Edtech Insiders
Edtech Insiders
Week in EdTech 12/03/25: PhysicsWallah IPO Shakes Indian EdTech, Google & OpenAI Escalate AI Race, Higher Ed Adopts AI, New AI Leaderboard Launches, and More! Feat. Andrew Carlins of Songscription and Eric Tao & Austin Levinson of MegaMinds
Join hosts Alex Sarlin and Ben Kornell as they break down the biggest shifts in global edtech. From PhysicsWallah’s major IPO to Google–OpenAI competition, higher-ed adoption of AI, and new benchmarks shaping the future of learning.
✨ Episode Highlights
[00:07:00] PhysicsWallah’s IPO reshapes Indian edtech and signals renewed global momentum
[00:15:00] Google and OpenAI escalate the AI race with Gemini 3, Nano Banana Pro, and OpenAI’s “garlic” model
[00:31:00] Higher education shifts from AI resistance to AI integration across teaching and majors
[00:39:00] Rural districts test new connected learning models backed by major tech partners
[00:40:00] Learning Agency launches the first Education AI Leaderboard for model benchmarking
Plus, special guests:
[00:45:30] Andrew Carlins, Co-Founder & CEO of Songscription on AI-powered music transcription and access
[01:01:10] Eric Tao, Founder & CEO of MegaMinds, and Austin Levinson, Director of Learning at MegaMinds on immersive AI simulations for CTE and AI literacy
😎 Stay updated with Edtech Insiders!
Follow us on our podcast, newsletter & LinkedIn here.
🎉 Presenting Sponsor/s:
Every year, K-12 districts and higher ed institutions spend over half a trillion dollars—but most sales teams miss the signals. Starbridge tracks early signs like board minutes, budget drafts, and strategic plans, then helps you turn them into personalized outreach—fast. Win the deal before it hits the RFP stage. That’s how top edtech teams stay ahead.
This season of Edtech Insiders is brought to you by Cooley LLP. Cooley is the go-to law firm for education and edtech innovators, offering industry-informed counsel across the 'pre-K to gray' spectrum. With a multidisciplinary approach and a powerful edtech ecosystem, Cooley helps shape the future of education.
Innovation in preK to gray learning is powered by exceptional people. For over 15 years, EdTech companies of all sizes and stages have trusted HireEducation to find the talent that drives impact. When specific skills and experiences are mission-critical, HireEducation is a partner that delivers. Offering permanent, fractional, and executive recruitment, HireEducation knows the go-to-market talent you need. Learn more at HireEdu.com.
As a tech-first company, Tuck Advisors has developed a suite of proprietary tools to serve its clients better. Tuck was the first firm in the world to launch a custom GPT around M&A. If you haven’t already, try our proprietary M&A Analyzer, which assesses fit between your company and a specific buyer. To explore this free tool and the rest of our technology, visit tuckadvisors.com.
[00:00:00] Alex Sarlin: I always feel reinspired when you sort of look at international ed tech because the idea that you can come out and do a high quality test preparation course for something many people are struggling with, get it really affordable, and then just on the back of that concept create this massively successful business.
That's what Ed Tech is all about, right? That's kind of the whole point of it. And when you have a system that has real gaps in it, gaps of affordability, gaps of access, gaps of not enough qualified teachers, you can just do incredible things.
[00:00:30] Ben Kornell: I'm not saying I believe this, but this is the view is meta is soulless and you go there for the money, and OpenAI is like so far valued that now you wouldn't go there, but it's attracting all the like follow on party bandwagon people.
Whereas Google Stock has done quite well, but it's already a big established company. You just never had this flux in flux out, like this rapid flux of talent. So it's looking much more like a, a stable place billing for the long term and willing to invest.
[00:01:11] Alex Sarlin: Welcome to EdTech Insiders, the top podcast covering the education technology industry from funding rounds to impact to AI developments across early childhood K 12 higher ed and work. You'll find it all here at EdTech Insiders.
[00:01:27] Ben Kornell: Remember to subscribe to the pod, check out our newsletter, and also our event calendar.
And to go deeper, check out EdTech Insiders Plus where you can get premium content access to our WhatsApp channel, early access to events and back channel insights from Alex and Ben. Hope you enjoyed today's pod.
Hello, EdTech Insider listeners. It's another week in EdTech final month of 2025. We're finishing strong. I'm Ben Kornell here with Alex Sarlin. We're about to dive into all things EdTech, but first, what's going on with the pod, Alex?
[00:02:07] Alex Sarlin: Yeah. Well, in this episode, we talked to the CEO and Director of Learning for a really interesting company called MegaMinds Ex Google person who started this company that does immersive learning.
They've shown some really effective results in actually getting students to speed up in math, but they're also working in career technical. It's a really, really interesting space, so tune in for that one. That's at the end of this episode, but we also spoke to Rene Kizilcec, who is the head of the National Tutoring Observatory.
He is the head of the Future of Learning Lab at Cornell that's coming out very soon. We talk to Nolan Bushnell. He is the founder of Atari and Chuck E. Cheese, legendary person, and he is starting a company with his head of the Two Bit Circus Foundation, Dr. Leah Haynes called ExoDexa. We talked to Susan Wang and Frank Wu from Aibrary, which is an incredibly interesting platform that basically turns any book into a podcast and then you can sort of tailor it to make it more useful for your exact context.
Really cool stuff coming up on the pod. Ben, what is exciting to you about the end of the year? You know, it feels like 2025 has been a crazy year in so many different ways, but what sort of is top of mind for you right now?
[00:03:13] Ben Kornell: Well, the holiday season is in the air. It's like a great time to have the apple cider on the stove and, and the smells, and we've got our tree up here, but it's also a great time to reflect on where we are with AI and education.
Oh yeah. Many of you'll remember it was chat GPT launching right around this time. Several years ago that kind of launched us into the AI era. And then we have our big episode with our predictions. You know, we have our reflections from 2025 and our predictions. So I've been thinking it's a great step back time at Art of Problem Solving where I'm CEO.
It's also a time where we're stepping back and looking at all that we've accomplished, the impact we've had, what are the priorities for next year. So I just think it's that reflective moment and gosh, we have so much to reflect on. I feel like January feels a long way away, like January, 2025. And it does, the rate and pace of change seems to be so fast now that the Gemini released, there's a new chat GPT release coming out.
You almost can't. Like you and I, we try to keep track of all this stuff. At this point, it's untrackable the nuance between each release and basically people are shipping updates every single week. I think it's like we've just gotten to this point where the change is just baked into the cake. It's no longer newsworthy that there's change.
It really is just how much are we accelerating and what business models will work. So that's what I've been thinking about. I know we've got a lot of news to cover today, but those of you who are listening now, please tune into our year end predictions. We generally do an episode where we've got our predictions, and then we also reach out to all these great luminaries in ed tech friends of the pod and just ask them what are their one or two predictions.
For next year. And it's always so much fun to hear what they're thinking about too.
[00:05:21] Alex Sarlin: Yeah. Last year I predicted by the end of this year, like almost every teacher was gonna be using AI in one way or another. And I think the number we are seeing right now is, I think it's like 86% is the last thing I saw.
So. Not every teacher, but it's pretty high up there.
[00:05:35] Ben Kornell: Yeah, you're just getting close to As in total. Yes.
[00:05:38] Alex Sarlin: Yeah, it's getting there. The other thing that's fun about the end of the year is you get all these sort of best of lists. I've been looking at the best of movies, best books that came out this year.
It's always so fun. And on that note, the Forbes 30 under 30 just came out this week for education. I highly recommend it. There's some great names and some great companies on there, including some really sort of fast moving, interesting companies. We saw Collegio and Royal and Jungle, the founders of all of those companies, study Fetch, many of whom we've interviewed on this show, and many of whom are just doing incredible work that we're gonna reach out to in the new year that we wanna amplify what they're doing.
So definitely recommend checking out the Forbes 30 under 30 in education. That's a fun end of your list. One story that has been absolutely huge in EdTech, but we actually have not gotten a chance to speak about it on this. Pod Ben and I'd love to talk now is the physics Walla IPO Physics Walla humongous Indian Ed Tech IPO launched, just happened a few weeks ago.
It iPod at a huge premium over, its asking price. It's come down a little bit since then, but not that much actually came down and it's sort of come back up. But this is a big deal. This is a big deal for Indian ed tech. I think we've mentioned it in passing, but it's a big deal for venture because it's actual return for some big venture capital firms.
I'd love to dig into it here because it's a story that I think maybe some of us in the US know a lot about Physics Walla. Others might not know that much, but I think it's worth digging in 'cause it's a really interesting story and very this era EdTech story. You know it, it started in 2016, launched its big app and website in 2020, got a hundred million dollars in venture by 2022, and then this massive IPO in 2025.
It's, it's something that sort of swims against some of the currents that we've seen fighting, EdTech and venture. In the last few years. So why don't we start from the beginning. Tell me your thoughts on the physics. Well, IPO we can also sort of lay in some of the facts about the company itself, for those who might not know a lot about it.
[00:07:31] Ben Kornell: Yeah, I mean, kudos to you and to Jen who helps produce our show. I was shocked to read these facts. Alex, why don't you just go, just so people have a sense of the scope and scale. It's
[00:07:43] Alex Sarlin: massive. It's humongous. So physics wall basically started in 2016. It has some parallels to what we saw years ago with Khan Academy.
Right. It was basically a tutor, single tutor who went online, started doing things on YouTube, started making sense of the space. Test prep is enormous in India. Yeah. And he got incredible following and started using it as a launchpad for courses, for basically test preparation courses. The big thing they were doing that was really differentiated physics while at the beginning was the low cost of the courses.
They were as low as a thousand rupees, which is like. $12, and that's a very competitive price even in India where prices are much more competitive and it grew really fast. But the numbers have been absolutely massive. For Physics Swallow, they have over a hundred million subscribers on their YouTube channel at this point.
