[00:00:00] Alex Sarlin: Welcome to Season seven of Ed Tech Insiders, the show where we cover the education technology industry in depth every week, and speak to thought leaders, founders, investors, and operators in the ed tech field. I'm Alex Salin.
[00:00:20] Ben Kornell: And I'm Ben Cornell. And we're both EdTech leaders with experience ranging from startups all the way to big tech.
We're passionate about connecting you with what's happening in EdTech around the globe.
[00:00:32] Alex Sarlin: Thanks for listening. And if you like the podcast, please subscribe and leave us a review.
[00:00:37] Ben Kornell: For our newsletter, events and resources, go to EdTech insiders.org. Here's the show.
Hello everybody. It is back to School edition of the week in EdTech. Classes have started. We're getting in a role. We're so glad to have you on the pod with us listening to everything going on in EdTech. I'm Ben Cornell, along here with my co-host, Alex Arlin. Alex, before we jump in, what's going on at EdTech Insiders?
[00:01:10] Alex Sarlin: Yeah, so in the long form interviews, we just put out a fascinating interview with Dr. Satya Tta of Merlin Mind. He ran AI education at Watson and is now doing really interesting things in the classroom, including with voice. And the episode that's coming out this week is from Marco de Rossi of We School, which is the one of, I think, the leading, I think, the biggest Italian EdTech startup that does all sorts of collaborative training platforms.
Really interesting guy. Very candid in what he talked about. I thought it was a really great interview. So I, I definitely check, we check that one out. That's where we're at on the interviews.
[00:01:45] Ben Kornell: On the event side on September 8th, we have a happy hour sponsored by Reach Capital and our season sponsor Magic Ed Tech.
It's at four o'clock in San Francisco at Driftwood Bar. We've already sold half of the seats. Well, by the way, it's free, but we're halfway full, so make sure you sign up. We'll be including a link to sign up in the notes as well as our LinkedIn posts. All right, well, without further ado, let's jump in the deep end.
We're gonna do ai, a little bit of K 12, a little bit of higher ed workforce, and then m and a and fundraise activities. So let's start with ai. There's a monopoly of stories here, Alex. What are some of the ones that are popping up for you as real headlines that we should be paying attention to?
[00:02:34] Alex Sarlin: Yeah, so a bunch of things happening in the AI space, and I feel like the headline for this week is Some heavy hitter.
New companies are really coming into play now, so, you know, we've seen lots of relatively small new companies and then our big incumbents, you know, adding lots of features. Now I think we're seeing this sort of middle size, like well-funded, you know, led by very serious tech people, companies that are, you know, have a lot of promise in the AI and education space.
First off, I think, is Prof. Jim Deepak, Shar, Dr. Deepak. Shar has almost 200 patents. He's a fascinating guy. Has, does not come from an education background. Particularly, but has this incredibly wide range of expertise. They just launched something in the last couple of weeks and just put out a paper basically saying that when they're testing their new product, they call it real books against text or against teacher created materials like live teacher created materials.
Their avatar videos are actually creating better engagement, better retention of knowledge. So that's really exciting to me just because we're seeing, it's one of the first times I've seen an AI company lean into trying to actually build an evidence base for the new thing they're doing. Everybody wants to do it, but there just hasn't been that many people in it.
So I'm excited about that. And we're actually gonna talk to Dr. Deepak Sekar and his co-founder Maria Wally later in the episode.
[00:03:54] Ben Kornell: You know, in terms of the mid-size rounds, I, I saw Atypical, which is Bethany Maple, she's a PhD at Stanford, and, you know, basically is building an EdTech LLM similar to the Merlin mind, EdTech, LLM.
And in this case, the idea is that it's a platform that you build multiple applications on top of. And she's starting with assessment. And I just think this is a brilliant starting place. You know, if you want to create transformational change in education, start by the measurements and then work your way back to, okay, how are we going to, you know, support students?
