that I was falling in love with how English learning was the key that unlocked visas for so many of the refugees that we were supporting, it unlocked employment and unlocked access to higher education. At same time that I was really geeking out on that my best buddy was becoming a prolific VR engineer stateside. And we were both dreaming around. Wow, VR is about immersion. What is more immersive than language learning? I grew up overseas and I was He's learning languages in a full immersion context. So I knew innately you can't learn Spanish in a classroom much less on Duolingo, you've got to go to Barcelona or Mexico City and talk to locals do the real life tasks. If we could bring that into VR, we believe it would be game changing. Probably the only other thing I'll add real quick is one of the elements I'm most proud of. I think this is why men really kind of taken note is early on, like 2018 2019, we hired a part time and then full time research manager and now small team, conducting third party efficacy research, comparing learning in immerse in VR against a control group in a traditional classroom. And we've just partnered with these research universities now over 30, around the globe. And they've reported back and they've published in like the scientific, peer reviewed, kind of top journals that learning on our platform has a three to 400% boost in all the key what I would call like KPIs, learning, retention, and vocab memorization and satisfied, like all the things that really matter. So I think that's the reason that meta has been like, oh, wow, this makes sense. It's fun. It makes money. But also, it's not a small 10 or 20%, change, there's a step change. It's a right, multiple and improvement. So I think that's the unique kind of cross section that we're playing in.
Ben Kornell:That's awesome. Yeah. So tell me a little bit about on the VR side, why you decided to partner with metta and tell us about like, a lot of people on the business side are thinking about the challenge here of timing, your product rollout with the availability of the technology, how have you thought about that pairing with VR devices?
Quinn Taber:Oh, smart. Yeah, no, we, when we first started dreaming about this end of 2016, we're like, now is the perfect time, VR is about to explode. Seven years later, I'm like, I think I'm still saying that. But we've we've thankfully, we've stuck around, we've stayed alive, we've built the right partnerships. And I actually would now be I'll pinch myself in three years, I would now say that VR is at a series of really unique inflection points, the quest three, I won't give a whole sales pitch. But the fact that it's now AR augmented reality with VR at an affordable price point, with some key new focuses, it's no longer just on gaming, a lot of the mandates within Mehta maybe share that confidential part, but a lot of their mandates are to prove the VR is more than gaming. So that's why they like language. That's why they like fitness. That's why they like one of my buddies, like meditation apps in VR. So trying to build out new use cases for new kind of user demographics, as well as proving out enterprise use cases for VR. And even. And this is the part that can't go too deep into the details. But there's gonna be some great announcements soon education for VR, that Metis pushing, this isn't just kind of our agenda. At the same time, obviously, the Apple headset made just massive waves. But I think even there, the focus is explicitly not on gaming. It's highly indexed around enterprise use cases, but also educational use cases. So I think there's just this convergence around, hey, we're trying to as a whole industry rebrand away from VR. It's for teenage boys to play like lightsabers on paintball. It's actually this is the new era of spatial computing. Yeah,
Alexander Sarlin:I agree. It really feels like we're moving past the, as you say, the sort of teens with a lightsaber stage the gaming phase of VR into a whole new world. Tell us more about what you've been seeing as you travel through this. This growing landscape?
Quinn Taber:Yeah, yeah. No, I think that's that's exactly correct. I think the the key opportunities that now really lay and I'm kind of speaking to my fellow educators, because I know I know the audience here. The key moment is, they're about to be breakthroughs, on kind of all the levers that matter. The first is a breakthrough around the quest three light, which will come out this year at an even cheaper price point, more affordable for educators, the tooling on those headsets to help educators be more successful. There's a couple of the other VR education founders, we put together a list of 10 things incentive to meditate that we need this in order for our apps to be successful in schools. And meta is taking that into account they've got for first time engineering teams and product managers focused on those requests. And now there's really like a litany of VR apps for education that are plug and play, where before it took a lot of just one effort for educators to try to experiment with VR in the classroom. So I would say, start looking into it right now, because all of those factors are coming to a head starting this fall. A lot of them are going to be launched September 1 intentionally. So now is the time to dive in, experiment and get ready because As I think this coming like school calendar year, is going to be the first time that VR has the chance of really entering the mass market and becoming ubiquitous.
Alexander Sarlin:It is really exciting me, you know, my cynical side says it's Lucy and the football we bulb and saying, This is the year this is the year while I remember we talked to Monica are as the next Head of Education. And she was saying amazing things are happening. But there are so many structural systemic issues keeping it from being in schools. And you know, they had been sending out with headsets all over the country. But it sounds like you know, when companies as big as Mehta and Apple start taking very seriously the actual education use case, we might see a real step change. And I am thrilled to hear tell us a little bit about we've been talking about the field, tell us a little bit of what's happening in immerse itself, have you launched new languages, you know, tell us how the product has been evolving and what your response has been from users.