They have over 36 million students on the platform, and they really expanded, especially in the last few years into hybrid, into physical spaces. And they have these offline and hybrid classrooms, but they use their online presence, especially their social media and video presence to engage students. And I see numbers like, yeah, it's at least 30 something million, maybe as many as 46 million students on the platform, most of whom about 95%.
First engage online through YouTube or through their online courses, but a huge number. Then come join the offline courses and actually go in person. And physics wall has also started to really think about education. In general, you know, it's a test prep company, but they've been expanding and thinking in a lot of ways about how the methods that they've used in test prep can actually influence in-person schooling in general.
And of course, their ambitions aren't just limited to India. They have been starting to look at international expansion in a lot of different ways, but it's really, you know, it's a very Indian story in that it's affordability, it's test prep, which is a huge part of Indian ed tech, and they've done a lot of AI injection in the last three years, as you can imagine, to make their products even more effective.
Which numbers jumped out to you most? Is it those, like, what, 30 to 40 million users, I guess,
[00:09:51] Ben Kornell: situating this in the Indian ed tech landscape? One thing we have to recognize is that the Indian ed tech landscape is defined by massive numbers of users. Because it's so huge and generally low average customer value because affordability is an issue or challenge.
The negative stories about the Indian ed tech market really have to do with high churn and the number of people who sign up for the $12 course and then just churn out so that your lifetime value is also incredibly low. But I think what Physics Walla has demonstrated is by being both comprehensive and oriented towards distinct milestones, which are the tests.
They've really been able to build a more dedicated user base in the sense that sure people graduate or age out, but you know, why wouldn't you take a physics wallet course? If you're going to take that test, it's $12. Why wouldn't you serialize that and do the full program? It becomes like a no brainer. I think the things that I'm thinking about as I look at it are the role of YouTube as a channel.
Yeah. To reach people and to create community. Yeah. Is a very modern thing that many other ed tech companies haven't mastered. And I think the US ed tech market is so B2B oriented that like learning from these B2C players in India, it's quite, quite interesting. Con is also b2c. Yep. Or started that way. So you draw a really good parallel.
I think the second is like. Supplementary education out of pocket is just, is an emerging space here in the us. The idea that you go to school and then you add on extra classes and courses is very normalized in India and a lot of China.
[00:11:51] Alex Sarlin: Yeah,
[00:11:51] Ben Kornell: and here in the US that just has not been the norm and yet we're seeing massive movement in that direction driven in part by new funding with ESAs, but also by dissatisfaction with like where schools are going.
The main thing I would say is, besides that context around Indian ed tech market and some of the parallels in our market, it also shows that if you can build a good business no matter how down like EdTech might feel, if you can build good cash on cash business, there's an opportunity to IPO and there's an opportunity for liquidity.
Deb Qua was at the physics wall IPO, and kudos to her. For investing, not just the money, but the time and energy to really know the Indian ed tech market to figure out who the right winners would be. And this is like a huge win for GSV as a fund and as a firm. And I think it is a really. It's a good lesson that like long-term thinking really pays off in EdTech.
[00:12:56] Alex Sarlin: And also, I mean, GSV has focused so much energy and time on India for exactly some of the reasons you said it, the number of users, the viral growth, the number of young people are just so stellar. I mean, it's worth noting a lock Pandy, the founder of of physics wallet is 34 years old. He's a self-made billionaire at this point.
He has about a half a million followers on LinkedIn. That is really not very large number. This is somebody who's really made this company very much on the back of his own celebrity. Again, we've seen a few, but not that many parallels to this in the US and we saw Sa Khan, we saw sort of Scott Galloway do section trying to build his own educational celebrity.
We've seen some masterclass or some places that sort of try to build this concept of an educational superstar, but we haven't seen that much of it, and I think this is an incredible. Case study of what it looks like to be a sort of educational superstar on a global scale. I think we're gonna see more of this.
I think we might see more of this in South Korea. I always talk about how there's this huge culture of sort of educational tutor stars in South Korea and in in South Asia and in East Asia. I. But it's exciting and I mean, coming off of the global ed tech prize, which is just a few weeks ago, we saw breast teaching, take home an award.
We saw a lot of really exciting things come outta that. I always feel reinspired when you sort of look at international ed tech because the idea that you can come out and do a high quality test preparation course for something many people are struggling with, get it really affordable and then just. On the back of that concept, create this massively successful business.
That's what EdTech is all about, right? That's kind of the whole point of it. When you have a system that has real gaps in it, gaps of affordability, gaps of access, gaps of not enough qualified teachers, you can just do incredible things. So kudos to physics Squala, and kudos to GSV and WestBridge and some of the other people who put money in to this star.
It also debuted at 40 something percent over the asking. It's come down now somewhat, but it's still higher than its debut price. So that was also really exciting to see because Indian EdTech has been battered over the last few years. It's lost a lot of reputation and this is a, a real counter narrative to what we saw with he or should not be named by Jews in the last few years in India.
Anyway, that was an exciting story. The other big news this week that stood out for me is. So years ago in 2022, you just evoked this. Ben Google had announced a code. Red famously announced a code red when chat GBT came out saying, this is something the whole company has to focus on because this is a serious threat to us, which doesn't happen that much with Google.
This week, Sam Altman put out his code red in response to Gemini three in response to Nano Banana Pro, which is kind of blowing up the internet and said, we need to make new models. We are now at serious risk of being lapped by Google, and I'm gonna give myself a mini pat on the back. I've been calling this for a long time.
I really have. I'm like, Google is a humongous company. Very, very smart people. When they corral all their resources to do something like this, they can go, they poke the bear. They really poke the bear. Yeah, exactly. Exactly. They really poke the bear. So OpenAI is developing a new model called garlic, and they're trying to speed up their development.
But Ben, what did you make of that news? You know, that's not EdTech. Per se, but it's obviously underlies everything that we do in EdTech.
[00:16:16] Ben Kornell: Yeah. Well first we often say that Google is the largest EdTech company in the world and it's a rounding error for them. No longer AI and learning primary use case. So we've been living this like search in Google's engagement around AI and learning huge.
And it's been so fascinating and so fun. And again and again, we see this distribution advantage. So that's what you and I have often talked about is like Google's strategy of being AI everywhere. Yep. But now they've got the technical lead too. They've caught up in the race and I think this acknowledgement that OpenAI, which has always been six to 12 months ahead.
Yep. They are now even, or even a little bit behind, just goes to show how once the momentum, it's like the physics of. A place like Google, it took a while for the machinery to get going and get going, but once you have that massive weight of organization, it's huge. I mean, Google's GDP is larger than almost all countries on the planet.
Wow. So that weight kind of coming downhill. And then I think the strategic acquisition of Deep Mind. Yeah, it's going to go down in the annals of business as one of the greatest kutas in the history of business because it almost, they bought themselves like a reinvention engine. That's true. And because Google, if you imagine like how do things work in corporate organizations, if you've got this outlier research lab that's pulling you forward, it really creates a nice and somewhat competitive dynamic.
Now all AI has folded into DeepMind, but for a while, remember we would laugh about the names of their models 'cause they were competing with each other. Yeah. That is really what we're, now that that fruit is here, those trees have borne fruit and now you're seeing Google surging ahead. It makes you wonder what's the staying power of chat?
TPT and of OpenAI. But for those who are worried, Benedict Evans had this great, I reposted on LinkedIn. He had this great analysis. From a usage standpoint, chat, GPT is still, yep. Weeks ahead on like a weekly active user standpoint. It's just not being embedded into workflows. It's still experimental use cases, whereas Google, not only now do they have a technical lead, but they've got this embedded advantage.
It did early days, you and I thought maybe Microsoft and OpenAI would just glom together and become like a mega Yeah. Company and that didn't happen. No, and I think that, I wonder, was that ais, was it that. OpenAI was incompatible with Microsoft, but I think both of them are behind because they haven't figured out how that marriage would work, and instead they're seemed like both of those parties are hedging all their bets by investing and getting entangled with everybody else.
[00:19:25] Alex Sarlin: Yeah. We covered how Microsoft is now incorporating Anthropic into a lot of its work, and Apple has been sort of playing with different models too. Who knows what happened in inside those high level meetings between Microsoft leads and OpenAI? I mean, we know Mustafa Soloman is the head of Microsoft AI sort of came to the work from a different angle as Sam Altman.
My guess it's just a guess is that Open AI was sort of high on its own supply, you know, being the. Being the product that sort of created a new category was the fastest growing product in history. I think they were just like, we can do anything. We really don't have to look behind us all that much. Even by companies like Google, like they weren't feeling like they had to get in bed all the way with Microsoft.
Microsoft put a hundred billion dollars, I think, or more into them. But I think they felt like they could do it alone. And now the question is can they, and your point about usage is important though, because I do think that Gemini is not used as much. It's not synonymous with AI yet the way that Chate is synonymous with ai, right?
I mean, people literally think of what is ai? They think it's chat, GBT. It's like the Kleenex, right? Of ai. Xerox, yeah, Xerox of ai and that that matters. But it turns
[00:20:33] Ben Kornell: out Xerox is not a great company today. Right. Good point. And I don't think this will happen, but they're at risk of being the Netscape. And why did Netscape lose the browser wars?
They didn't have distribution. Right, exactly. And everybody used them, but ultimately over time they got cut away. And Google, this is the scary thing, is that Google has the technical advantage now. And the distribution.
[00:21:00] Alex Sarlin: Yeah. It also maybe begs some of the questions about the distribution, like open AI's models are fueling huge amounts of the ai.
I mean, so many companies are using open AI's APIs, but they mostly using them white labeled. And then Chue itself has sort of created this app store or these GPTs and now apps inside GPT. But that's sort of the other direction, right? That's them trying to be the platform They've created their web browser, that's them trying to be the platform, but they're starting from like a standing start with trying to be the platform.
I mean, not the Yeah. Well, and, and
[00:21:31] Ben Kornell: the models themselves have become commoditized and the switching costs are so low. Right. There's no lock in there, so you've gotta be the platform to win.