How are we gonna provide personalized learning for them? How are we gonna, you know, dynamically deliver, et cetera, et cetera. So what's interesting about Atypical is, you know, with this 4 million round, they're both a student, LLM, but also really leaning into the educator LLM and then sizzle ai. You know they made some sizzling news with a $7.5 million seed round for their tutoring chatbot focused on middle and high school STEM courses.
This one was led by Owl Ventures, and when you look at the kind of who's who on their team, it is really an embarrassment of riches. I don't know. With the atypical and sizzle raises, what's your read into that? What does that tell us? Both about each of these companies, but also our space in general.
[00:05:19] Alex Sarlin: Yeah, I mean, I think what these all have in common is highly technical and seasoned founders or, you know, highly technical founders, many of whom are seasoned in various ways.
Atypical, as you mentioned, is a AI PhD, Bethany Maples deep in the field, really, you know, understands it very well. I know they worked at Google Mind, right, exactly. And they brought on James genone, who has been working at Minerva for many years as their chief product officer. They're sort of really, you know, trying to gather the forces around things that are deep understanding of, of EdTech, I mean, of, uh, LLMs of, of ai sat netted Merlin minds doing that deep detox core at, at Professor Jim and this sizzle group, as you say.
It's an amazing crew that they brought on. The founder is an XVP at from Meta and obviously he has a few people. His, his product lead, his marketing lead are both X meta, but he has people from Google, he has people from Ave Ismo, from Twitch. So went right over this. But that led by Owl Ventures is not something we've said a lot recently.
Right. Talk a little bit about that. And they also have Lawrence Hal and, and Mike Yates who his from GFA as advisor. So they have a really cool team putting together there and all of these guys do.
[00:06:27] Ben Kornell: Yeah, I mean, I have several takeaways from an investor standpoint. What's interesting is, you know, investors have largely been sitting on the sidelines trying to figure out where is there going to be defensibility in AI and education.
And what we can see as a trend line is clearly there's a bias or an orientation towards more technical founders creating a defensible moats. The other thing that I think is quite interesting is. You know, there's a danger for most funds. If you go in too early with a startup, you might end up, you know, for competitive or conflict of interest, reasons not be able to invest.
And another challenger. And what I find quite interesting when you read about all these companies is the 10 year vision is actually merging, which is basically like we want to be the one-stop shop ai. LLM, you know, toolkit for all of education. And it seems like the venture capitalists have wrapped their head around the inevitability that all these companies are eventually going to be competitors.
And I don't see how they find a way around that, to be honest, Alex, you know, with generative ai, with many of the AI tooling pieces, once you kind of land an entry point with an audience, the pressure's on to like build value and build multiple, multiple tools and use cases for the ai. So I just think that the space is ultimately going to be really, really crowded here in.
Three or four years with people offering many, many overlapping pieces. I guess my last point would be it'll be fascinating to see which is the right entry point, because if they can find a killer use case that is the right starting point to build an audience and then can layer from there, that ultimately will from a product standpoint, tip the scales towards winner on a growth standpoint.
You know, many of the companies that have highly technical people, they're gonna need to find folks who really know the education space and distribution well to kind of pair so that they can find scale and find product market fit.
[00:08:40] Alex Sarlin: Or they can partner with companies that already have a lot of distribution in the K 12 space, which could be, you know, the canvases of the world.
It could be, uh, textbook publishers, it could be, you know, the Duolingo or news, you know, places that have reached a lot of penetration in either the commercial or the school space. One other thing about this that I think is, I totally agree with what you're saying here, that it feels like I'm getting flashbacks a little bit to 2013, the sort of movement where, you know, you had suddenly all of these companies all at the same time.
You had Udacity and edX and Coursera and Udemy to a lesser extent, and all these companies just appearing on the stage at the same time, and there was sort of a. Feeling of, okay, this is a little, this is an arms race. They're all gonna adopt and pivot and do certain things as they grow. And they, they did, but there's a little bit of a sense of it being like, yes, but these are, they probably are competitors.
