Edtech Insiders
Edtech Insiders
Week in Edtech 09/08/2024: OpenAI’s $100B Valuation, Google’s AI Innovations, Anthropic's Hive-Mind Vision on AI, PhysicsWallah’s Global Expansion, AI’s Role in Dyslexia Intervention, and More! Plus Special Guest, Kumar Garg of Renaissance Philanthropy
Join Alex Sarlin and Ben Kornell, as they explore the most critical developments in the world of education technology this week!
✨ Episode Highlights:
[00:03:25] 💰 OpenAI in Talks for $100B Valuation, 200M Weekly Active Users, Considers Higher-Priced Subscriptions
[00:15:18] 🖼 Google launches Gems and Imagen 3
[00:18:53] 🐝 Anthropic CEO predicts hive-mind AI, partners with Amazon for Alexa revamp; Claude for Enterprise: 500K context window, GitHub integration, and enterprise-grade security
[00:35:04] 🎓 PhysicsWallah’s ‘Alakh AI’ making education accessible in India, expands 4x in UAE
[00:38:50] 📚 AI could revolutionize dyslexia intervention and drive educational gains
[00:43:42] 🎓 Yale announces $150M for leadership in AI
[00:46:36] 💻 U.S. Higher Ed Faculty increasingly reliant on digital course materials
Plus special guest:
[00:52:34] 🎙️ Kumar Garg, President at Renaissance Philanthropy
😎 Stay updated with Edtech Insiders!
- Follow our Podcast on:
- Sign up for the Edtech Insiders newsletter.
- Follow Edtech Insiders on LinkedIn!
🎉 Presenting Sponsor:
This season of Edtech Insiders is once again brought to you by Tuck Advisors, the M&A firm for Education Entrepreneurs. Founded by serial entrepreneurs with over 25 years of experience founding, investing in, and selling companies, Tuck believes you deserve M&A advisors who work just as hard as you do.
Welcome to EdTech insiders, the top podcast covering the education technology industry from funding rounds to impact AI development across early childhood, K 12, higher ed and work, you'll find it all here at edtech Insider.
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 Ben and Alex, and we're back with another episode of week in edtech. So exciting to have you here. So many great guests and topics to talk about. Alex before we dive in, let's just start with what's going on with the pod. Yeah,
Alexander Sarlin:so so much is going on in the pod. We talked to our good friend jamira Herrera from reach capital, and Lyman miser, who's ex Newsela and now at owl ventures in collaboration with New York edtech week, that's going to be coming out very, very soon. Ben, you talked to Jen Holland at Google for Education. That sounded like a great conversation.
Ben Kornell:Yeah, a 10 year anniversary of Google Classroom, and she's been there every single moment along the way. Very cool, amazing.
Alexander Sarlin:And we talked to tal havivi from ISTE and ISTE and ASCD, that's doing all kinds of new things for teacher development and technology, and also our good friends, Ilya and Mel, two Columbia professors who started a really, really amazing edtech tool that uses Netflix and Disney TV content for emotional intelligence purposes. So lots of cool stuff coming up on the pod in the next few weeks. And then we have a great guest today, right? Ben, yeah,
Ben Kornell:we have Kumar bar, formerly of Schmidt futures. They have spun out to create Renaissance philanthropy, and it's just exciting to see about venture philanthropy, and how is that space evolving? And Kumar's been a long time advocate for the intersection of R D and education. So great combo today,
Alexander Sarlin:yeah, and that ISTE loves that stuff too. It's really a big part of their mission. So we have some events coming up. You know, this week is the whole on IQ back to school event in New York, which is a big investor event about education technology going really deep with lots of lots of interesting people. We will be covering that in some ways over the next few weeks, but if you are going there, you know, have a great time. It's a really, really interesting event. We also have this Thursday, so that's September 12, a session with Matt Walton of Future Learn, an online, free session about product market fit in edtech, one of the most elusive and tricky things there is. And then Ben, we have an in person event coming up in just a couple of weeks, right? Yeah,
Ben Kornell:join 100 of us to celebrate the back to school. It's both the celebrate and commiserate. September 18 in San Francisco at Salesforce Park. We've got a reserve space out there in the sunshine, and it'll be a great coming together time, and then also at New York edtech, which is October 9 and 10th. We will be in New York. We hope to see you there. We're doing our famous rooftop party with magic edtech, and we hope all of our listeners and members can get there. Just a reminder, if you want to come to our events, sign up for edtech insiders, plus that is the way to secure your spot for all of the events. You know, when we fill up, we can only release a couple of spots to the general public, so please go to our substack and look at the subscribe options. We'd love to have you at all of our events. Let's jump into what's going on in education, technology, AI, etc. I wanted to start first with a little bit about founder mode, which is becoming this hashtag moment in Silicon Valley. Just the backstory Paul Graham and Brian Chesky. Paul Graham, leader and founder of Y Combinator Brian Chesky, founder of Airbnb, kind of had this two hour, two and a half hour insider session for Y Combinator alumni, where they talked about founder mode. And the kind of essence of it is that founders have a unique of. Ability to cut to the core of issues, solve problems, you know, disrupt bureaucracy and drive things forward. And I think there's a frustration among founders, around a professional management class that comes in after founders, and their inability to drive the same kind of dynamic results that the founder did. It's a really interesting thought piece that has captured the zeitgeist of the current moment, in part because there's just frustration overall that many tech companies are stalling out, and they need to point the blame somewhere, and it's these like next gen managers who are getting a lot of that frustration. I think the other is that AI is coming along as a disruptive technology, and there's concern about tech companies at scale being nimble enough to actually respond and react and take advantage of the opportunity. And I think third is a little bit of the darker side. There's been, you know, a big political shift in Silicon Valley where there's a number of folks who have endorsed Trump, who have kind of come out with a little bit of an anti dei stance, with a concern that the policies, the progressive hiring practices and policies of tech have hurt their competitive capability, and so they point out founder mode, almost as an alternative way of talking about that issue, and saying we shouldn't have hired in all these managers. Founders need to shake things up. We need to get back to breaking stuff and innovating and so on. And so, you know, I have a like this is one of those where I have a balanced view in that there are times where founders need to get in, or even managers, who've been hired in need to go shake things up, because you have a business that's either struggling, or you have a nascent business that needs the startup skills. But this founder mode hashtag flies in the face of a lot of research, which also shows, and this is from like Stanford and Harvard, that often your founder skill set that is needed to get you from zero to one or zero to 10 on your product journey is not the skill set that you need to get to from 10 to 100 or 100 to a billion. And there's also been a lot of employee pushback, which is, you know, people saying this is just going to be a call for founders to get back and micromanage everything, and that is not an effective way, long term, to run a company. So it is, literally every day, a hot topic. Everybody's giving their take on founder mode, and it's rapidly becoming part of the canon of Silicon Valley ethos. I don't know, Alex, if it's penetrated the out of the bubble universe, but what are your thoughts on it?