[00:21:40] Alex Sarlin: You do. But then they're starting from so far behind a company like Google or Apple or Microsoft in terms of being a huge ecosystem of interconnected platforms, which all of those companies already are.
They're working on it. And I wouldn't count OpenAI out quite yet. I don't think this is sort of, I'm, I'm not, no. Is. A little bit of sort of the natural order being restored here and it is really impressive. I think to your point about Google and DeepMind, like I think there were two moments, right? One was the acquisition of DeepMind, which was a while ago at this point, and that was a sort of strategic bet to stay in front of this stuff.
But then it was the decision in which you mentioned it to sort of shift the locus of control and the center of AI to DeepMind rather than it being out of the sort of regular Mountain View headquarters. And I think that was an acknowledgement that Google is a big enough company, that it needed to have a little bit of a fresh eyes and not be like, we're gonna develop this in this very systematic, corporatized way.
They're like, we have to move as fast as this incredibly small. Startup at, at the time, open AI and philanthropic. So we actually really need to sort of shift how we're thinking about this. And you and I have mentioned, I mean, we came back from the Google AI event a few weeks ago. It's like even within education, there's like 40 different products.
They're just trying all kinds of things in a way that I just am not used to seeing a giant company sort of move that quickly. They're like, let's try this. Let's try this. Let's add this. We're gonna do quizzes, we're gonna do mind maps, we're gonna add video, we're gonna do video in this way. We're gonna do EDpuzzle, we're gonna do, and you're like, that's not usually this sort of like scatter, shotty approach, which there's pros and cons to it, but at least it's a, it means they're moving quickly.
And I think the moving quickly at the model level has created Gemini has created Nano Banana, has created their incredible video tools, and OpenAI is scared.
[00:23:25] Ben Kornell: The reality is Google had a clear. Linear strategy around cloud, cloud infrastructure around search. And then all of a sudden, Chatt BT changed the game and they were just spitting out new releases all the time.
And we know some people that were in the AI organization at that time, and Google was scared to release things because of the political blowback of like racism in some AI like generator. And so what has changed is they've adopted some of the practices of an open AI where they're giving freedom and they're doing scattershot.
It's reminiscent of the, you know, 80, 20 days where Surge and Larry would say, Hey, every engineer has 20% of their time to work on whatever they want. So they adopted some of those things. The challenge is that I'm not sure that OpenAI adopted some of the strategic discipline that Google had. What's happened, basically, if you look at the time when Sam Altman was like voted out, yeah, he was ousted and then he was reappointed.
I feel like that's another critical moment where he kind of, because he consolidated power in that moment with the failed coup. Then it became we're gonna be startupy, ship, ship, ship, ship, ship, ship, ship. And they haven't really come with a coherent strategy other than we're gonna say yes to everything.
Yeah. And then if you look at Anthropic, they've been savvy. We're gonna go B2B, we're gonna do enterprise AI quality partnerships. And not to say that they're gonna win, but I feel like you can better understand their strategic direction. And so with Google, you're a little bit like, oh my God, I can't keep track of all the releases.
But it's also like, this is cool, this is cool, and it's all here in my browser. And then, you know, you have even the OpenAI chat GBT browser to compete with Chrome. That was like a may fly boom. Here it is. Uh, it's gone.
[00:25:36] Alex Sarlin: I assume that they're trying to get it to get off the ground, but did they pull it over?
I don't even know.
[00:25:40] Ben Kornell: I mean, it's still here. You can still use
[00:25:43] Alex Sarlin: it, you can still download
[00:25:43] Ben Kornell: it,
[00:25:44] Alex Sarlin: but it didn't affect much. Yeah. And then Google got to keep Chrome, right? I mean that happened right at the same time.
[00:25:49] Ben Kornell: Yeah. That's the other thing too, is yeah, they, Google's racking up some like regulatory wins that are a little out of their control.
But by the way, what better case can you make to say that I am not stifling competition than that open AI is opening a new browser against me. Like open AI's done a great favor. Yeah, Microsoft didn't have the rival and so that's why they had all the antitrust stuff. Google, right when they were under the scrutiny has this awesome rival that's like moving fast and it's just like the greatest foil.
So kudos to that team for figuring out where to go. We've said this many times before, we're in still in the early innings of all of this stuff, but you know the Benedict Evans post. Talked about like the current best value is like automating tasks. That's really all that AI has demonstrated value of.
And you need to automate tasks where like perfect accuracy is not required. Yeah. But this idea of, you know, sentient AI or of the platform winner, if that plays out, chat, GPT OpenAI is still super well positioned. If we get to a GI. There's gonna be a lot to do there. And they are well
[00:27:07] Alex Sarlin: positioned on that stuff.
There's also the wearables. I mean, they purchased Johnny Ives company. There's something coming out next year at some point where they're gonna do wearables and try to, and that, that's a different type of platform to compete in that is not fully owned by anyone. I mean, Google has Fitbit, right? There are lots of wearables out there, but like, you know, I could imagine that being a, a new level of ai, commercial AI that changes the game potentially if people are willing to try it.
But we'll see. We'll see if they are. I your point about Sam was really key though. 'cause the other thing that happened during that moment was a whole series of the original folks from OpenAI left at the same time. Right. I mean, Ilya Sr. And Mira Mirati and, and a lot of that original crew left. And I think who knows exactly what they were pushing for, but I think as you said, it sort of solidified this strategy of going commercial, being very.
Startupy sort of moving in lots of directions at the same time and not making the kind of deal with a Microsoft or an Apple, frankly. I mean, I, I don't know if that was ever fully on the table, but it could have been right to be like, if we're gonna compete with Google, we need a massive player in hardware, A massive player in browsers, a massive player in, in devices.
Like they could have gone that way and it, and so far they haven't.
[00:28:18] Ben Kornell: There's a great business book to be written or series of books to be written about, like how this is playing out and who's winning and so on. On a personnel side, you know so much of this stuff too, it's easy to get fixated on these big figures, but there's real teammates working on these projects, like real people for sure.
And there was like a missionary element of the early open AI people through common sense. I got to meet a bunch of people and. There was this real sense of like holding a mission and this understanding of what the future might look like and we've gotta make sure that it's good and right. I will also say that the kind of escalating valuations and like the stock dynamics that have affected the space, I think ruined some of the missionary elements because you're at OpenAI, a nonprofit, but people are now like 10 to 20 millionaires from what they've done.
Like you lose a little bit of your edge there. Meta basically throws money at people Yes. To come work at it. So it's like it's in the bay. I'm not saying I believe this, but this is the view is. Meta is soulless and you go there for the money and OpenAI is like so far valued that now you wouldn't go there, but it's attracting all the like follow on party bandwagon people.
Whereas Google Stock has done quite well, but it's already a big established company. Company. You just never had this flux in, flux out, like this rapid flux of talent.
[00:29:56] Alex Sarlin: Yeah.
[00:29:57] Ben Kornell: So it's looking much more like a, a stable place billing for the long term and willing to invest and no one's buying a small island, but also you're getting paid well and you're doing that.
So we do hear a bunch of people talking about, ultimately the talent war is about like 500 people. It's like a thousand people that really make the difference in this space. And I feel like Google's been solid and because of DeepMind, they haven't lost a big tranche of talent. And by the way, we're, I'm also, I've heard that Anthropic has managed to retain a bunch of people and keep a missionary focus.
They're also raising at crazy valuations. So that could be at risk. But in a weird way, this like business model of like valuations and equity kind of throws. Is throwing a curve ball into a moment that should be about what's the technology of the future and how do we deliver it? Yeah, it's a really great point.
[00:31:00] Alex Sarlin: So speaking of stability versus sort of disruption, there was a series of articles this week that really got me very excited actually, because higher education in ai, especially the sort of zeitgeist, the sort of like public perception is that. Professors are so angry at ai. They think that students are using it to cheat.
They think that it's impinging on their teaching. They think it's ruining writing. We saw like a dozen or you know, big articles in the Times in the Atlantic over the last six months about how it's ruining college, and this way we actually saw a flurry of. I think pretty positive articles about AI in the higher education context.
So there was a great editorial in The Times called I'm a Professor. AI has changed my classroom, but not for the worse. Talking about how a lot of professors at this point are starting to adapt. They're starting to adapt their testing, they're starting to move to a more formative assessment. They're starting to lean into the sort of humanities of the humanities classes.
It was a very nuanced, and I thought very positive take on that. We also saw some really exciting things happening. On the ground. We've seen a whole bunch of schools start creating AI majors, including MIT, the University of Wisconsin. And Madison just proposed moving some of the largest majors to a new AI focused school.
And we saw Johns Hopkins. This was like, just was so cool. Johns Hopkins has this amazing new technology they're putting out about basically surgeon training, surgical training, using AI there, because it's Johns Hopkins, it's very sophisticated technology and they're studying it very carefully and they're already seeing really amazing effects for surgical training.
So we've had this knee jerk reaction and across the entire higher education sector for like two years about what is this thing? Is it gonna screw with us? And now we're starting to see some sort of institutional change and especially the idea that not only are these AI majors being created, they're incredibly popular, people are flooding to them.
And I think they're starting to realize as a sector starting to realize that. This stuff is here to stay. We've been saying this for a long time. It may be threatening to certain aspects of, of higher ed, but it's not an existential threat to higher ed. At least not yet. It can be put into the process. It can be incorporated into some of the existing structure and infrastructure and sort of systems.
I thought all four of these just were so exciting to see and it was such a, a nice counter narrative to all the miserable editorials we've seen about I'm a professor and I'm gonna throw myself off a bridge if I have to review another AI paper. It's just like, I get it. But there's other things happening here.
What did you make of some of these stories, Ben?
[00:33:29] Ben Kornell: I mean, my overall take is that we're now starting to adapt.