I mean, maybe one will swallow another, maybe one will supersede the others more quickly and they, and the others will have to pivot really early and they won't seem like competitors. But I agree. It feels like atypical Merlin mind sizzle, all realizing, oh, there's, I gotta say the killer use case though, that you're saying, I don't know if you need a killer classroom use case.
I think that what they're all realizing is if you are the platform, if you are the toolkit, if you are the middleman that allows classrooms to use AI safely and well, and in ways that are really helpful for people like you can become a platform. I mean, this is also what we're seeing with Canvas, right?
It's like, I don't think you need to identify exactly what people need to do quite yet. You just need to make it. Able to use it in ways that, you know, it already does such incredible stuff just out of the box, some of these LLMs that you can, just making it safe and comfortable and easy to use. And having some prompt libraries and some built in, you know, parameters and things is still gonna get people really
[00:10:33] Ben Kornell: far.
I think. Yeah, I mean I come at that totally from the opposite point of view. It's rare that we disagree quite this clearly, and I'm sure that's how they're thinking. Like, just get in and our LM will do it all. This is one of the kind of classic tropes though in education as we focus on the technology and not on the solution.
Oh, AI is awesome. Oh, you know, let's get laptops in the hand of every child. And then what it turns out is you get a laptop and instead of writing your essay on a piece of paper, you type it out. I mean, it's like very, very low substitution value. Whereas the ones that create the opportunity to create killer use cases that create viral spread and ultimately cement your place in market share, it's really around like what is the killer use case?
And I'll give you two examples. One, when Google had Chromebooks, nobody was buying the Chromebooks and they were trying to pitch the value of a flipped classroom and kids could take it home and all of these things. And then one day, you know, no Child left behind mandated that everyone's gonna do online testing and you have to be able to lock your laptops down.
And it turned out like Chromebooks had that feature and all of a sudden every director of technology is putting in order for like 5,000, 10,000 Chromebooks. And that's why we see Chromebooks everywhere was standardized testing for a one week window in May, and it turns out you can use it the rest of the year.
Uh, second example is Newsela, you know, amazing English language arts program that also incorporated the new social studies standards. But the killer use was when kids are coming in from recess or coming in in the morning and you need to take attendance. You play the video and you have them read the level reading and you have them do a little bit of work.
It's the perfect do now. While you take attendance. That was their channel and that's why it took off is teachers saying, Hey, use this while you take attendance. You know, I just encourage these folks to really think about, you know, maybe it's three killer use cases, but what is the thing that's just super solving the problem that nothing else solves?
[00:12:47] Alex Sarlin: Yeah, I hear you. I definitely not quite on the same page there, but I hear you. I mean, saying that, you know, standardized testing and lockdown is a use case that's so coincidental. I remember talking to Doug Levin, who was the head of the State Education Technology Directors Association years ago and saying.
Argue they're gonna be able to do this online testing on iPads, which were very hot at the time. And he was like, no, because you can't lock 'em. You can't control the camera. You can't keep them from get, you know, that one button that allows 'em to get out and go to any other app and look things up or go to the internet.
Nobody's quite figured out how to lock it down. And it was, I mean, that's such a stupid problem. That's the thing. It's like that's not an educational use case in a meaningful way. That's a like bureaucratic logistical loophole that like is so dumb.
[00:13:34] Ben Kornell: This is precisely my point. Okay, why haven't we been doing project-based learning?
Why haven't we been flipping the classroom? It's these types of technical challenges. I agree with that. And so look, I believe in the technical excellence of all these leaders. Hopefully they're getting counsel around. Where they can find these like, you know, scalable and really, really small problems that if they solve them just creates unlock.
The other thing that I think though, you know, in this headline news, so Atypical Sizzle Prof Gym, we also had Course Hero launching an AI academy to help educators navigate AI in the classroom. It is the moment of B2B in education, like everyone's moving. Like remember how much it was in the Coursera era?