Alexander Sarlin:I've been heads down for the last couple weeks, so not something that I've come across too much outside of the Bay Area, but that, you know, Paul Graham and Y Combinator. We know Sam Altman was also formerly Y Combinator, and Brian Chesky is one of the big hero founders, quote, unquote, in, you know, still sort of running the company. I think he still runs Airbnb, right that? Yeah,
Ben Kornell:let go of a large number of mid management people as part of his founder mode. You know,
Alexander Sarlin:I mean, I've been at a number of companies that have experienced that shift, that very, very clear Cultural Management shift, where the that goes from sort of the founders to the management class, and I agree with you in that there's reasons to do it, and there's definitely a sort of professional CEO skill set that includes a lot of financial maneuvering, especially if you're trying to IPO. It includes a lot of legal maneuvering for patents, and it's covering from competitors. Like there's a sort of, like mature company skill set that I get. Why the pattern for, I guess, a decade or two now has been, you know, the founders then step aside in favor of professional managers at the same time that cultural change is usually pretty depressing, right? I mean, let's put it that way, like, as somebody who's been inside these companies, the shift when you go from this incredibly idealistic Silicon Valley, like, you know, moonshot Masters of the Universe, all of the, like, ridiculous Silicon Valley stuff, but you feel it when you're in a company like that. You feel inspired. You feel like you're sort of close to the core of this comet. And then as soon as that sort of managerial class comes in, you feel like you work at a company. And, I mean, all the HR picks up, all of these things start to really, like, get in place, and you start to, you know, just all sorts of stuff changes. And I can see why, not only founders, but investors. By a whole lot of people say, hey, you know, we've seen this pattern for a long time, and it feels like, you know, the trade off is, are you bringing this managerial class? And they blow it through the roof, right? I mean, that you saw that with Google, for example. You've seen that with Apple. Those are. Maybe not even the best examples, actually, because those are both companies, especially apple that grew a lot under one of its founders, but you've seen it with a number of companies. Let's just say it's not that it has worked for a lot of companies. Yeah, at this point it's not working at that systematically anymore. It's not clear that you bring in those seasoned pros and it just works. And I get why there's a pendulum swing, yeah. I
Ben Kornell:mean, there's like, the Adam Newman examples, where it's like, where it goes wrong, and I feel like it's precisely what you said, a pendulum swing, where it was like, Oh man, these founders have been unfettered, and now look at all the crazy shit they did, and that was kind of in the era of free cash and capital where, you know, boards were stressed because they didn't have voting rights and so on. And you know, the founders had this superclass. And then we swing back now, investors are crowding out founders. And, you know, at the small and mid level company size, more and more investors are taking majority stakes in the companies where there isn't a founder place. So you can almost see this coming from a defensive standpoint of like, Hey, we've got to get the investors to realize founders have some unique and special talents and skills. And I think that your other point, that's totally spot on, is that there's not a one size fits all approach. If your company is thriving and it's really about building and sustaining the engine, the machine to, like, keep cranking it. Yeah, that's a great spot for, like, professional managers. If it's like, oh no, our market is disrupted and we need to pivot dramatically, that might be a good time to have somebody who has the founder skill set and can just break stuff and move things, and has the kind of unilateral authority to make stuff happen
Alexander Sarlin:exactly and test really rapidly, sort of put out products just to see what works quickly, rather than it being this massive, long water folly process where you have to have the marketing team weigh in and all that stuff. It's, I totally see why the tension is there. And I think, you know, it obviously relates to EdTech, because a lot of edtech companies are in these spaces where they have founders their venture backed, and they have to, you know, figure out whether the founder stays in. I mean, the companies, in my experience that I've been at that had this were edtech companies. And, I mean, edtech maybe doesn't have as big a problem with it, because there are just simply not that many edtech companies that sort of reach this stage of, okay, you're blowing it up. You're doing so well now we've got to bring in the pros. I just don't think that happens as often in edtech as consumer tech or, you know, the ride sharing companies all have that. But it still matters, I think, for edtech founders, because that dynamic and that sort of tension between the founder class and the investor class, however you want to, you know, frame, it is something that every edtech founder should be keeping their eye on, whether they're in Silicon Valley or not, because it really might affect some of the terms that come towards you in a term sheet, or it might affect the way you make decisions and try to Maybe present yourself as somebody who's getting that management skill set over time so you don't get pushed out when, if you know, you don't become a victim of success.
Ben Kornell:Yeah, maybe this is a signaling of a larger shift where in the High Times, there was a real alignment between investors and founders. And now that we're in a tough economic state, yeah, there's much more inherent tension. And so the founder mode folks are going to stand up and say, Hey, we need to back our founders and investors. You need to decide which side of the fence are you on. And it is really interesting to see in I've been seeing some investment rounds where you know the terms are essentially the same, and one investor group is saying we are more founder friendly than the other one go with us as the sales pitch. So it is a fascinating time before we go deep into our edtech universe, let's go to AI. Let's maybe around the world in AI. What are the headlines that are capturing your attention?
Alexander Sarlin:Yeah, so a couple of big headlines from open AI that I think are worth keeping an eye on. One is they announced that usage of chatgpt, sort of core product, has now doubled since last year, and they now have 200 million weekly active users. That's, you know, two thirds of the population of the US. It's five times the population of France, that kind of thing. So that's a pretty good number of weekly active users. And on the back of that kind of release, they're also talking about more funding at a humongous valuation of over $100 billion and they're also starting to differentiate some product pricing and considering maybe some higher priced subscriptions. You know, the chatgpt Plus, right now the consumer version is still, I think, set at like $20 we may see things that go much, much higher, maybe even orders of magnitude higher, like 1000s of dollars a month for incredibly sophisticated, very, very powerful, AI, so we're. Starting to see some of the dynamics change in the AI space as they move into enterprise. Those things all stood out to me. I've gone back and forth on OpenAI. I think they're doing amazing product, but they're also sort of the outsider, compared to the Amazon's, Microsoft's, Google's, but these are all pretty exciting headlines from that company, so definitely something to keep an eye on. What stood out
Ben Kornell:for you? Yeah, I mean, the other thing that has, you know, addition to OpenAI, the two things are Google and their announcement of gems, which are kind of like gpts, which are your ability to customize your own version of Gemini. I would say gpts Haven't been quite the success story that OpenAI had hoped. But I do think in an enterprise context, this idea of creating custom bots with my own data is really appealing. And the kind of custom GPT is a great place to start, and a custom gem, in this case, is a great place to start. What's noticeable is all of the use cases that they pre populated. The Gems with are education use cases, and one of them, in fact, is powered by the Learn LLM, which is basically trained not to give you the answer straight up, but actually help you experience that cognitive friction of stretching and learning, and, you know, it asks you questions back, and it's Socratic, and it's really interesting to see the investments that Google is clearly making in outcomes. The article that Google had in the Atlantic, which I'm quoted in, is really about metacognition, and what they're realizing is anyone can get you the answer, but helping you lay out your thinking process, that's really interesting, and that's going to be even more valuable. The second one that I'm paying attention to is anthropic. They were just rated kind of top in class in terms of the safety of their AI systems and tools by Common Sense Media and you know, safe for kids, safe for families, accurate, clear citations, etc. This week has been a big week for them, because they launched their enterprise version. And as part of that announcement, you have your chat window, which is where you can kind of put in your normal, kind of similar to chatgpt, your normal prompts, but it also has a second window which has output. And so you can actually create a website. And as you're telling the AI, here's what I wanted to be doing. On the right hand side, you can see that evolved. And I think what we've had is one shot AI where you ask it to do something, and it generates it. And then when you ask it to do something again, it's a fully new generation. What this is going to allow is multi shot generation. And you know, we covered this like six months ago. Anthropic was one of the first to allow teams to collaborate on one, you know, prompting window. So you could actually have four or five people all working together, typing in prompts, and then the collective outcome could be on the right. Where I get excited about this in edtech is think about that as a learning tool. Oh, you've got four or five people all prompting together with an instructor guiding it all in the same chat window. It's unlimited what you could really do in terms of co creation. And if you've got some really, like strong instructional designers, how amazing of a canvas is this going to be to paint learning? So I think, you know, dark horse in the race here continues to be anthropic, and they continue to have this drumbeat of great use cases, tools and features, while the models underneath are somewhat becoming commoditized. Anthropic is making like interfaces that are differentiated?