[00:33:33] Alex Sarlin: Yeah,
[00:33:33] Ben Kornell: exactly. The AI moment, and I'll come back to a theme that I've talked about. There was a long time where we talked about internet companies and regular companies, and then it just ended up being companies, right?
And I think that we're like maybe a year away from like the AI being irrelevant or like, my kids use AI for this. My kids use it. My kids use technology for this. My kids use, like I have them do the assignment at home. I have them do it in class when it's proctor could end up, meaning they can use AI on that and they can't use AI on this.
We're starting to see the practitioners adapt to reality. And by the way, this is why I love teachers. Teachers are the most adaptive. They understand that context for their kids matters so much. And so while I think we give a lot of volume to the Luddite or the resistors or the naysayers, and by the way, also some of the adaptations are moving in a Luddite way.
I'm using it less screen time. I'm doing more books. But on the whole, teachers are constantly tweaking and adjusting whatever they're doing year after year, they're getting better at finding a way to connect kids and learning, or in this case, university students. So I think that's great. On the rebranding of courses as AI courses, I kind of see through it and I'm like, come on in.
Like are you really? This is a money grab. And I feel for, I was at home for Thanksgiving and you know, some of my nieces and nephews are in college and it's so hard for them to know what's the job market gonna look like? What major should I have, should I not have? And you have a university pitching do this AI major.
Do you think any of those kids are gonna be getting a job at OpenAI Google? Or maybe, maybe. I mean this is easy. Maybe could be mit, they're very good schools,
[00:35:33] Alex Sarlin: you know?
[00:35:34] Ben Kornell: But could they have gotten that job with a math major? Probably for how much of this is substance versus how much of this is branding? I see you universities, I see what you're doing here and I'm not sure that it's good for kids, but I'm watching you.
[00:35:51] Alex Sarlin: That's a very fair take and I'm sure there is definitely some people looking at some of these syllabi and being like, if you add one module on AI here, we can rebrand this course as AI applications of AI rather than computer science 3 0 1. I mean, no question that is happening. At the same time, I'm gonna take the ProView of this.
I mean, for one thing, this article in The Times is by Natasha Singer. We've attacked Natasha Singer. I personally attacked Natasha Singer for being such a naysayer on ai, but she's basically saying, here, look, computer sciences majors have gone down. 62% of computing programs have reported declines. And meanwhile, this AI is shooting up.
It's definitely a time of great uncertainty, but in a time of great uncertainty, it's good for the educational institutions to admit that there's uncertainty and that, that we should choose some changing, right? Like, like I, I think it's probably good for the world if uc, San Diego or South Florida is mentioned in here, or, or MIT or or Wisconsin are saying.
Kids are here 'cause they want ai, they're trying to futureproof themselves, they're trying to get ready for a new world. We could change the names of some courses. We can try to, you know, do a little bit of surface a facade. But at the same time we have to grapple with this in a real way. And try to imagine, and we should look this up, I don't know the numbers on this, but how long after you know, computers were created?
Did they have the first computer science major? It's probably 15 years. I'm not even kidding. I mean, it is a long time. And here we are, three years in almost. I I, right. And And you're having some of the biggest brand name schools in the country create new majors, new programs, dozens of new certificates.
Like that's progress. That's progress. They're responding to market demands in a way that we don't always see at that speed for higher ed. So I wanna give 'em some props for that, even if some of it might be a little bit of a show.
[00:37:35] Ben Kornell: Yeah. And I think both of these things can be true at the same time, and. I do wonder about what would be thrilling to me if there was a university that took a total counter bet and said, we're not gonna do any of those things.
'cause we believe in our education program because AI. Whereas the headline thing may be the sexy topic, but what's really matters is can kids do X, Y, and Z? And which is timeless, and we've always believed in it, then there's choice. But if it's good, like I worry, this could be a rush to the fad of the rebrand.
And I, I
[00:38:15] Alex Sarlin: worry about what schools have said for, for 50 years. They said, oh, we're gonna ignore everything happening in the world because we believe in our, that's seeming more romantic to me now, Alex.
[00:38:27] Ben Kornell: I, I agree that I criticized it before,
[00:38:30] Alex Sarlin: but now I'm taking the other side. Okay. And I'm sure people are doing that.
I'm positive you could find some college presidents saying something very similar.
[00:38:38] Ben Kornell: It's probably not in the news. It's true. On the K 12 side, uh, you know, a couple things in the news. One is an apple backed classroom is focused on bringing top-notch education to rural Alabama. I think there's actually a really interesting renaissance in rural education.
This one's called the connected rural classroom. And the idea is around bringing technology that allows people to zoom in and bring top-notch courses, curriculum instructors by things like Zoom. Nothing totally game changing here other than I think where the innovation is happening is developing countries and rural education.
Yes. These are the people that are really frontline of how do we reimagine what learning looks like? Yes. The second one is, which I'm really curious to get your take on, is the learning agency. Yeah. Launched the education AI leaderboard, and we've been calling for some sort of like hugging face of EdTech and ai.
This is it. And it's great that it also combines research around this ASAP 2.0 essay scoring and the ED misconception mapping, which we talked about with Ben Caulfield on on a previous episode. It's really cool to see this. It's dynamic. If you're building, this feels like something to really pay attention to and it's probably also never going to pierce the consciousness of the average classroom teacher, but man, is it like deep ed tech and I love it.
What are, what are your thoughts?
[00:40:14] Alex Sarlin: Yeah, I think it's the beginning of a movement that we've all been really wanting to happen and it's really exciting to see. I mean, I think the benchmarks that they have on right now, as you said, it's the asap, it's an essay scoring benchmark and the ED misconception annotation benchmark, these are coming out of Kaggle competitions that the learning agency has sponsored over the last few years.
So they put together data sets, they put, brought together teams, and they're trying to create these really meaningful benchmarks and I think it's a great start. The thing that's tricky about education, you know, compared to other fields, is that like. There isn't consensus on what makes good education. So what they're doing is cutting it into small pieces.
Right? Can a model do essay scoring? Well, can it identify math misconceptions? Well, can it do this, this, this, this, and this? We are going to be seeing some bigger benchmarks coming out over the next few months. Philanthropy has been putting a lot of money into developing benchmarks. Digital Promise got a enormous grant to help develop benchmarks over the next three to four years.
You've seen a, there. A lot of people are sort of heading in this direction, so that's why I'm really excited because I think the combination of having. There be platforms that look, just like you said, Ben, look and feel like hugging face. You can sort by, you know, which models are the best at doing this, which are the best cost ratio, which are the best latency.
It's like, it's a very technical approach, which is good to what models work for educational. Context. That is fantastic. I think until the set of benchmarks is a little more fleshed out, it's gonna be hard to pierce consciousness of almost anybody in the space, even the model creators, right? I don't think they're gonna sort of scramble over each other to say, we wanna be better on the ED misconception annotation project, you know, index that benchmark because it's so specific, right?
But as we start to have bigger and bigger and more transferable education benchmarks and sets of benchmarks that can work together, so you'd say, oh, let's run this model against the top 10 education benchmarks. And whoever does best is the most education friendly LLM, and they can go tell all the colleges and all the K 12 schools that they're the best on the suite of benchmarks.
That's gonna be a huge moment. And I think this is the beginning of that moment where you start to see a couple of benchmarks in there. They're pretty specific, but as you have a set of them, you're starting to get to this dream that you and I have had for, for years now of how do we actually. Get LLMs to understand what education looks like and compete on being the best educational model.
I'm really excited about it, but I still think it's early days.
[00:42:41] Ben Kornell: Yeah, this is also one where it's just gonna raise the bar or the boat for all EdTech. So maybe people don't design specifically towards these, but if you're trying to kind of come up with the right mix of AI models, what a great guide to have and also great way for you.
To communicate to your customers. Look, we use this as our benchmarking, as an underpinning for
[00:43:06] Alex Sarlin: quality control. Yeah, right? I mean, you're, you're bringing up something really important. I, I'm looking at it from the model perspective, but if you're an ed tech company, if you're a startup, and if you're in the math space, you definitely do wanna use the ed misconception adaptation project to see which underlying model is best at understanding math misconceptions, because that actually is something that will improve your product and allow you to go to your customers and say, we're benchmarked against something very legitimate.
So, no, you're right. Even these very specific benchmarks can matter. Essay scoring, math, misconceptions, they can matter for startups. I think for the broader space, we're gonna need more. But e even right now, if you're think verse right, a a, a math tutoring platform, you want to go and check and see whether the models you're using are aligned to what's happening there, or, or one of many, many different, you know, math specific tools.
I agree with you. So yes, I'm glad you brought that up. I think in the future, this could affect the entire field. Right now it's gonna affect. Subsets of the ed tech field that are focused on specific areas.
[00:44:03] Ben Kornell: Yeah. Well, I think that's a great point to end on. We have a interview coming up here, so we'll do that intro now.
But thank you all for joining the week in Ed Tech. If it happens in ed Tech, you'll hear it about it here on the week in EdTech.
[00:44:18] Alex Sarlin: For our deep dive in the week in EdTech this week, we are talking to Andrew Carlins. He's the co-founder and CEO at Songscription and a current student in Stanford's joint, MA in Education and MBA program, which is a great program, by the way.
He's passionate about music and education, having performed in musical theater and played the baritone, saxophone, and piano in various ensembles, primarily in school settings. Now he's bringing music education to the masses. Really exciting to talk to you, Andrew Carlins. Welcome, ed Deck Insiders.
[00:44:50] Andrew Carlins: Thank you so much for having me, Alex.
I'm really excited to be here.
[00:44:53] Alex Sarlin: I'm really excited to talk to you. So first off, give us a little bit of the backstory. You've been involved in music through your life. We just mentioned you played multiple instruments. What is your musical background and how did it lead you to creating Songscription, which is a really interesting idea to use AI to support music educ.