It's like B2C. Everything is going to be consumer and we are full on B2B because why? One, the LTV is good and schools and universities have money to spend on these things, but two, this is the fastest way to. You know, lock in large numbers of users and if you're going to be a platform that is multimodal, multifaceted, having like an installed, you know, school, university B2B corporate install base is a great place to grow from.
So I, you know, of course, you know, I love that stuff, so I'm like. This is the moment for building a B2B sales motion.
[00:15:04] Alex Sarlin: There's also no meaningful B2C within actual school systems in the us, right? I mean, there's no buyer there. You can sell to parents, but that's not really school anymore. Like I agree with you that this is the moment and I am very pumped about it.
I think one of the reasons why this is the moment as well is that I have sensed a pretty meaningful sort of turning of the ship over the last few months, I think over the summer, where at the end of last year I still couldn't talk to anybody about AI without somebody saying, but it's about cheating, but it's about plagiarism and you know, we've been talking about this this summer.
I think things are starting to change. There's a really interesting article this week about some of the, yeah, to be clear, pretty cutting edge districts, very urban districts. Mountain View in, in Silicon Valley, New York, LA just launched an education specific chatbot. Like there's starting to be a turnaround where the narrative, instead of being, oh man, this is really scary.
It's gonna replace teachers, or it's gonna be all about cheating, or it's gonna, people are starting to say, wait, this stuff is actually really cool and we really do wanna use it. And I think that's what's causing this B2B moment, because I think there's a sort of realization on both sides, on the startup side and on the school side of like, well, if we could actually figure out how to use this and enable this to be used in interesting ways in the classroom, like.
This is actually kind of a big deal. I, I feel like there's a little bit of a, a zeitgeist shift that over the last couple months. Do you feel
[00:16:29] Ben Kornell: that? Oh, totally. Totally. And, you know, I'm not thinking that it's uniform across, but, and there's a little bit of like red light, yellow light, green light that we're playing here where it's like.
Yes, we've gotta lean into the technology, but red light, it's gotta be safe for kids. And so I do think teachers are a little bit caught in the cross hairs because you're getting signals that you need to be using, you know, 21st and 22nd century tools to help kids and unlock their, you know, creative talents.
But at the same time, you're getting a lot of concerns around data privacy, around bias, around hallucination, and then also around cheating. And so I think, you know, there's a, I was at an event on Wednesday with the superintendent of LA Unified, and he was talking about ed, but it was in conjunction with a group called A-I-E-D-U, which was basically like, how do we infuse AI into our lesson plans around core curriculum in ways that is safe and productive.
For students. So like an example is, you know, learning about genome sequencing and then leveraging AI to think about, you know, how they discovered new proteins. And so I just think this is going to be a year of figuring it out and I will not be surprised if there's also some sensationalist headlines, you know, in two or three months about where it goes wrong.
Oh, for sure. But there is no choice. And for our ed tech folks out there, I do think being a trusted partner to K 12 schools, being a trusted partner to corporations and being a trusted partner to universities, actually there's a services layer here around like being a thought partner that I actually think is pretty high value add for many of these.
And you're seeing who's getting the deals, who's not getting the deals. Some of it has to do with trust in the team and the ability to kind of work through things together and figure it out.
[00:18:29] Alex Sarlin: Yeah. One funny piece of that that I think has been missing so far, and I totally agree, that sort of consultative partnering, actually helping a school and or a set of educators, you know, cross over the safety threshold to be able to use this stuff safely, but also in really interesting ways that actually a curriculum line that are actually useful.
It's not coming from any one place. It's like you see the, you know, A-I-E-D-U and AI for education. We interviewed Amanda Bicker staff who, she does this all the time. She's doing professional development in schools all the time. Trying to, you know, first off, show people what this stuff is and then say, here's how you can actually use it in ways that'll save you time, or teacher students, something really you couldn't have done before.
I think a missing piece is the visual pieces. I always find it so funny that like everybody's always focusing on the, the poetry and the essay writing and the, the writing pieces. Like I get that in some ways. But to me, the thing that's most crazy about the AI right now just that just jumps like madness, is the visuals and the videos and knowing Gen z, knowing these kids, I mean.