Alexander Sarlin:Yeah, not only are they making differentiated interfaces because they're focused on enterprise use case much more than the others, I think they're really thinking hard about products that are enterprise friendly. So safety and sensitive data is an Enterprise feature, not only but it certainly is. And that kind of two sided like you can build on the fly and actually make a website. For example, you're you're mentioning as a possible you're making a website on the right side, and the code is on the left, but it's not just the code like it's actually the AI bot that's writing the code, where you can talk to it in normal language. You can see how this becomes a incredibly powerful work tool as well as a learning tool. The other thing that stood out to me about about anthropic this week, you know, we see how some of these smaller companies have been sort of serving as proxies in some ways, or sort of like almost the faces of bigger companies behind them, we saw that with OpenAI and Microsoft. Amazon has put a huge amount of investment into anthropic, and they just announced that the new Alexa, which just launches in a few weeks from now, is based almost entirely on the Claude models, on anthropic LLM models, rather than Amazon. On in house models. So you can very easily imagine some of the conversations happening in Amazon right now. You're the one of the biggest companies the world, one of the most powerful companies in so many different ways, but AI has just come through the whole world like a wrecking ball, and you're like, do we have the internal capacity to build something at the level we want? Maybe not, or not yet. So we're going to work with these really smart people at anthropic so Claude is going to be inside Alexa, and that will allow things like ongoing conversations with Alexa. Speaking of your one shot, multi shot, instead of just asking and getting a question, you can have a deep conversation. It'll remember what you're actually saying. You know, to some extent, at least, it can give you advice. It can do all sorts of things that AI can clearly do. So that's really interesting from a product and sort of go to market standpoint, I have to stop and say, congratulations, Ben. It was so exciting to see you quoted in the Atlantic in collaboration with Google. That is huge for you and Fred Tech Insider. So that was amazing. If anybody hasn't seen that, go check it out. What is the name of the article? Oh, I
Ben Kornell:don't remember something about metacognition Google and Atlantic. You can find it, but I appreciate it. And by the way, I will just say in our WhatsApp group, we're getting a ton of great conversation about that article and several other articles. So if you get a chance to sign up for edtech insiders, plus, you can check out that WhatsApp channel, but you know, a lot of the ideas that I even talked about in that Google article come from the dialog we're having within our community, which is basically like a standardized test, is rapidly losing value, and the metacognition that goes into how kids think is rapidly ascending in value. How do we capture it? And there are startups that are thinking about it this way. Snorkel comes to mind as an example, where it's all about showing your work in math cognition. At my meeting at Google, Damir, who's the founder of Photomath, all of their focus now is not on just giving the kid the right answer, but it's around helping them lay out their thought process. Super interesting moment,
Alexander Sarlin:no question about it. And there was a terrific article this week from Ulrich Boser, who is the author of a great book on learning. He runs the learning agency. His take is that some of the way that AI is really going to drive education is through what they call, quote, hidden tools. What he means by that is it can actually be about improving. In this case, the way tutors work with students, and the way that that's going to be improved is by metacognition. It's basically about helping tutors understand that they should, draw out the reasoning, draw out the actual thinking behind the thinking, and get students to be metacognitive about their own abilities. That's what great tutors do, and that's what AI can train tutors to do. So there's a sort of like, he's pulling it back and saying, you know, whether or not you're worried about kids interacting directly with AI, we're already seeing meaningful impact from Ai teaching teachers and tutors how to think more aligned with evidence. And that's really, it's an interesting take, and I think it relates to your metacognition point. Just one more quick thought about this gems thing. You know, I agree that GPT is obviously when open AI, put out GPT creation ability, and then has adapted in various ways. The pitch there was, we want to be the App Store, right? We want to be the App Store for gpts for AI. We want people to be able to create and sell agents, AI agents that do all sorts of super specific things. And that's a great pitch. I mean, of course you can see why that would happen, or why people would want to run with that, but it hasn't quite happened to your point, right? I mean, I think we're just too early in the AI world for people to not only be using chatgpt, 200 million, right, weekly users. That's pretty good, but also be going through it and saying, I don't even want chatgpt. I want a super specific agent model that does this exact thing. It's all about, you know, car mechanics, and it knows absolutely everything about that, or knows everything about Roman history or anything like that. I don't think we're there yet, but we did see Dario emoti from anthropic this week talk about how he believes the future of AI might be this sort of almost like organizational structure model, almost like a corporation, where the top level is the generalized AI, like Claude or GPT four, and then there are this whole set of agents underneath them. So when I hear about this gems idea, I think, Hmm, I wonder if you know, you could ask Gemini, and if you're asking Gemini a question that is education related, it passes it to learn LM or to an agent that's using learn LM like a gem we talked a little bit about this with Claire Zhao when she was on but you can start to see this sort of model emerging of how the AIS might all work together to do really amazing things. You have AIS for image, you have AIS for video of AIS for education. You have AIS for advice. You have AIS for writing, and you know, a triage AI, a top level AI can say, I understand what you're trying to do here. You're trying to get help with your homework. I don't want to give you the answer, because that's not going to help you. I'm going to give you a really good tutoring session where you get to be metacognitive and. Mistakes and do productive failure and all these great things. It's pretty exciting for education. And I think that combination of the generalized models and these specialized models is really powerful. I'm curious. Just Ben, I don't know if this has come up yet, but you know, you see on one side of the AI landscape places like Synthesia and hey Gen that are making these incredibly realistic characters and people, or character AI, which is just sort of purchased, basically, and then you have these ideas of gpts and gems, which sound like apps, they sound like software, right? Are these going to come together? I mean, are we going to get to a world where it's like, it's not just these sort of abstract models and agents, but it's actually like, there's a character, and that character is, you know, a great teacher. That character is a great car mechanic. And, you know, when you think, hey, I need a great car mechanic, I'm gonna go ask, click and clack, whatever it is like, are we heading there? I can never tell. It feels like the pendulum swings back and forth and it's happening in parallel. Yeah,