[00:45:11] Andrew Carlins: My own musical background is pretty unique, and I guess to set the stage, I'm not a very talented musician, but music did change my life. I grew up with a stutter, and interestingly, when I sang, my stutter went away. And it's actually true for many people who, who stutter. They could sing or recite Shakespeare, which has a rhythm rather fluently.
So for me. Music and musical theater is quite literally how I found my voice. I was actually part of an organization called Say, or The Stuttering Association for the Young, which is a nonprofit that I benefited from, who connected young people who stutter with the arts and used the arts as a form of empowerment.
So flash forward, I did musical theater and baritone, saxophone and piano all the way through college. I went to Duke for undergrad, and there they have the Duke University marching band or dumb. So I was a proud dumb saxophone player. And then eventually I put my instruments down because again, I wasn't particularly talented and it was really hard for me to find notes to the songs that I wanted to play.
So either the songs that I wanted to play, which on piano, maybe I could find after a few hours of searching. But then it had to be at my level, and I wasn't very, very good. And on baritone saxophone. I remember in high school spending hours writing down notes by hand, note by note and transcribing. And I just, it was too much effort for me to continue the instrument.
And so when my co-founders and I. Two or three of us met at Stanford. We were in a class called Lean Launchpad for education, and the purpose of the class was to try to build a startup to solve a problem. And we realized that music transcription was a bottleneck across all of music, music, education, music creation, music learning for the hobbyists.
And we saw this really cool paper that had just come out written by Tim Bayo, who ended up being our fourth co-founder, and that paper created the first commercially viable model for automatic music transcription. Flash forward all four of us, Alex, Katie, Tim, and I came together and founded Songscription with the mission of making it easier to learn any song on any instrument.
In the format that makes the most sense for the individual.
[00:47:10] Alex Sarlin: So I mean, this is really appealing to me personally because I am also a music hobbyist, as you mentioned. I play piano and guitar, and I'm constantly looking online for tablatures or for chords. I use a lot of fake books. I'm also very much a diante, but I love it.
And it's something that, the idea of being able to take any song, anything you wanna learn and immediately get true musical transcription is incredibly powerful for hobbyists, but it's also incredibly powerful for educators and for music education. Tell us about some of the needs that you've seen in the K 12 or higher ed music education world that you're hoping to address with Songscription.
[00:47:47] Andrew Carlins: When we think of education, we do think of it in two regards. One is in the formal education space and the other is in the music, the individual music learning space. Yeah. Respect to the former, it's been like some of the highlights of building Songscription so far has been the cold outreach that we've gotten from public school systems both domestically and abroad.
Interestingly, the majority of our user base is actually not, not in the us it's because music is a global language and people around the world are finding us, and for us it's rather exciting. So we've gotten cold outreach from high schools, from school districts that include elementary schools and also from university programs, and they're interested, or their teachers are interested in using our application to create sheet music tailored to their students.
Makes sense. We imagine a future where an eighth grade band teacher can create a song tailored to the preferences of their ensemble at the level of play. So you could play your band students' favorite songs, and Joey could get the beginner level for trumpet, and Sam could get the advanced level for trumpet.
So it's individualized education at scale.
[00:48:53] Alex Sarlin: It's
[00:48:53] Andrew Carlins: incredibly
[00:48:54] Alex Sarlin: exciting. I mean, it does two things at the same time. It personalizes, you know, individualizes differentiates music education, but it also raises the interest level. It allows band teachers or individuals to actually play the songs they love, the songs they want to learn, not just whatever's available in the store in already written down.
So you bringing interest into the music education space and personalizing at the same time, which is incredibly exciting. I think the sort of secret story of AI is that it's going to completely transform creativity. We know it's out there, but I don't think we talk about it enough on the podcast. It's obviously can change the things from medium, from one medium to another in all these different ways.
Can create video, can create images. This is such a unique and interesting use of AI to be able to create musical notation that is used for educational purposes. Tell us a little bit broadly about how you think AI is going to enhance the creative education space. What does it look like when AI is just a major tool in the toolkit of music educators
[00:49:51] Andrew Carlins: everywhere.
That for us is a really exciting future and where we as a company stand in terms of values and our mission is to empower human creators. And so all of our tools are creating derivative work. So it's taking something that Alex you created or something that someone else created and turning it into a form that allows it to be more accessible.
Right now, you could take your favorite song on piano and turn it into notes, but we're not taking away the process of creating that song. And so our hope from a music education standpoint and also from a music creation standpoint is now Alex, you as an amateur instrumentalist or hobbyist, can actually interact with your favorite artists in a more profound way by being able to bring their songs.
For your instrument tailored at your level into your living room. And similarly, the educators can hopefully advance the interests of their students and keep their students more excited. A lot of the educators we spoke to spend 20, 30% of their time not teaching, but actually finding sheet music and or writing the notes hand by hand.
So for us, it's an exciting future if we could allow the teachers to focus more on the musicality and the human aspect of teaching.
[00:50:56] Alex Sarlin: Absolutely. I have one more personal story. You know, when I was taking piano lessons as a teenager, I would talk to my piano teacher and say, oh, I love this song right now.
I'd love to learn it. And he'd have to go transcribe it. He'd have to go, okay, tell me where it is. Let me go buy the tape at the time. Buy the cd, listen to it, transcribe it, and then transcribe it to your level and bring it to you. And it was a huge amount of work. So the idea of being able to take all of that and do that in a much more efficient automated way gives music teachers.
The ability to do it much more often for many more students and focus their time on other parts of the music education process, which is really exciting, and it just empowers individuals to learn. So one of the other things that's interesting about the music world is that you have huge numbers of students.
It's almost universal that teenagers love music and that young people love music. There's a whole research behind the adolescent brain and music and how they're connected. People love listening to music, but a small percentage of them think of themselves as creators. People can actually play and perform, and I'm imagining that one of the goals for Songscription is that you can change that ratio a little bit, and people can, rather than only listening to music and loving Taylor Swift, or Chapel Rowan or whatever, Sabrina Carpenter, whatever the popular people are, they immediately can see themselves as performers.
People who can actually jump in and sing or play instruments along with these songs. That's a really exciting world as well. I'd love to hear you talk about it.
[00:52:14] Andrew Carlins: Yeah, that's exactly it. Alex, one of the things that. We actually see our, our users coming back most for right now is our piano role feature. And so a lot of musicians are aspiring musicians don't actually know how to read notes traditionally, and so the sheet music alone isn't gonna be helpful.
We do predict chords or have tabs as well that helps a broader audience, but still doesn't help that the profile of person that you mentioned. And so what we created is something called the piano roll, which is kind of like guitar hero for piano. Now, if you put in an audio, a piano, audio, for example, into our website, we actually create a video to shows you the individual notes to press and how long to press that.
Wow. The future for us is being able to level at, make it easier or harder. We'll have hand separations. You could see where you should place your right and, and, and your left hand. We're looking to build a future where Alex, if you want to play an instrument, you could just do it without having to be just reducing the barriers
[00:53:07] Alex Sarlin: to play.
Exactly. And then once you're in and you're excited and you're playing music, you wanna play that, you start to move up that progression and say, okay, well now I can do the piano role. Maybe I should start to learn chords. Maybe I should start to learn notes. Maybe I should start to learn musical notation.
And another thing I I love about AI is it takes the eventual goal and brings it up to the front. You can actually do something really aspirational early, which then creates an entirely conducive environment to education rather than, I mean, so many students in history have given up on playing instruments because they sit there trying to do these three notes songs that they don't know, they don't care about.
They have to play them over and over again. I mean, it, there's so much pain, frankly, associated with learning music. The classic Suzuki with a, with a Twinkle, twinkle Little Star. And the idea of being able to have your first song that you play, be some song that you actually love. Already is just transformational.
So let's talk about equity. This is an interesting aspect of education and music education as well. You know, we talk about piano lessons. Often music lessons are not taught in regular public school as much anymore. They're reserved often for people who can, who have the means to do it privately. This is a really interesting, you know, reduction of cost, reduction of it has the potential to change the way that music education is accessed, especially for low-income learners who may not be able to afford private lessons.
Tell us about how you think about that. How can it ensure greater accessibility and equity in music learning?
[00:54:30] Andrew Carlins: That's one core way, which is we now make music education accessible to people who may not live near, near music teachers and or may not be able to afford traditional music education. And that's a core impact that we're excited about with respect to accessibility.
The other area that we're equally excited about too is around physical disability. So I mentioned my own personal story of music being quite empowering to me because I grew up with a stutter. We've seen a few users come to us who have shared with us that, for example, they were, they were blind and now they were able to take what they created on piano and transform it to note so it could be replayed elsewhere.
We've had other users come to us saying that they had a, a physical or a, a motor disability. And so while they were able to say, create a song on Suno because, uh, they could voice, text, and voice prompted, it was impossible to recreate that in real life. This one individual I'm thinking of is actually a singer.
And so they uploaded the Suno output into Songscription. Were able to get the notes because they couldn't write it down by hand. And then finally were able to recreate their own composition in, in real life. And so for us, the accessibility is both socioeconomic, but also bringing in people who, you know, maybe music performance was out of the question for them previously.
That is
[00:55:46] Alex Sarlin: such a fascinating use case of physical disabilities. Is inspiring. Incredibly inspiring. And then I would never have thought of this, but of course it makes perfect sense that some of the music tools that you can use to literally create music. From scratch in different styles. You can then use a tool like Songscription with that created music to turn it into transferrable musical notes that you can perform on your own or you can transfer to a band or you could even sell as a song that you've created.
I mean, what a crazy world we are entering with what music is gonna look like. So we only have a couple minutes left. I'm gonna ask you the big, big question. You've already answered parts of this, but I wanna ask the big question. You know, Songscription is, is relatively new, but you're getting some traction.
I think you're getting some interest from investors in a world where it blossoms and other music tools also blossom in the AI space. And music just becomes a very different endeavor than it has been traditionally with the, you know, the classic lessons. Some people get lessons, some people, most people quit their lessons, A few people stay with it.