When they're unleashed on the visuals and the videos and they really can build these unbelievable things. I mean, just like with Minecraft, people started making whole cities and Minecraft, whole pyramids, like you start giving this stuff to, to kids. It's going to be madness. In a really good way. In a really, really exciting way.
Yeah. Popular with madness.
[00:19:46] Ben Kornell: Yeah. Yeah. I think this is also converging with some other trends that we're seeing in K 12. So let's go now to K 12. You know, I am also thinking that when we interview. The Prof gym team. We should also circle back with some cooler talk afterwards, water cooler talk just around like, okay, how do we make meaning of all of this?
But like you said, basically kids are learning in new modalities, gaming, interactive video, short form, collaboratively. And a new study came out basically saying that Gen Z has a permanently declined interest in college and even middle schoolers are less interested in pursuing college. You see that basically only around 80% believe that they need a college degree of any kind to be successful.
And it's, this is a Y pulse survey. It basically shows from 2019 to 2022, a very precipitous decline in that belief around college being a path to success. Interestingly, they also segmented Gen Z and millennials. And millennials tend to score about 10 percentage points higher than Gen Z around belief in college.
So there's even like a generational gap here where our younger students are even more skeptical about college as a path to success. I do feel like that also plays into this of this new tooling and the new engagement and the personalization. Is it making college pathways irrelevant? The other headline that we, we were looking at is the rise of virtual teachers.
The idea of remote teachers where basically if I can't fill a classroom or, so maybe I only have like 10 students, but I need to offer AP physics, I can combine with somebody else and have a remote teacher, or maybe I can't find a teacher and hire them. So I have somebody from, you know, elsewhere in the rural part of the state teaching them this is another trend that is coming.
So both the delivery systems and the outcomes. Up in mind. K 12. What are your takeaways from this? And you know, this is, it almost feels like we're in a blender heading into the school year because it's just hard to know which way is up and which way is down.
[00:22:10] Alex Sarlin: I think a potential through line between several of these stories, one of the things coming outta this big Why Pulse survey, but it also says two out of five Gen Z students say the pandemic has made them less interested in higher ed.
The number of, uh, people who are planning to go to college went from almost virtually a hundred to 83%. So you've just lost a, you know, you've got some real changes in behavior there and belief about college, but it also goes with Gen Z students saying they're much more likely to choose YouTube. Or Google over a teacher when asked if you wanna learn something new, what resources would you use?
And that feels very interesting to me. 'cause you know, YouTube in this very sneaky way, we talk about Google being the biggest education company. YouTube has grown and grown and grown as an education go to. We've seen TikTok do it to some extent, but YouTube, I, I still think it's sort of the go-to for this young generation.
You wanna learn something. YouTube, it's a video. Somebody will explain it to you, they'll show it to you. If it's something visual, like I think we in EdTech, don't always, you are always saying how Google is, you know, the number one, the biggest EdTech company. But I don't think we always sort of realize how much power there is in that sort of media shift.
And when I see that and think about all these AI tools that are coming, that many of them whom once we get through the sort of safety issues, I think will be delivered directly to students. These sort of AI teaching assistants or tutors or coaches or chatbots or characters. Like if kids are already using YouTube over teacher and starting to say, I'm not sure college is as important as you'd think.
They're already just on this totally different path that is much more autodidactic, it's much more, you know, just in time learning. It's just a very different outlook on what education is. You always have to check out Claire Z's ai Roundup from GSV. It's one of the best things in ai in EdTech, you know, any week.
And all of the big companies are leaning real hard into EdTech, PowerSchool and Instructure, which combined with Google, are basically the, you know, 70 plus percent of the LMSs in in K 12 are all building AI at this very rapid pace and integrating it in, I think this is like a really interesting moment just societally, it's just zoom out and you say like physical educators, people have always been the sort of channel by which we learn.