Ben Kornell:there's no doubt that we're headed to some sort of agentic AI future, right where you know whatever your prompt is gets cascaded through other agents. And today, the user challenge is that you have to direct that flow and that kind of org chart. And of course, like a real business opportunity is for someone to either standardize that or allow you to make selections on your agentic flow. And that, by the way, I think the challenge with any agentic flow is that the best image generator changes over. Do you the user want to pick which generator you want to use? Or do you want someone who's basically like the best in class right now, or this is the cheapest, et cetera, et cetera. So startup founders were already seeing building these Atene workflows, eventually, one individual would have a personalized flow. I actually think this is less an AI question about how it will play out, and more of a human question, yeah, you want your one character to be your therapist, your learning coach, your, you know, website developer? I don't think so. I actually think that you turn to different profiles in your life for different needs, and so the kind of billion, multi trillion dollar question is like, can somebody build a one stop shop where you actually have your collection of four or five different personas that you rely on that then can create that have a cascading agentic flow that you can relate to, and they would own the whole ecosystem? Yeah, what we're seeing right now is that the way it's playing out is a more fractured experience, where I go for my companion, AI in one set of app ecosystem, I go for my learning, AI in another one, and I go for my work based AI in another and you may not you as an individual may not be in control of what AI system your school or university or your workplace uses, too. So I think that is a big question. The people who could be that universal source are people like Google and, by the way, Microsoft also, which is affiliated with open AI, Microsoft's doing quite a bit to bring their AI copilot into everything. Remember Clippy? Well, think about Clippy on steroids everywhere you want to be. I personally can't really deal with the Microsoft ecosystem and interface. It's just like, too hard for me. I'm like, I'm imprinted in Google, but I think I could see people picking a ecosystem or universe that they're comfortable in and actually getting a bunch of agentic flow. Our listeners, I'm sure, are listening with some skepticism, and I think it's well founded that right now to do a prompt that then goes out to different agents and then comes back with something is expensive and takes too long, and so what we're postulating about is a world in which this is instantaneous and the cost is approaching zero, because that's where the trend lines are going. But that's another breakthrough that will have to have to happen, and you might find that this, like agentic coordination is a big advantage in terms of cost and time that somebody might figure out we have the folks from grog on the pod. Part of their bet is that actually, speed is more important than the overall, like, compute capacity of your request, and they're redesigning the queries in ways that make it cheaper and faster. So I think that's where we're heading. Then you know, we in learning, have our fundamental problem, which is the user who wants to learn, they generally want less friction than the teacher or the parent who wants more friction. So if you. Want user satisfaction. You know, chatgpt has now 200 million weekly active users. We thought 100 million was a lot, and then, you know, we thought, oh, it's kind of dying down. No, it's just basically people are finding their use cases and using it every day. That's right, though, what we've got to understand is that actually having a different agentic ecosystem for learning is probably better for the learning outcomes, even though it might near term, get, you know, less than five stars from an individual
Alexander Sarlin:user. Oh, fantastic points. I think that's that's such an interesting future. I love the term agentic flow. I'm gonna, like, totally adapt that one. I think that's such an interesting way to look at how a query, or, you know, how a prompt could be directed to the right agent, whether that is a character or a bot, and whether that's within one ecosystem or across. I mean, when you hear about Amazon putting Claude and Alexa and one of the things it can do is shopping advice, you're like, well, that it seems like some pretty good vertical integration. You say, hey, Alexa, you know what's the capital of Thailand? I'm thinking about going on a trip. And then it goes, Oh, I can give you 50 things on Amazon you should buy for that. Like, let's
Ben Kornell:be real. Like, we're seeing a proxy war with the large platforms. Anthropic is sneaky, not in a bad way, but they are coming up, and Amazon is behind them. And imagine if you're building a website, all Amazon has to say is, post this on AWS, like, let us set up your commerce website, and then, boom, you're like, in the Amazon ecosystem. That's powerful, hugely,
Alexander Sarlin:hugely. I mean, makes me think just of some of the other times in the past when these giant companies have actually bumped up against each other. And this is probably a silly example, but I think it may be interesting. Like, remember when PowerPoint was everything, basically the only tool that anybody used for presentation was PowerPoint. Even if you were a Mac user, even if you were in the Apple ecosystem, they didn't have anything they compared with PowerPoint in terms of ubiquity and ease of use. So it was one of these one moments where you'd sort of cross over, even if you're an apple person, you'd use PowerPoint. And then eventually Apple came out with keynote, which was pretty much a complete dud, as far as I think about it, I don't think that many people use it, even though it's beautifully designed, has all sorts of amazing, beautiful, yeah, and then Google came out with Google Slides, which crossed both ecosystems, totally web based, and sort of ate both of their lunch in some ways. And I think the same thing is going to happen here. You're going to see, like, to your point about agents, you're going to see some ecosystems be better at some things than others. It seems like Google's trying to basically lock up education, or at least move really, really quickly on serious educational evidence based, you know, responses. So if Google becomes the education place, people who are in OpenAI and doing it for education may not get quite as much as you get from or learn LM, but if a lot of people are doing it, it puts pressure on OpenAI to improve that part of their ecosystem. And I think if they have these parallel ecosystems with different use cases, obviously some are more important for business than others, like, you know, cloud hosting or shopping, but you have these different ecosystems, they're all going to continue to have to compete with one another to try to have enough of a suite that you don't cross over and don't say, you know, I wish I could ask my Alexa about this, but I know they're going to try to sell me something. So let me go over to chatgpt and ask it like it's going to be a really interesting proxy sort of war between the devices. We heard that mid journey is thinking about doing hardware. I don't even know what that means, but mid journey is thinking about doing hardware like these crazy things are happening, and it's just going to be consistently interesting to see how these companies and these models sort of compete with each other in different areas. We also saw just a quick note of in edtech. We saw physicswala, which is a really interesting Indian edtech company, put out a couple of announcements this week that I think are worth mentioning in this context. Right? One is that they are growing very quickly in the Emirates and in the Arab world, which is really, really interesting, they're really trying to sort of go global. And physicswallet is not a brand that is super known global yet, but it's very big in India. But they also announced a whole suite of AI tools, including a 24/7 companion, student companion that's really basically like, to your point, trying to be the one stop shop. They call it the AI guru. And it's like you can ask it about anything. You can ask it about your homework, about something academic or something non academic, and it's just there all the time, available to everybody in the physics wallet ecosystem, including millions of users in India. So you can see that, you know, there's a outside chance that an edtech company might actually compete or even supersede some of the big tech companies when it comes to very specific learning outcomes or learning learning measures. I'm not sure that will really last, but it's interesting to see. You know, pretty big companies like physics Wallace, sort of head that way, and they're not they're not saying, hey, just go to GPT. You want to ask something about about this. They're trying to own the ecosystem themselves.