They do marching band, they do various things, you know, the, this traditional path in music in a lot of ways. Or they do a garage band, that kind of thing. Or. What do you envision might change about the music education world? You know, writ large If tools like Songscription and suno and others become just woven into society, it's quite
[00:57:05] Andrew Carlins: tragic for us that this is a statistic from the NAM Foundation.
50% of children will quit their instruments before graduating high school, and 10% of adults currently play an instrument, but around 50% say that they would really want to. And so a tool like Songscription for us is doing two things. It's increasing the top of funnel, so now anyone can learn any song on any instrument at their level.
It's also making it more engaging for people to stay playing their instrument. And so it's. It's improving that middle funnel for that bottom funnel that people, Alex, that you mentioned, that are doing marching band, that are playing in their garage bands for us. Hopefully subscription helps as a productivity tool and just allows them to access songs tailored for their ensemble much quicker.
Another area that I didn't even have a chance to mention was actually around religious music too. We've gotten a number of cold outreach from both nonprofits and also from religious institutions that have weekly needs for. New songs. And so for, for us, a better world is a world with more music. And we hope to create that.
We don't see ourselves as replacing humans and certainly not replacing formal music education that I think in the future will continue to be the best way to become a virtuoso. But we think that many people who could become virtuosos are just not because they're putting down their instrument too early.
And so hopefully we could be part of the change so that fewer people put down their instruments and more people are able to enjoy live performance. I'm
[00:58:35] Alex Sarlin: really excited about this mission and I think it's, uh, just one of the most exciting noblest thing that 50% of students who put down their instruments turns into 10% of adults.
I mean, what a sad drop off of creative energy. So before we end the conversation, is there anything else you'd like to share with our audience about Songscription and your vision for music education?
[00:58:54] Andrew Carlins: Yes. We also allow for MIDI download. And so we envision ourself playing, plugging into the creator's toolkit.
You can then upload the MIDI output in into your favorite doc.
We're looking to work more directly with more schools as we now have better models and hopefully are looking to make impacts on, on students' lives.
[00:59:30] Alex Sarlin: And Andrew, you have some exciting news to share with your audience about funding and how you're starting to bring Songscription to even more people with some investment.
[00:59:39] Andrew Carlins: We're super eager to announce that we'd, we actually just closed our seed round. It was led by Reach Capital and we had participation from Emerge, which is a UK based EdTech future of work, vc also 10 X Founders and Dent and a number of other angels, including Ron Bumble Al, who's former Guns N Roses
[00:59:58] Alex Sarlin: lead guitarist.
That's amazing. Amazing set of investors and can't wait to see how this, the investment is used to scale Songscription and build even more tools for music educators and musicians. This has been fascinating. Andrew Carlins is co-founder and CEO at Songscription. He's a current student in Stanford's MA and and education and MBA program, and he and his three co-founders are building a tool that can turn things into Tablature, into full musical notation, into piano roles, making music education accessible for everyone.
Thank you so much for being here on ATech Insiders. Thank you so much. Welcome to EdTech Insiders. We have really special guests today. We are talking to the CEO and Director of Learning for MegaMinds. It's incredibly interesting platform. So Eric Tao is the founder and CEO of MegaMinds. It's an AI powered learning platform that puts students inside immersive simulations with AI tutors that guide them through lessons and tasks.
He's a former innovation lead at Google, and Eric is redefining how AI can help students learn in more engaging, equitable and future ready ways. Austin Levinson is the Director of Learning for MegaMinds. He's a 20 year educator with a track record of developing impactful, gifted, and steam programs with deep expertise in project-based learning, design thinking, and AI integration in classrooms.
He ensures that MegaMinds delivers meaningful student-centered learning experiences. Eric Tao and Austin Levinson, welcome to EdTech Insiders.
[01:01:27] Eric Tao: Thanks, Alex. So happy to be here. This is awesome.
[01:01:30] Austin Levinson: It's a pleasure to be here as a fan and as somebody who looks forward to this conversation, I'm really excited to be here as
[01:01:36] Alex Sarlin: well.
Oh, that's so nice. Well, I'm really excited to share some of what you're doing at MegaMinds. I've seen the platform, it's really a cool platform and very immersive. So, Eric, for those who don't yet know MegaMinds, and I wanna get their heads around all the things they just heard, AI tutors, immersive simulations.
Give us the overview. What is MegaMinds and how did you come to this idea of rad tech?
[01:01:55] Eric Tao: Okay, great. So high level MegaMinds is an AI powered platform. It uses 3D environments. We have an army of little in world tutors. They're sort of like NPCs, the video game term of them. They're like characters that are placed in a video game.
And what we do with them is we basically, it gives us the ability to create simulations. So what does that mean? That means we can create, like for CTE, we can create a job center and we can, inside that job center, we can place six different little AI tutors and they each assume a character of a different role.
So if you want to interview in the construction, in the trades, you can talk to a manager there. If you want to interview an education, you can talk to a school principal. And we create those characters and we put them into this dynamic environment that it's really. A lot like a game like Roblox or Minecraft, the games that kids, we know 91% of kids under the age of 18 play these games oftentimes for hours a day.
Certainly my two daughters would play all day if I let them. So within these simulations, they give the students the opportunity to build confidence via practice in low stake situations. So for those job interviews, they can go and they can interview at different positions. They can do interview at different roles, and they can actually practice that specific interview for that specific career pathway.
And it's low stakes students, they love talking to Ouris. Some of them will talk to our ais more than they'll talk to a human being.
[01:03:24] Austin Levinson: And that's an important point, is that from the student point of view, many students don't want to ask for help. Students, it is hard to ask for help. It's hard for adults to ask for help sometimes.
And in that space, the students feel comfortable asking for help. Their affective filter, those who know Steven crashing is lowered and they're really much more open to ask for help. And then the teachers are able to actually provide that help or they're gonna get the help within the module. And so the teachers are gonna get information that, ah, this student is struggling.
If that didn't occur, the teacher would've to wait for an exit ticket, a formative assessment. They'd have to actually analyze that, and they might not get that right away. And so it's immediate and it's giving the students the opportunity to develop not just their content area knowledge, not just their conceptual understanding, but also that confidence in the SEL component that goes with it.
[01:04:11] Alex Sarlin: Yeah, so you're putting together a lot of really important aspects of EdTech and education right now. It's future readiness, career preparation, CTE, immersive simulations, authentic learning, and of course, you're using AI to support the whole thing by creating these AI-based characters. When I hear all those things together, that gets me very excited.
It also feels like something that, I bet different schools have very different reactions to this type of learning. I bet some say amazing AI plus career and technical education and relevance and lowering that effective filter I'm in, and others say, wait, AI. I don't know. We have seen a lot of that sort of nervousness around, especially the AI component of this right now.
I'm curious, Eric, when you're out in the field talking to all the different potential customers for this, how do they react to the different elements?
[01:04:56] Eric Tao: That's a really great question, Alex. I think that's something that we, and probably a lot of other companies that are, uh, similar stage as us in using AI are encountering.
So it's very broad, right? So we have some districts even in the same state where we'll get 30 seconds into our demo and the admin will be like, okay, so let's pilot this. And he starts literally charting a pathway to how they're gonna deploy it within the district effort. 30 seconds. We've been on calls like that.
We've been on the exact opposite. Where we're showing the admin and the admin, I can just see them starting to get nervous. And one admin was like, this is really impressive, but there's no way this will fly in my districts. And we're trying to find right now the balance of what, as a company, which districts should we be targeting and where DI districts.
Find value with what we're presenting to them. And it does, you're right, Alex, it comes down to the AI aspect of it. Is the district ready even more so are the parents ready of the children in that district? The kids typically are all ready. They're the ones that are already adopting ai, probably even more so than the educators of that district and their teachers.
And they, and their knowledge base probably extends beyond them as well. But a lot of districts don't have the policies in place. They might not have the teacher buy-in or the parent buy-in. And that's the thing that they're thinking about first before what is actually beneficial or most beneficial for their students.
Unfortunately.
[01:06:20] Austin Levinson: And there's a lot of fear and uncertainty, and rightly so. This is, we're at the precipice. This is a massive fundamental change in how we learn and how we work and how we partner and hopefully think. And so it's not a surprise that there'd be a lot of fear and uncertainty by certain stakeholders.
We've been able to help. Some overcome some of that with our AI literacy bundle that we have. That's a great way to get some of the students and even some of the adults learning about some of the aspects of it. But it is, it is a challenging moment.
[01:06:51] Alex Sarlin: Yeah. But let's dig into the AI literacy. I mean, so you mentioned that there's different modules within MegaMinds, and I know AI literacy is one way you go this, the training and career in technical education is another way.
Tell us a little bit about how you built this AI literacy module, what it looks like in practice, and sort of walk us through what it looks like when a district is implementing MegaMinds. What does it look like when it's going perfectly Right.
[01:07:11] Austin Levinson: Alright. Well, so first I can talk a little bit about the AI literacy bundle where we've partnered with code.org for their hour of ai, which is coming up.
We've also, we're helping out for the AI Presidential challenge. We're a turnkey solution for that. And the AI literacy bundle is a set of modules that walk students through really what is ai. What is it? How does it work? Where does it get the data from? What is machine learning? What is natural language processing?
What are some of the ethical issues? How does bias work? And then eventually, what are some of the careers and some of the career paths related to ai? And then the students have an opportunity to develop a solution to a real world problem in their school, in their community, or even in the world, using AI as a lever for that.
And so that's part of a capstone project that's incorporated into it, so it becomes a PBL. And this whole bundle was designed to help students interact with AI in ways that are gonna promote critical thinking, that are gonna actually promote empathy and social emotional learning, and are gonna also promote creativity and the opportunity to really develop a solution.
So it's, it's using ai, it's digging in with ai, it's struggling, it's failing, it's being challenged by the components of AI in ways that will really. Frame what they're going to need in the future in a professional setting. It's what we use, the three of us, and many of the people are listening using in our professional settings.