That's and books, I guess. And now I just think there's meaningful competitors to that in a way that I don't think we've ever seen before.
[00:24:33] Ben Kornell: And if you change the goal of the system. Then you have to like reanalyze every piece of the system. So you know, if you look at the agrarian purpose of education, it was basic reading, writing, math so that people could run their own farm or be part of the agricultural system.
And let's also be clear, many people were left out of that system either intentionally because of race and class or unintentionally. Then we went to the industrial era, which the idea was to give everyone some standard of education, but so it was mass produced and had like a minimum bar of quality, but ultimately it was a sorting hat for the different roles in the factory.
And then you kind of had this World War II kind of GI bill kind of come back where the purpose of school shifted yet again, which was everyone must go to a four year college because that is the path and. Now, you know, and whatever you feel about that, I think the like overall perception is that the ROI on a four year college degree has declined.
And so if that's not the goal, what is the goal? And if K 12 is merely a piece of a lifelong learning journey, which, you know, you and I both agree in, around like the need for chunking, the learning and upskilling, and more rapid time to value in the learning cycle. How do we rethink our K 12 systems? How do we rethink our, you know, longitudinal systems?
And I keep coming back to during the pandemic, our ahas were number one, school is for childcare. Number two, school is for socialization. And number three, school is for learning. It's definitely number three because when you have little kids at home, like you need a place for them to go and interact with other kids, that's like safety, like caring, all that stuff.
So I don't think that there's a world where teachers, I. Are going away totally from K 12, but I do think there is this question around if the goal has changed and we're not going to four year universities, can YouTube be a teacher? Can these virtual or remote experts zoom in and teach? Can we think about the school almost as a collective of resources, tools, and supports that they're amalgamating this kind of list and these resources in a way that's personalized and structured for students.
All that said, I read the article about the virtual teachers and Jennifer, Carolyn made a really good point, which was she was basically like, I can't invest in this technology because it's not something I would send my own children to. While we talk about this distributed model, I. We still see the most affluent families still sending their kids to low tech all day schools with fully staffed teachers.
And so this is the balance that we're gonna have to figure out and that also has implications for higher ed. So on the higher ed beat, what are you watching and what's the news that's percolating up this week?
[00:27:47] Alex Sarlin: Yeah, I mean, well as a sort of segue, I mean you mentioned the four year university changing as sort of the goal of K 12 school and I think, you know, we've seen that coming in and out of, of focus for the last few years, but it's becoming clearer than ever now.
And I think where universities are gonna have to evolve, and I think some of them are really seeing this, is it's just not about a four year period as students. Become more flexible in their lives as they become more skeptical of the sort of accepted paths that their parents did or that their grandparents did.
I think that colleges really have an opportunity, and we've seen people start to do this. We've seen Southern New Hampshire, we saw the concept of open loop University from Stanford. We're seeing some of this, but the idea of like a subscription based university where you can go starting at 13 and you can go until you're 75 and you just pay a certain amount, which is not that much, should not be that much per year anytime.
And, and, and the university gives you a whole suite of different meaningful learning opportunities that could include virtual classes, they could include AI tools. I mean, I can't help but feel that that's where we're eventually going with this. And one of the news pieces that stood out I think to both of us this week was the University of Michigan, one of the, you know, big.
State schools, the flagship state schools in the entire country, just put out a whole suite of AI tools for the University of Michigan community. And that includes, you know, not only a sort of safe version of an LLM that people can use, but also APIs that you can build so that you can connect your own data sources to these LLMs really a pretty smart suite of tools.
That is quick, and I am thrilled about it because that feels like, you know, university of Michigan's have always been very tech savvy and they've had a lot of really smart initiatives around digital learning for a long time. But this feels like almost like a new level of being ahead of the curve, because I haven't heard of any other school.
Maybe they're out there, but I haven't heard of any other school leaning into AI in this way saying you go to, um, you get your own ai, all the professors get it, all the students get it. The grad students. I'm thrilled about it. And then, I mean, we'll see what happens with it. But what did you make of that news?