Ben Kornell:Yeah, I think one, gosh, man, there's three or four different great points to pull out from there. One is that our assumption that the leaders in AI and education are going to come through the US market, I think that that might be a flawed assumption, and we've got a little bit of the Clayton Christensen's innovators dilemma here in that, you know, for the established players in the US, you don't want to disrupt your market. For schools, it's like slow you're competing against real teachers, real classrooms. And then you go to developing countries, or countries like in the MENA region, where there's huge oil wealth, but there's also an underdeveloped education system, and the AI enabled options really represent a leap forward. Second is that the door to the classroom is actually likely a child's first experience with AI. And you know, there have been many business strategies won or lost around customer acquisition in K 12 that is essentially a loss leader for lifetime alignment to an AI ecosystem. And so when you talk about crossover around Google as the like learning system, or with physicswala being kind of a tip of the spear, potentially for some acquirer who wants to be in those systems, I think that's a really interesting space to watch. Even if it's not moving billions of dollars, it might be embedding people in an ecosystem. And then last, you're just rewinding to your point about PowerPoint. I think it's such a great analogy, because what we're seeing play out is a business strategy question, feature by feature, there's no doubt keynote, if you lined up the features, any product manager would say keynote is a better product. It should eat Microsoft for lunch, and then you add Google Slides, especially gen one, you would say it loses on all of those things. But what I think you found is that actually PowerPoint was a threshold product, where you had to meet a certain bar of functionality. And then there were other factors that unlocked it, and Google immediately built it, cloud based and immediately collaborative with others, whereas PowerPoint was locked in a license and not collaborative at the time, and keynote was locked into the Apple ecosystem, so very hard to share and so on one feature, dynamic Google Slides, beat the other two, which was around ability to collaborate, that created a product virality, where to work with somebody else that invited you to join, to their their Google side. Now you get exposed to that new product. This is what is going to win. Ai. Things like this. Who has the best AI is becoming irrelevant. Who has the AI right there when you need it, at the time you need it and when you need to collaborate with others, that's really interesting. And as we said, anthropic is actually in the lead on collaborative AI. So is it going to play out that same way, like you and I are certainly not the smartest people in the room on figuring that piece out. But I think what anyone from the outside needs to realize is, don't just do a feature by feature comparison. You need to figure out what are the threshold qualities that that you've got to meet to be in market competitive, and then what are the viral growth qualities you need to have to drive? You know, user acquisition and market share capture,
Alexander Sarlin:such fantastic points. It makes me think about how, you know, when I first played with Gemini and I realized that you could just export an answer from Gemini directly into a Google Doc, I was like, well, that reminds me of the moment when first played with Google Slides and you realize you could export it as a PDF or a PowerPoint or anything like there's these subtle dynamics. I mean, they're not subtle when you're actually experiencing them. But if you're looking at our set of product features, these subtle dynamics that really create lock in or the opposite, that create the ability to switch. There's, like so many really interesting dynamics that happen in this, and I definitely think this is going to happen with AI. What's really interesting about it? So, you know, some of the other things that were have been talked about this week. There was really interesting article about AI for dyslexia intervention, and there was a number of companies starting to figure look at special ed, look at dyslexia through an AI lens, which I think is really smart. So you play that out parallel to what we're talking about here, let's say one edtech company really does create an incredible dyslexia intervention tool that uses AI. It can maybe diagnose dyslexia, it can change text, it can do all sorts of things to support students. Well, suddenly that company could. Stay on its own and try to work sell to schools, but it might be a whole lot better idea for them to be acquired by Google's learning suite and be part of the Google system, because the Google system's already at every school, and you already have students, like millions of students using it every day, it would reduce every kind of barrier to entry and every kind of business issue. And for Google, which is, you know, ostensibly, it's a big company now, but ostensibly trying to, you know, do good for the world, that is such a powerful thing that they probably couldn't do on their own, frankly, right? At least not immediately. It's certainly not a priority, right? That is dyslexia intervention. So I really wonder how the sort of edtech and big tech AI ecosystem is going to play out, and I'm hoping, I really am hoping, that as anthropic and OpenAI and Google and Microsoft and Amazon and all these folks at Gemini are all you know against each other and trying to build that they might actually be looking for really powerful use cases, like you look at like illicit or some of the really good research tools. Or, you know, there's a potential. I don't want to get over excited about this, but there's a potential for these big tech companies to start pulling in specialized functionality as a way to differentiate itself from other ecosystems.
Ben Kornell:I'm of an opinion that we will eradicate dyslexia over the next decade, and when you know dyslexia is not a disease, dyslexia is cognitive difference. And I think there's a bunch of information that goes to show that neurodiversity is actually a real positive as we get into an AI future, but the barriers that dyslexia represents towards literacy and towards academic achievement should be totally gone in 10 years, in part because the science around dyslexia is so compelling that if you are able to identify dyslexia in a three or four year old by the time they're in second or third grade, you would never know that they have dyslexia in the first place. Because of neuroplasticity, the therapies and strategies for early learners are profound. The challenges our diagnostic system relied on kids failing to learn to read in kindergarten and first grade, so it wasn't till second or third grade, that they're actually getting those therapies or those treatments, yeah, and, you know, for our generation, many folks just went undiagnosed altogether. So companies that are doing this, I think, have a huge opportunity. One in the article that they talk about is called dissolve, dissolve. AI dis like dyslexia, but with solve, I'm also an investor in early bird, and there's also marker learning. There's just a number of companies that are using AI to move the detection window up, because what they're able to do is like, understand the cognitive difference without the decoding components of reading. And so, for example, with early verb, you're playing a game that actually tests for the cognitive skills that kids would have or wouldn't have with dyslexia. And then is screening therapies and interventions. So I think there's other parts of education that are going to be slow to change, hard to change. This is one where all resource energy should just go into making this happen. And I'm going to be sorely disappointed if we're sitting here in 2034 and dyslexia is a like a recurring issue, because it's totally within our hands to solve globally, like even the cost of delivery of this thing is going to go down like an insane amount, yeah, yeah. It's
Alexander Sarlin:a fantastic point, and I think it's a great example of a type of educational problem that we've struggled with for a long time, mostly because of capacity. Right? There's just not enough capacity to test every kindergartner for dyslexia with the current system without having smart tools to do it. So, you know, other capacity constrained problems I think are really, are really on the table to be solvable through AI. It's very, very exciting. There were a couple of quick higher ed things I wanted to quickly mention. We don't have to talk about these very long but we've talked to Georgia Tech. We've talked to a few different, you know, universities that are really trying to figure out their AI strategy and doing some interesting things. This week, we saw Yale a little school in Connecticut. I think Yale announced $150 million to support leadership in AI. You know, it's a decent, decent investment, even for a Yale that has a lot of money behind it, it's for our targeted faculty hires. It's aI curriculum development. It's basically an investment that's supposed to go across the whole university to sort of up level it in the AI space, including infrastructure, which is really key. When we talked to Georgia Tech, that was a big thing for them. They're saying, Yeah, we you need computers. You need really strong supercomputers to be able to actually use AI at the level we should be using it at to train for the. Future so interesting to watch that obviously, you know, Yale is always a school that is trying to pave the way for others. We also saw something where results of a survey this week that said that faculty, higher ed faculty, basically the number of print textbooks that are required for students had just continually gone down that only print. And at this point, less than 10% of textbooks are only print. According to this new survey, two thirds of them are either print or in digital format, and almost a third 28% are digital only as of this year. That's really interesting. So 92% of textbooks are available in either digital or are in digital, at least in some way in digital and that's lower than it's ever been, by far. I mean, that's less than half of it was even two years ago. So I think we're starting to see some real comfort level with a digital material that we haven't seen in a long time. We also that same survey also said that more than half of higher ed faculty are now aware of open educational resources. That's a lot more than we used to see, and that 40% of them use OER as either required or supplemental material in their courses. This is all really good news for the EdTech world. I think it depends on how you look at it, OER NBC, is competitive with paid curriculum, but I think this is really good news that people are starting to really embrace digital curriculum and content and start to realize that they can actually offer lower cost options, in many cases, to students. So just wanted to call that out. This is a survey from Bayview analytics, and we'll put the link, of course, in the show notes. If you have any comments about either of those things, please jump in. But is there, I know you were thinking a lot about phones in schools. Is that something?