We've really designed it in a way that will set them up for success in that setting, and we'll set them up to ask the right questions and, and think about it in that way.
[01:08:45] Eric Tao: And I think what's really unique about our literacy bundle is because we take advantage in all of our modules, we really take advantage of the immersive qualities of it.
So for instance, in our AI ethics room, we talk about bias a lot and misinformation and both the positives and the negative aspects of ai. And we give students an opportunity to first be the subject of bias. So we have an AI character that the student interacts with, and the AI is sort of like a police officer, and it's like, he's like, Hey, I've been watching you, you've been doing this and you've been behaving this way.
And the student's like, what are you talking about? I've never even met you before. Where are you getting this information from? And the student learns that that security officer doesn't have the right information. Is judging the student based on biased information. And so it gives them an opportunity to actually experience firsthand being the subject of bias.
So as a result of improper information. So we're able to really put the students into that experience and let them feel it genuinely. So, and I think that's really kind of what's unique about our modules is we can do that.
[01:09:48] Alex Sarlin: Yeah, you mentioned that sort of firsthand experience and that feels like it's really key to your approach to education and ed tech in general here.
You have immersive, it's scenario based. When I hear you describe that being accused of something you didn't do, I mean that's a very visceral experience, right? It's very memorable. It feels very different than the idea of reading a text about AI bias or taking a little quiz about AI bias, and I feel like that's really core to both of your and MegaMinds in general approach.
And Eric, you mentioned Fortnite and Roblox and Minecraft and some of these sort of gamified environments, your daughters and everybody else's kids are in all the time. As a sort of parallel to this type of firsthand immersive learning, I'd love to hear you talk a little bit more about why. You chose immersive simulation-based, character-based interactions as your primary way to deliver information.
It feels like there's a lot of advantages to it. It also, I bet some people have to be convinced, so I'd love to hear you sort of talk us through it.
[01:10:46] Eric Tao: Sure, sure. Certainly people who may not be familiar with the context. And really, if you're over the age of 25, you probably have never played as a character in a multiplayer universe.
But as I said, off the top, 91% of right of under 25 have and even more shocking that daily active users of these games are on these games for 2.6 hours a day. And Alex, just as a watch out, both of my daughters started playing Roblox at three years old, so Wow. It's coming quickly for your kids. I think the first thing I wanna talk about is just set up the context for these games so they're not just a trend or a fad.
They were a lifeline for these students during the pandemic. It was how they made friends. It was how they maintained friendships. It goes a lot deeper than a game to Gen Z and Gen Alpha. It's creates FOMO if they're not on these games and, and the only thing I can relate it to is maybe social media for the millennial generation.
If you weren't on Facebook back in the day, or if you weren't on social, you felt like you were missing out. And that's really the appeal of these games for this generation as well, is if you're not playing one of these games, then you're missing out on all these social opportunities and potentially the groups that are playing them.
So number one, we started thinking about immersive as well. Back in the day when I was working at Google, I was working closely with the VR team, so that's really where it started. But during the pandemic, I got to see how my daughters, and they were not successful with virtual learning at all, but yet they would play these games with their friends for hours a day, right?
So this was really where the, the sparks started to come together for what we eventually built. So the first thing that we're trying to do with these environments is we really wanna meet the kids where they are. We wanna meet them in the environments where they're excited to be, where they spend the majority of their time anyways.
And I think adding AI to these environments, where we started was creating these virtual spaces for students for learning. And then we started adding AI to these spaces in February. So it hasn't even been a full year yet. And that's when we started to see things really kind of take off because adding the AI to the environments did a number of things that completely changed our value proposition, is it allowed us to create these spaces, these simulations for productive struggle.
Even more than that, we were able to use the capabilities of AI to let teachers see what was happening inside these environments. So, you know, as I mentioned earlier, the students will, oftentimes, they'll talk to these ais where they wouldn't talk to a teacher. So I'll give you a quick example. We did an intervention study with a Title one school school in Florida in the spring in math.
And they had a group of eighth and ninth graders who were 0% of this cohort were at grade level in math. In fact, 83% of them were two grade levels or more behind. Now this is, they were just the unfortunate group who were in the fourth and fifth grade at the outside of the pandemic, their Title I. So maybe they didn't have devices, maybe they didn't have access to virtual learning, but through no fault of their own, they never learned their foundational math skills.
So they're teenagers now and they've grown up their entire lives thinking that they were not. Smart that they were not good at math. These are kids who are not gonna be raising their hands in class. They're not gonna be active learners. What we found when we put them through this intervention, first of all, the results were astounding by the end of the semester.
So by June they started in March, and then by June, 67% of the students had achieved grade level.
[01:14:18] Alex Sarlin: Oh wow. And only
[01:14:19] Eric Tao: 22% of the control group were able to achieve the same. So the students using MegaMinds learned at triple the rate that the students using traditional intervention techniques were able to achieve.
Now, Austin and I were able to do a two hour download with these students afterwards, and what kept coming up over and over and over and what we saw was the students would work with these ais. They had no problem talking to these ais because it was non-judgmental feedback. They could talk to them, they could say, Hey, I don't understand this.
And they didn't have to worry about the AI going, oh, come on kid, you're 14 and you don't know this. You know, it was completely non-judgmental feedback. It allowed them the chance. We heard multiple of these students say, it helped me build my confidence. It helped me get my confidence up. I couldn't do this before, but after working with the ais, now I feel like I could that moment really kind of crystallized things for us and really crystallized what is our value proposition for students.
And that's what exists. Whether it's, whether you're talking about math or whether you're talking about CTE or whether you're talking about special education. That's the through line that we provide is these environments provide a safe space. For productive struggle for the students to try things, to fail, to try them out again.
[01:15:26] Austin Levinson: And from the teacher perspective, there's a lot of talk, and Alex, you know this very well. There's a lot of talk about personalized learning. It's thrown around a lot. And as an educator who worked with MTSS for years, I know the challenge to differentiate is hard. Anybody who says, I got differentiation, I got it under control, they're lying.
It's really difficult to differentiate. What we've provided is a space for actionable space for responsive teaching. So this, the teachers are gonna get almost instantaneous feedback as the students are going through the module and are able to see, and this is what happened in the study, but it's what's happening with, with many of our clients now, they're rolling out, now the students are working through a MegaMinds room and if somebody's struggling, somebody's having difficulty, the teacher is alerted basically immediately they can go over.
And they can find out from the student what's needed. They can take a whiteboard, they can draw some things out, or they can wait until the end of the session and they say, you know what? There's six students. They're having a challenge with the same thing. Boom. That's a small group for tomorrow. And so the MTS has integration and the ability to differentiate and create actionable results the teacher can use.
And I mean, it's really assessment to guide instruction. It's a constant formative assessment. Now, the 3D model is getting much richer data than a chat based model, because if a student turns to look at something. The AI agents capturing that and is able to tell that to the teacher how they're being resourceful.
They're looking at something, oh, they're walking over to the board or the graphic that's inside of the module. They're asking the AI question multiple times. From the student perspective, why it's working is that the AI are keeping the students in their, in their zone, proximal development. The students, if they need a lot more help, the AI's gonna go back and provide some foundational information.
If they need a challenge, they might push them to go further and explain their thinking and make their thinking visible. And so there's a lot of aspects on both sides. I often call it a smart use of AI in, in layman's terms, but it's a use of AI that's really creating a situation where the students can feel successful and the teachers have a thought partner.
The work that they're doing, it's also gonna be able to respond in 90 different languages. Right. So for ELLs that might be struggling, say in math, they're often relegated immediately to tier three. Oh. Can't teach them in English because they don't understand English Tier three. And they actually were good mathematicians in their native language.
Well. This gives them an opportunity to continue to pro progress and continue to feel good about themselves as students In a situation where it's a foreign country, it's a foreign language, they're often in a very difficult situation experiencing culture shock, but they're able to feel successful in something that they were always good at, which was math in this case, and be able to continue to proceed in math despite the language bias.
There's a lot of components to it that lead to those results. A lot of nuance there.
[01:18:14] Alex Sarlin: It's amazing. First off, congratulations on those results. That's really astounding. It really is amazing and one through line. I'm hearing from what both of you're saying and from some of the previous answers, that gets me very excited about the potential for this type of work is that you're able to bring all of these different types of.
Personalities, you know, experts like career training experts, math teachers and tutors, nonjudgmental peers that can talk to you. Rather than having, you know, one teacher, if you're lucky, maybe a teaching assistant in the room, you have this huge cast of characters that the students can interact with and they all have their own personalities and their own way of interacting.
And it creates this dynamic where, you know, students can go, as you say, they can feel not judged. They can talk to somebody about concepts that they wouldn't want to talk to a regular human about because they'd feel embarrassed or they'd feel lack of academic self-efficacy. You know, just feeling really problematic about expressing that they know all the languages, as you just said, Austin.
And I feel like that metaphor of sort of. Supporting a classroom with a huge cast of different characters, even like the policemen you mentioned before, right? The idea of like somebody to come in and play a biased cop who accuses you of something, well, the teacher might not be able to pull that off, right?
But if you have a, a character, you can do that. It's really interesting way to see the sort of role of ai. And, and I guess it leads to my next question, Austin. I wanna pass this to you, which is that, you know, one area of discussion a lot right now is what is the role of AI in relationship to the educators?
You just mentioned a few different ways, right? The data in MegaMinds, every time a student turns every answer, every question can get bubbled up, can be used for differentiation in, in MTSS, it could be used for differentiation in, in any type of classroom. But I'd love to hear just like almost philosophically, how can the ai, this huge cast of AI characters or AI NPCs and then the human cast of educators and support staff and admins and specialists in the school, how can they play together in a way that feels.
Truly positive. And that's immediately removes the concern that some schools have about these being competitive forces.