Quinn Taber:Oh, I love this. Yep. So we we started with a conviction that we still hold that we don't just want VR to become like a Duolingo knockoff, where you're memorizing vocab words. Instead, from 2017, memory launched through 2021, almost all of our focus was around innovating the classroom, teacher led experiences in VR. So that's still kind of our bread and butter, I think the things that we've layered on top of that, in the last year or two, the obvious one is AI. So we now have aI tutors that you can practice with outside of your formal class time. So right now, if you were to sign up for a monthly subscription to emerge, you could jump in at any level in French, Spanish and English, to live classes from like 6am to 2am, we take a little break. But there's a live class every 45 minutes every hour, the AI now complements that really well. And really this larger learning journey that can take someone from an language that we would call a one like kind of that brand new beginner all the way through to like, right around fluency. I think the additional levers that have been really fun to see kind of, like our user base really take advantage of is now they're starting just our students to schedule their own meetups that they facilitate in our virtual learning scenes to continue practicing just as peers. So in order to like really equip them, we've been building up things like community tools to galvanize that. Even learning games, three students that students want to meet up and they want to play like a Pictionary in Spanish or any of those fun party games, I play with my buddies on a Friday night, those are now really becoming popular and immerse. I think the last thing that up until now we've kind of waited on is now we are launching. And this is much more to be kind of in press releases shortly. But we're launching an increasing focus on both enterprise sales and higher ed partnerships. So we're for a long time, I was kind of the sales guy with one other person, we're now building out those teams to get ready for metas launches this summer in this fall, so that we have the teams and the kind of the product ready to really start shipping.
Alexander Sarlin:It's really incredible. I have two quick follow up questions on that. One is the AI that is an amazing and very useful use of AI. Because you know, talking to other individual humans is an incredible way to learn talking to an AI tutor can work in other situations maybe lowers the embarrassment factor for certain people when they're starting. Yeah, so that makes a lot of sense. I'm curious, one of the things we talk about a lot on the podcast is the potential for AI to do sort of auto translation. And I know that that's some language learning. I do get that. But I'm wondering if Is there a world in which a global team of people from different countries could collaborate within immerse and it actually you can do the little like Star Trek auto translation and so that everybody is speaking the same language? Yeah,
Quinn Taber:I was. So those tools are basically coming out as we speak, how we're integrating it into the like, learning journey is, if you let's say maybe you're an English learner, but you're from a fairly small country, you pick the whatever it is Colombia, or Poland, or you name it. And the teacher, we're trying to, especially at higher levels, spend the entire class time in your target language, Bollywood, you want to learn English. But if you get stumped, and it's like, oh, the teacher just gave us instructions for this little partner activity. And I can see the Instructions in English they're written down. But my first language I once again I'm making this up Canada is Polish. And this example, I can click an AI is now able to via these plugins that we've been building, translate the instructions, anything that the teacher is leading you through into your first language where before Although we could really do is like Spanish and a couple of large first languages, now we can do it in Polish. So I think that's probably the most compelling, coupled with, it's the same interaction when you're practicing with an AI tutor, that I don't know if we have all the languages launched yet, but that's definitely on the roadmap, that you could have a conversation with the AI tutor in your target language. But when you're stumped or confused, you can translate it into your first language for us. Those are huge unlocks.
Alexander Sarlin:It's incredible. It really is amazing to think about amazing use cases, you clearly thought through it a lot. My other question and maybe this is a little bit of a silly one, but I'm sure you've gotten it before is how people used immerse to meet one another and, you know, start romances or friendships or, you know, do you envision Sunday, attending, you know, the first immerse wedding that was always talked about in Second Life and World of Warcraft? People can be up till Oh, I
Quinn Taber:think about it all the time. I will admit, it is on my bucket list. But I have not been doing a murse wedding yet. But it's gonna happen. I'll call you when I when it does. Yes, please
Alexander Sarlin:do we want to cover that
Quinn Taber:I exactly will live streaming for the podcast. Yeah, the friendship side that was awesome to watch. We have like a discord channel for our users where I read it every day, and get to see all the interactions where we just spun up a Facebook group as well, that kind of has the same same tooling. But ultimately, it's all around, hey, let's encourage and equip our users, our students to really make immerse their own. So watching those interactions, like we have now, two nights ago, I got sent a screenshot from one of our users, there were like close to 15 students, half of which were learning Spanish, the other half are learning English. And they coordinated and facilitated this, like awesome partner meet up like this little language exchange. So there's an infinite number of possibilities there. There really
Alexander Sarlin:is. It is really exciting. And I mean, you hear your enthusiasm and excitement for all the potential and all the things that are already happening in the immersive world. I'm excited about all the work you're doing. Partially from a field building standpoint, I think that immerse has done an incredible job of sort of getting both VR and edtech on to bigger stage around the world already. And I hope you continue to do so because you know, there are only a handful of edtech companies that are household names and it feels like you are a very strong candidate to be the next so always great to have you on the podcast. Quinn, Taber CEO of immerse check out what they're doing and keep an eye out for some of these press releases or announcements from meta that we might see over the next few months that might take VR to the next level. Thanks so much for being here with us on tech insiders.
Quinn Taber:The curtain is a total treat. Thanks.
Alexander Sarlin:Thanks for listening to this episode of Ed Tech insiders. If you like the podcast, remember to rate it and share it with others in the EdTech community. For those who want even more Ed Tech Insider subscribe to the free ed tech insiders newsletter on substack.