Ben Kornell:Yeah. I mean, I think there's a national dialog happening with our back to school right now, around phones in schools, and there's a range of solutions proposed from outright fans, don't bring the phone to school to these pouches that you can store your phone in to software. You know, it's like the distracted driver software, but asking kids and families to install that on their phone so that you know, only the emergency apps or learning apps are available. I think the step back question, and, you know, this is probably a different answer for elementary, middle and high school, by the way, I think for high school, just focusing on kind of the meat of the challenge, what educators are seeing and saying is that kids are constantly on their phones. It's really, really hard to teach when kids are distracted and the phone is out on the desk, under the table, in the backpack, whatever it is, there's just a kind of interactive, like compelling call. And as somebody who myself like has to be thoughtful about my own cell phone usage. I think it's a really important dynamic to acknowledge that high school kids and educators have to deal with. My concern is that this kind of move to outright ban doesn't actually build the executive functioning skills that kids are going to need in college or in career to find the right balance a binary band actually isn't teaching them life skills, which is what high school is supposed to be about. And I think high school educators may be told that it's about getting an A on your, you know, math or history exam. It is about that. It is about learning the knowledge, but it's a lot of it is the skills and competencies to thrive in your life, and we're seeing already this understanding that social emotional learning should be a part of our schools. This is an extension of that. It's a life skill to understand how to manage one's own cell phone usage time. I think the other challenge that we see is that many educators aren't aware of the technological solutions that do exist for this, and so it seems like the only option is a band or a pouch or something ultimately like what's going to happen is what is possible for schools to manage. The software can be as great as it can be, but if it's not easy to implement, and not easy on a day to day basis, to make sure those who have their phones are using the, you know, distracted driving software, and those who aren't, you know, aren't bringing their phones out. So I think there's a lot of logistical concerns here, but in our WhatsApp group, man, are people talking about it? New York Times is publishing about it. It's becoming a real mainstream conversation. You know, the software providers that are out there are seeing a big opportunity, whether that's a geo located, you know, ability to restrict application, basically block social media, or whether it's, you know, parents thinking about buying a phone that actually has some of these protections built in. I think BARC is the most well known, you know, safe smartphone. As you get down to middle school and elementary school, these issues become bigger and bigger. But the last thing I'll say, you know, we have really terrible news about a shooting in Georgia schools. There's, as we're talking. Talking about cell phone usage in schools and the distraction factor, there is an overlap with safety in schools that, you know, I think we're we've reached emergency levels in the US where parents don't feel like their kids are inherently safe at schools, and the cell phone represents an ability to connect with the outside world that parents are concerned about giving up. That's, you know, kind of the wraparound of what's going on on that and rad tech providers. There's, you know, real opportunities to help solve that problem.
Alexander Sarlin:Yeah, that's a very nuanced take, and I appreciate it. I'm sure our listeners do as well. There's, it's such a complicated issue. I mean, for what it's worth, if I were an educator, I would probably want a pouch. I would probably want to ban phones, just because I can only imagine how it's they're so addictive, they're so distracting, and you cannot just by definition, you can't tell if somebody is on Instagram or Tiktok, or if they're actually on a learning app, or if they're on doing something you know, relevant. I would want that as an edtech professional, I definitely do not want to ban phones. I think that they're not only useful for for safety, but I think I want a world where a student can have an AI, you know, note taker, on in their classroom in eighth grade, and they can be taking their own notes, but also get AI notes downloaded at the end with images in them. Like I want students to have full access to the incredible power of technology in the context of learning, but I totally understand why administrators and educators don't, and why, increasingly, states are now shutting it down. I think your point about, you know, distracted driver, or whatever, the limited usage, whether it's through it, whether it's through devices, whether it's through, you know, any other solution, it makes sense, but, yeah, it's such a tricky issue. My own kids, you know, I don't want them to have phones until they're 16. That's the Jonathan Haidt suggestion. I think that makes a lot of sense, if you can help it. But at the same time, you know, I want them to know how to use AI when they're 13. I want them to be creating, you know, creating stories and creating movies. So this stuff is just, it cuts both ways. And I don't know, I don't have much smart to say on that, intelligent to say on it other than that. But I think you had, you made a lot of good points. There's one more story that I think we should probably cover next week, just for time. But states received a big investment, almost $200 million from the DOE, basically, around literacy and assessment. And there's, like, a whole bunch of different there's 30 million for 10 states to deliver high quality assessments. This is really interesting, and I think we should cover it, but I don't think we we can do it justice here. Maybe we should move to our guest.
Ben Kornell:That sounds great. Hi everyone. I am so excited to have Kumar Garg, who is co founder of Renaissance philanthropies joining us today. Kumar, Welcome to EdTech insiders.
Kumar Garg:Thanks for having me.
Ben Kornell:So we're having you and a bunch of the tools competition winners on the pod. Can you just tell us a little bit about the origin story of the tools competition?