[01:20:19] Austin Levinson: I'll start by saying that when I've given PD sessions to schools for AI onboarding, I always say that the most powerful piece are the people right here in this room. The teachers. So never removing the human element and the human element here is absolutely essential.
And so I think that the way that the teachers, that the way that educational leaders can interact with MegaMinds as a, an effective tool to maximize student learning, to maximize student self-image, that's a key component. So I, I'd say, stepping back a little bit ideologically, you know, we know a lot of, a lot of cognitive offloading is occurring, A lot of shortcuts are occurring.
AI detection, you know, is not working. We know that it's gonna single out neurodivergent students or English language learners. And so what are effective ways. That AI can be used and we can capitalize upon the power of AI to further student learning, to further curiosity, to further creative thinking, to further critical thinking, divergent thinking.
And what we're doing is we're finding ways to really unlock that potential by having the students interact in a space that they're at the helm. I mean, we talk about, again, another term that's thrown around a lot is student agency, and what does that mean in a learning environment where students are, they're outside of school spending hours in some cases on TikTok.
And there's this massive algorithm pulling them and trying to draw them in these dopamine hits that are occurring. How do you, how do you compete as a simple teacher with that? Well, you need to compete with ways that are going to really bring the student to the fore of what's happening. So for a student to flip through a slide deck.
For a student to watch a video and answer some questions. There's a little bit of agency there. They're controlling the click, but they're not really controlling the, the experience. And this, I mean, a student in MegaMinds is physically is moving around the space. They're turning and looking. They're using 'cause the same controls as they would use in Minecraft, and they're actually moving around the space.
They're clicking on an object and it's giving them information. They're clicking on a question mark and then they'll have a chance to opera, answer that question. They're interacting with an AI and asking for help. They're moving around. They might be discovering some things in the room that the second or third time round they didn't see.
And so there's this real experience of doing it themselves. That's a different level of agency and that is a way to combat this amazing drain of what's going on. And we see it everywhere. You walk on the street, you see people just engaged in their phones, passively consuming. And so that's, I think a piece of it.
I think as well, the teachers as the sage on the stage, and we're the, the goal is to move them to the, the guide on the side. What we're providing in that space is that the ai, in this case, the AI agent that's analyzing all the student movement, is giving the teachers relevant feedback. It's really creating a thought partner for the teacher.
People have asked us, well, is this kind of like an IXL or a a, a program like that? Well, that's kind of a drill and kill and that has its space. But no, this is much further up the chain because it's providing information to the teacher that then the AI and the teacher are really working together to formulate groups to formulate an action plan.
And so that shift in teacher time, you know, there's a lot of platforms that talk about saving teacher time, and I don't like to think about it as saving time. I think about it as a shift of time from getting away from having to do some of the mundane tasks of, of certain grading and certain aspects of grading a math paper.
Well, that is actually something AI can do effectively and incredibly accurately. But instead, it's where the teacher can then take that information and the data that the AI has been analyzing for them and using it effectively to form small groups, to individualize the instruction, to figure out where the student, maybe the emotional piece is affecting the students and blocking them from learning.
And so it's that shift that our teacher dashboard, which is very robust, is providing. And so that's where, uh, getting back to your original question, the ideology, the teacher needs to stay at the center, the educational leader, the, the superintendent, the principal, they need the instructional coach. They need to stay at the center of this process.
And we're keeping them at the center of the process. But the students at the same time are wildly engaged. And so the conceptual understanding, the transfer is being amped up.
[01:24:41] Alex Sarlin: That continuous formative assessment loop where you're finding out both from a knowledge and an affect perspective, where is every student on every different aspect of what you're teaching?
What do they know? What are misconceptions? What are they missing? How can you continue to evaluate that? It's just something that, as you mentioned, no individual human can do that for 30 students all the time. It's just, it's literally not possible. So creates this army of formative assessors that can speak every language, can do all the, maybe army's the wrong metaphor, but this huge support staff of formative assessors that can, you know, figure out where every student is, identify gaps in learning.
Identify those students who, you know, came from another country and knew their math down cold. And it's, what they're struggling with is language, not math. It might take a long time for an individual teacher to suss that out, but they can get it very quickly. It's all, it's very, very exciting, you know, vision of how this all works.
[01:25:31] Eric Tao: Yet one of our teachers describe it as the equivalent of having a one-on-one meeting with every single one of their students every single day.
[01:25:38] Austin Levinson: You know, they're engaging in metacognition and reflection questions. They're asking them about where they still struggle, how this can connect to real world.
And so that metacognition piece is incorporated into some of the questions that the AI is asking the student. And so the teacher not only gets content data and performance based question data, they're also getting metacognition that information that will help them help the students develop them as learners.
It's, it's powerful, powerful information that they're getting.
[01:26:05] Alex Sarlin: Yeah, it, it really is. We only have a few minutes left, but I wanna ask something specific about what you, something you both mentioned earlier, I think it's relevant to a lot of people in our audience. You mentioned that you're, you're part of the hour of AI with code.org and that you're a part of the AI presidential challenge, and there are these initiatives.
The starting right now that are really trying to address, you know, AI literacy, trying to address gaps in learning, trying to sort of enhance the distribution and visibility of some of the most exciting ed tech tools out there. I'd love to hear you talk a little bit about your experience with these types of programs and movements and what it's looking like.
It's gonna mean for MegaMinds in terms of visibility. What does it look like to join the hour of AI and then have this whole initiative at your back?
[01:26:50] Eric Tao: Yeah, so first off, for what it means for MegaMinds as a company is, is obviously it means the world. It basically, it's two orgs and the administration creating what is essentially a giant tidal wave of AI literacy that we can imagine ourselves as like just a little surfboard on top of that wave.
You know, that's our goal. The reality is of what we're finding is, and why I think mega mites might be able to be an interesting offering within this, or o verve, is that what we're seeing is districts are a bit hesitant to participate in these challenges because of a number of factors. Number one, the complexity involved.
So, you know, certainly in the presidential AI challenge, they're asking students to use AI tools. To create a project that can solve a problem that they've identified within their communities. Now, a lot of districts, they're facing two, two main issues here is number one complexity is, is which tool do they use?
Has it been vetted? Is it safe? And then there's the fear aspect as well. Is, is it private? Is it secure? Yeah. You know, what is the data, what are the inputs being used for? And I think that is preventing a lot of momentum or preventing some maybe, uh, more momentum that, that these challenges are seeking right now.
The fact that they're happening right now is obviously a great thing. So where I think that we can play a role in that is the fact that we are that one stop shop. It's a space where students can learn about AI literacy. They can use our tools to create solutions. We've been vetted by, you know, a four month review in New York City schools that the, the Irma process is, is frequently thought of as the, the strictest privacy review in the nation.
We've also been embedded by multiple state orgs as well for the same purposes. So we have been deemed a safe tool for schools to use. So I think we have an opportunity there where we provide a turnkey solution for districts to participate in these challenges.
[01:28:41] Austin Levinson: Yeah, I think as well right now, we know Alex, that there's a lot of companies out there.
There's a lot of different things. And how do you, as an educational leader, how do I sort through, I'm an instructional coach, how do I sort through all of this noise to find really the innovative companies that are gonna benefit our students in meaningful ways? And we've heard from Orange County to virtual school leaders to people in New York City, that MegaMinds is different.
The way we're using AI is a way that is really directly benefiting students in terms of engagement and agency and teachers in terms of being able to differentiate. And so I think that it's giving us exposure and helping people cut through a lot of the noise. If you haven't heard of MegaMinds, if you haven't seen Megamind.
Take a look at it because it's, it's very different and it's producing much more outcomes, academic outcomes for students. It's producing much more actionable data for teachers and for administrators, and I think those are salient points so that these organizations are helping us gain the exposure that we need and deserve.
We had somebody, a major consultant with a newsletter with 36,000 people the other day say, this is a tool that people have to find out. This is gonna benefit students and teachers so much, but people don't know about it yet. And so I think that it's gonna give us that exposure that we need to, at the end of the day, help students.
AI literacy bundle is free for all schools in the United States during this entire school year. That's something that we're trying to do to help the students of, um, of the United States be ready for the future, which is an uncertain future. And the things that we do know is that critical thinking, that creativity, that adaptability, those are not gonna go out of, out of style.
Those are always gonna be there. The ability to establish criteria for evaluative decision making, that's not gonna go outta style, reflection, metacognition. So we're trying to really hone in on those things that we know are still gonna be there in a market, a job market that in 30 years will be completely different than what it is now.
Five years, it'll be completely different than we did now.
[01:30:39] Alex Sarlin: That's right. What you're both saying really resonates and it's really interesting to hear is that you know, these initiatives can bring needed exposure. They can get your tools that are MegaMinds is really different than other tools, but it might not be something that people can immediately find it.
RA raises the visibility and brings it to a lot more schools at the same time. What you're saying, Eric, I think is really notable as well, that the initiatives by themselves doesn't always just mean all the doors are open. The ed tech companies, especially one with, you know, like MegaMinds that's vetted against privacy, vetted against the Irma process, can actually support the, the momentum of the initiatives is an, a synergy or whatever, uh, there where they go together and can really help raise momentum for both the movements and the tools within them, which I think is exactly the point that's really exciting.
There's been really interesting conversation. If you haven't checked out MegaMinds, you absolutely should. They're doing really, really interesting work in this space. Eric Tao is the founder and CEO of MegaMinds. He was a former innovation lead at Google. Austin Levinson is the director of learning for MegaMinds.
20 year educator has done gifted steam, project-based learning, design thinking and more. Thank you both so much for being here with us on EdTech Insiders.
[01:31:47] Austin Levinson: Thank you, Alex. It's, it's amazing to hear your insightful comments in the way you are bringing things together in ways that are very meaningful for us, for your audience.
So thank you for all that you do.
[01:31:57] Alex Sarlin: Thank you. That's so nice to hear. Thanks for listening to this episode of EdTech 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, EdTech Insider, subscribe to the Free EdTech Insiders Newsletter on substack.