I know we, we've sent it to each other and we're like, this is really interesting.
[00:30:04] Ben Kornell: Well, I think there's going to be a dividing line here of like, you're either all in or you're all against. And if you're in this like wishy-washy space, you are actually, that's the most dangerous space to be in because then you have incoherence across what students are doing, what faculty is doing, what your organization is doing.
You know, I don't think you're gonna get a dissenting view on this one, like generative ai and just in general, AI tooling is where the future is headed. And if the goal of universities is actually to prepare people for 50 to a hundred years from now, you have to lean into ai. In fact, you want to shape ai.
And so this is like a proactive step. And when we have universities that are preparing leaders for 50 years ago. You end up getting the kind of short-term, thinking conservative, I don't mean politically conservative, but more incrementalist view of where society and technology is headed. But if you have universities that are thinking about the frontier and what, you know, 2100 needs to look like, this is a great thing.
Now what I will say is most universities are leaning into AI and generative AI in their technical fields. So if you're a computer science major or something like that, you'd be getting a lot of ai. What I love about the U of M approach is that this is not, you know, philosophy students are gonna be talking about the ethics of AI and leveraging AI tools.
Medical students are going to be thinking about AI and examining medical studies using the latest ai, and that is where I think actually like a liberal arts, multifaceted university is poised to be a leader and win. So basically over the last year we've talked about a lot of universities narrowing and specializing.
This is actually one of the great counter use cases where you actually say, technology is part of the fabric of everything that we do, and AI is the latest evolution of that and kudos to them for leading the way. A hundred percent.
[00:32:22] Alex Sarlin: A couple of other kind of big deal things are happening in the higher ed space that people may have seen this week.
One is the University of West Virginia has made a lot of news recently. Basically, university of West Virginia cut pretty much their entire liberal arts curriculum. You know, that's the flagship state school of West Virginia. And they made enormously serious cuts to their majors. They left all these students sort of high and dry that they were majoring in something that suddenly doesn't exist anymore.
And that, you know, it's something that we've talked about on the pod for a while, that these liberal arts majors have been increasingly unpopular for decades. More and more students have been saying they wanna go to college to get a job and to major in something that's gonna really help them in their, in their career.
That said, this is a, like, it's a scary moment. It doesn't feel good to see a school cut so many departments that are so much of human experience. What did you make of that news?
[00:33:18] Ben Kornell: Again, goes back to what is the purpose here, and if the purpose is to get people jobs and provide the highest ROI, then it's probably a smart move.
You know, just being really, really practical around what's the nearest step to get someone a job. If the vision of society though is one. You know, the other alternative is that school is also for building citizens that are well-rounded and supporting a community of professions that make, you know, society richer, you know, core to democracy, core to a vibrant civil society, all of that.
It seems like a big abdication of that role by a state university system. And I think, you know, you're gonna see this kind of thinking drive a lot of decision making in red states, and I don't feel like there's a clear prioritization framework in blue states around what universities are for because it's, you know, they're kind of trying to balance this like path from poverty to middle class with also this need for a like just civic society.
But I, I think we've been predicting this would be happening and now here it is.
[00:34:37] Alex Sarlin: Exactly, and, and it's, I mean, they're getting rid of 32 majors, 20 graduated, 12 undergraduate, like all of linguistics, all of math, art, history, you know, landscape architecture. I mean, it's a lot. A lot of these are master's degrees, but Chinese, French, German, Russian, Spanish.
It's a wacky moment. I think you made a really fascinating point there about the red state, blue state split. You know, we have seen that red states Republican run states have been much more comfortable embracing the role of universities and the whole higher ed sector as a sort of. Preparation for work, you know, career advancement than you've seen in blue states.
And it's probably not a total coincidence that this is West Virginia, one of the reddest states, although West Virginia has Appalachian all sorts of interesting things going on. You would not see this happen in Massachusetts, let's put it this way, right? It's a crazy moment and it's a little bit, I don't know.