Kumar Garg:Sure, yeah, started actually, during the pandemic. So if you go back to that time, early 20, 20 billion kids suddenly out of school, the whole economy's shut down. One of the questions that I was really interested in, and we started to ask inside Schmidt futures, where I was working, working for Eric Schmidt, was, hey, we've been working on this idea of education and technology. For a long time, the role of innovation everything else. But look, suddenly we're like a billion kids, suddenly at home, some of them now using various forms of online education, and it's all really bad. Like, you know, people are super frustrated. Supposedly, we've been waiting for this moment of a huge tech adoption moment, and the results are mixed, to say the least. And one of the things that sort of jumped out during that experience, where everyone was jumping on and trying different strategies, was that even though the EdTech community had been working on this moment for 15 years and been trying various variety strategies, a big question was, is like, Well, what do we know that works in these settings? How do we actually build on the evidence? And then people said, well, actually, the evidence is really mixed. We don't know what works. And so it was born in that moment of crisis. We said we had a couple of different instincts. One, we said, Okay, well, we don't know how long. If you go back to that March, April, time period when we first had this conversation internally. Well, we don't know how long kids are going to be out of school, so we, you know, if you the future was very cloudy. So we said, look, we actually need to, like, start over and actually have a conversation around what are the tools and technologies they're going to help us get through this crisis. But then we're not going to try to make the same mistake again, which is to say, we're just going to build the point solution that will solve everything, because there's always, like, a little bit of the silver bullet phenomenon that always exists. We're going to say, how do we actually build tools and technologies where they're actually designed for continuous improvement? And so we designed the competition to be very. 30. So for anybody who's applied to the duals competition, they will experience this, which is, it's not a standard ed tech competition where we say, wow us with how cool your solution is and how many kids it's going to serve and how big of a deal it's going to be. Obviously, we ask what the solution is and who's going to benefit, but the key question that really makes it different is to be eligible for the tools competition, you've got to be working with a researcher. So right from the start, to be eligible, you have to have an orientation towards an active relationship with the research community, rather than, Oh, you call up a researcher as you start to grow and you're like, Oh, can you do a trial on me so that I can have it from my website. So a research partnership is built in. The second is we look under the hood and we say, are you building a tech stack that is designed for continuous improvement? Can you run an AB experiment? Can you actually test different learning strategies and see which ones work over time? And so those two ideas a research partnership and the idea of a tech stack designed for continuous improvement were the two underlying below the hood, things that we said, and then we said, Okay, well, what are all the different things you're going to solve for, whether it's supporting teachers, supporting learners in a variety of settings, whether it's tutoring, whether it's accelerated learning and everything else. So that was the original idea. It was money from Eric and CO funding from Ken Griffin and his foundation. And then we first did it just as a first year experiment, you know, and partly because it was in the middle of covid, we ran it global, because kids were out of school everywhere, and we ran it virtually. So we said, you know, we're not going to have like, a big in person pitch session, we're just gonna have all of the steps of it, the reviews and everything else, happen virtually. The pitch sessions will happen to the judges virtually, and selection will happen virtually. And what was interesting was, like some mix of that combination really took off. So the fact that it was global, and you didn't have to fly to the US for some pitch day actually got us a range of performers from outside the United States that had never applied before. The fact that it was kind of oriented a little bit more towards research and interaction with the research community and a tech stack meant that a lot of these edtech developers said, huh, I applied to this competition, but it seems like to get to phase two, I need a research partner. Is there a researcher you could connect us with? And we then actually had a research community that we had been already funding, and so we that led to a range of conversations. So one of the most common emails I get is, Hey, I didn't win the tools competition, but we're going to keep working with the researcher we connect got connected to, and you know that relationship is going to keep growing. So that was the context. And then over the years, instead of that just being a one year experiment, we felt pretty good about one year. So we said, All right, let's try another year. There's been a growing set of funders who have now started to invest into the competition, whether it's the Walton Family Foundation, whether it's you know, public funders like DARPA, whether it's you know, tech companies like open AI and others. So it's iterated every year, and it's expanded now to cover both higher ed and adult training. It's gone earlier to pre K, so it's expanded in its scope. But that core idea that we shouldn't just be thinking about point solutions, but be thinking about developers who are building tools that can continue to both advance on learning science, but advance the field, but then also continue to improve. Has been like a sort of core tenet of the competition, yeah, and
Ben Kornell:part of the benefit once you are part of the the winning group of the tools competition, you're part of this global community of entrepreneurs. There's a community aspect here. You talked about connecting researchers and entrepreneurs, but also the entrepreneurs connecting with each other. How have you seen that community element benefit the group of winners and benefit the field? Yeah.
Kumar Garg:I mean, one of the things that I find fascinating is that for folks who've been working in educational technology for a long time, they sort of think of this as a pretty mature field. They're like, well, since 1993 we've been at this for what feels like 30 years. But you have to remember that technology is a constant wave in their new entrance all the time, and so often what we find is the personas of people who apply to the tools competition are really varied. There's the person who had a background as a computer science student who ended up becoming a teacher who then built a tool as a side project that now says, Hey, maybe I should try to expand this tool. Well, their skill set, they might be very technically strong. They have something to education, how to build a scalable organization. They have never done that before. You might have folks who are deeply researchers. And so one thing we found on the community side is one person's superpower is another person's weakness, and one person's like thing they need is another person's skill set. And so one is like, just this idea that you can build these Avenger teams of amazing researcher paired with a technologist paired with a provider that's already in a. Bunch of schools. That's, I think, one. The second is that, at least outside the United States, because a lot of people end up biasing towards the United States, you can actually build these mega partnerships with the national government and suddenly be like serving half a million students and people sometimes underestimate how much scale some of these early teams are having, especially when you take sort of a global perspective. So I think there's a little bit of that. And then I also think we just lucked into a couple of, like, complimentary things along the way. We created this learning engineering listserv, which was just meant to be like a bit of a discussion group, but then has ended up being a live community, both of tools, competition competitors, but then, just like, there's tons of funders on there. There's a bunch of, you know, CS students when they're still in university and in graduate programs. And we found that there's a lot of like, teaming that happens because it's a community that's people are showing up for lots of different reasons, but then finding teammates through it and having that just be a live community that isn't just a community that's live for two months just before the competition or just after has been like, I don't think we necessarily designed it that way, but then has grown to be this, like, pretty active side community that can be unruly and fun interesting, but also like just people find a lot of teammates through it. I constantly am bumping into researchers who are like, Oh, I found my next job through this, and I found my next research partnership through it. So sort of underestimate, like, how much role you can play. And one of the things I'll just say is, as the AI wave has come, I actually think there's been more fragmentation. I'm constantly having people say, Well, I put my code up on GitHub, I put my paper on here, and where is all the work happening is happening in a bunch of different spots. Like, where's the data set? Where's the Kaggle competition, where's the research paper, where's the thing on Product Hunt. And, like, we haven't yet figured out what is the commons, where I can kind of see all the work the community is doing. So these, sort of, like, at least these forums, end up becoming even more important, which certainly, I'm sure, something you interact with all the time, which is that the watering hole is very important in a moment of a tip shift totally.
Ben Kornell:And it also speaks to the fact that we're in the nascent phase of our industry. We are not in a mature industry where we've seen like an alignment and consolidation and an orientation around the few things that work, or the few business models that scale, or the few distribution channels that are optimized. So it is a really, I think, the community aspect. People often focus on, what is the cash prize, and do I win? Or do I not? What really, you know, I see, especially looking backwards, is that you're it's step one in a journey for many of these entrepreneurs to be part of this community and to be building iteratively tools year after year after year that do mature. And some of the I'm a lurker on your listserv. I don't post very much, but I read a lot on it, and I do feel like there's a lot US centric ed tech companies can learn from leaps forward in developing countries. And just to see the kind of scale and utilization and also simplicity of some of the products is really compelling. So let's talk a little bit about your organizational evolution. So, Renaissance, you announced this early this year, but I'm sure there were some pieces in the works for a while. Can you just tell us a little bit about the decision to spin out of Schmidt futures and tell us a little bit about the focus areas for Renaissance? Yeah,
Kumar Garg:so it's definitely has all the like fresh pain of a startup. You know, we are at the four month mark. We announced in early May. The theory behind it was kind of a instinct I've had now in my career arc. You know, I first worked for President Obama, for his science advisor. So in this Office of Science and Tech policy, worked a lot with philanthropy back then, and then spent six years building a foundation for Eric Schmidt. And one of the things that I was sort of interested in is just where is science and tech philanthropy going? And one of the dynamics I sort of noticed in scientific philanthropy is there's a missing middle problem. I call it like the billion or the million. There's a lot of gifting, especially recently that's been happening at the billion dollar mark, a major school, a major university endowment effort, a big capital campaign, or there's like lots of great grant making that happens to individual researchers you know, that are advancing on a particular cause, but work in the middle often struggles in science and tech, which is a particular effort to say, okay with 2030, $50 million we can actually have a transformative impact on a sub area of science and take it to the next level. We can accelerate the pace of math research that's using AI by building a set of tools and technologies that can really up level that area. Same thing in education, where we can actually build a hybrid field between AI and. Learning science and change the way learning science research is done, those sort of mid scale theses struggle to get put together because they have to be super fresh. They actually have to understand where technology is super is right now. They have to be driven by a field leader, someone who actually knows what that technology weight is. And they need committed capital, right? They need like somebody to actually say, here's the resources that I need to deploy over a 357, year period. So the idea behind Renaissance was, hey, there's a huge number of donors that we interact with who are just early in their journey. They're interested in science and tech, but they don't know if they want to build a big Foundation. They don't know what for those people, they need like a new vehicle by which they can actually deploy that money in ways that are strategic. And one of the things I was just sort of interested in is taking a bunch of that previous experience and saying, can we build these focused funds where they're targeting some area in science and tech for broader public impact? They say, here's what I'm going to get to in three, five or seven years. And then they say, here's what I here's the field leader who's going to lead it, and then here's what I need on the philanthropic resourcing to get it done. And so that'll be the kind of Renaissance function. You know, I sort of compared a little bit to like what substack did for the idea of being an independent writer, where it's not that people didn't want independent writers before, but it sort of said, here's how you actually be an independent writer that can drive a whole strategy. We want to make it possible for field leaders in science and tech to be able to build these focused funds, philanthropic funds, and drive a focused strategy and up level an area of science or technology in a focused period of time, using philanthropic dollars and for us to be the umbrella entity that makes that possible. So still, you know, like I said early in that phase, but what's been exciting is that it's really unlocked a lot of ambition. You know, we've got folks who are leaving government or venture capital, and others who are like, I think I could build a philanthropic fund on a science and tech topic that's really compelling. Here's my idea. We've had donors say, you know, I don't know if I want to build a big foundation, but we definitely have interest in this area, what could be a fund in that area. And similarly, we've had interest just from, you know, sort of like folks who already have established philanthropies, who are like a fun model, just more lightweight and allows them to just deploy the capital, rather than having to do a multi year internal process, so that, you know, so we're at that early period, but like what you'll see coming out from the organization over the next six months are these funds where they'll have a focus, they'll have a person leading them, they'll have money, and then each of those funds will be out there deploying money against the particular sites and tech goal.
Ben Kornell:Yeah, it's so interesting how parallel it is to some of the challenges in in investing. Like, you have this chasm between seed and a and like, you know, D and E rounds, and it's that middle that is really, really hard to fund when you're trying to scale in for profit circles. So there's a parallel and then there's a collective action problem, which is, you know, billion dollar single actors, high wealth individuals, sophisticated and then, you know, single actors at the lower level scale smaller wealth and more bespoke. And one of the things you've done really well is bring the funder community. Even when you were at Schmidt futures, you were always collaborative with the others in the funding space, where you know your, you know philanthropic investment was actually leveraging the philanthropic investments of others. So I think one thing that's really exciting about your approach here too, is it's, it's actually solving some of this collective action problem, where, in philanthropy, you know, getting people lined around a vision and coming in together to move the work forward, versus each one trying to push it on their own, meets us in that middle, that where gap is.
Kumar Garg:I mean, we're gonna focus more on the science and tech side, where I think these problems get really acute, right? Because there's a huge fear of looking stupid, you know, which is like, Oh, well, I know AI is important, but how do I philanthropically invest on AI in a way that is actually sensitive to the moment that there's a huge amount of private investment in AI, or it's, you know, the iPhone effect of, like, maybe I'll wait for the next iPhone. I don't want to, you know, maybe it'll slow down, and then I'll catch up. And so how do you invest on an emerging area in a way that is catalytic? Requires, like, a sharp sense of what's not happening that could be a really high impact. So it just caused a lot of people to be both interested but like not doing anything well.
Ben Kornell:And in our space, in education, like R and D investment, which the R stands for research, we just tend to go with the D and, you know, it's costly and it's challenging, and you don't want to invest in some. Thing that's not going to work. And so you do need, you know, more than the million dollar budget, but you also don't need the billion dollar budget for some of these sub areas. So I think while I'm going to be, you know, Alex and I are super interested, both in the Tools competition, will be really interested to see how your model can be transformative, and how philanthropy works across science and technology and education. So Kumar, if people want to learn more about the winners of the tools competition, or learn more about Renaissance, what's the best way for them to find out more?
Kumar Garg:Yeah, sure. So we've got toolscompetition.org and you could just search for learning engineering, tools competition. It's got tons of information on there. We've got a new round of the competition launching in September, and so folks should, of course, check that out. But then we've got tons of profiles of previous winners, you know, competition resource and everything else. I would definitely recommend people check out The Learning Engineering listserv and join I think it's a pretty active and amazing community. And then Renaissance philanthropy has a teaser website up. You can learn a little bit about our team, a little bit about our early work, and it was very much meant to be a teaser, because as we get up and going and start to roll out some of these projects and funds, we'll have lots of exciting updates coming in the coming months. Great.
Ben Kornell:Well, when you have those updates, let us know and edtech insiders listeners, you'll hear about here on edtech insiders. Thanks so much. Kumar gard, co founder of Renaissance philanthropy, thank you so much for joining us today.
Kumar Garg:Thank you so much for having me, and thanks so much for building this amazing community. Thanks
Alexander Sarlin:for listening to this episode of edtech insiders. If you liked 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 you.