I don't know how I feel. I agree that there's been something very weird about. Charging students 40 grand a year to get, uh, BFA in puppetry. That was one of the things that was eliminated. Like that is probably not really morally sound. At the same time, it feels very strange to get rid of all languages, all art history.
They even got rid of the higher education degrees, you know, the higher education administration and the PhD in higher education. It's
[00:36:00] Ben Kornell: very weird. And by the way, we should also just acknowledge, I think it, it doesn't do it fully justice to paint it with a red state, blue state brush here. I mean, higher ed is under financial pressure and there is, like you just said, like the ROI is not there for a $40,000 degree in puppetry.
So, and we've been talking about the accessibility via online. To all sorts of courses and you know, our friend Matt Tower would often say, how many economics 1 0 1 classes do you really need? If you've got 10 really good ones that are available to everyone, why create the redundancy? So I, I mean there's a way in which the other angle on this is, this is fiscal responsibility.
And you know, the Wall Street Journal article that just detailed how we've had incredible bloat in administrative costs at universities across country. You know, maybe this is to be celebrated 'cause here's a university doing something about it. I think it's, it's both. It's all of these. It's both.
[00:37:04] Alex Sarlin: No, it, it really is.
I mean there were $45 million budget shortfall. This got rid of a hundred something positions and admin and professor positions. It's just a wacky moment. Another interesting article, and we, we won't spend much time on this, but there was a really interesting little sort of expose article at the Hecker report about test optional admissions.
'cause we're still at a moment where like 80% or a huge percentage of schools are test optional right now. But it's basically saying, Hey, you know, they say they're test optional, but they still look at test scores if you give them, and people who do give test scores like the SAT or a CT have a higher acceptance rate.
Significantly higher in most of these schools. So they're saying, Hmm, you know, it may not be as fair as it sounds to make these tests optional because people with the advantages to take them are still taking them and they're still helping enormously get to get those students into the school. So, interesting article that will link to it in our show notes as always.
I don't know if you have thoughts about the test optional world, Ben, or should we move to funding?
[00:38:03] Ben Kornell: I mean, on the test optional. I think there's a lot around like signaling and the competitive nature of admissions that makes the kind of policy shifts in name only. And there's also was an article around, you know, affirmative action and how schools are working around the affirmative action ban, but still using other indicators to build diverse student populations.
I just think that one of the things that we have to be careful about is if we say to kids. It is test optional. The majority of kids who have parents who have not gone to college will take things at face value. And the majority of people whose parents have gone to college will understand that there's a game here and that you play the game by the unwritten rules as well as the written rules.
And so just another example of us not doing any favors to ourselves, to our society, and especially to our least privileged kids. Yeah, really well said.
[00:39:08] Alex Sarlin: So, you know, a couple of the big funding rounds this week were pretty specifically about addressing sort of teacher burnout and giving teachers tools that make their lives.
Easier and better and more comfortable and make their teaching better. One is Magic School, which raised $2.4 million. That is a suite of AI tools. Similar to some of the other ones we've talked about on the platform. They have over 50 tools now, and even though they're only a few months old, they raised 2.4 million because they already have a hundred thousand teachers.
You know, they're getting traction in many different districts. So this is interesting because it's starting to, you know, we saw lots of different lesson plan generators and quiz generators and things, but I think these, you know, IEP generators, but I think some of these tools that put them all together.
Are starting to get traction and you might see some sort of separation where some of the people pivot away if they're not getting traction or try something else. And the ones that do are, you know, getting funding, leaning in, being able to hire and accelerate.
[00:40:07] Ben Kornell: Yeah. And on the magic school raise, the interesting thing that I find is the number one value prop is save teachers time.
It's not AI can do amazing things, which it can, and which they're using, but they're leading with save you time. And then the use cases are like, make a quiz, level the text, do a quick assessment, generate a lesson plan. Like it's the kind of ease of use and time to value that is the time saving part of it.
And I thought that, you know, very, very well positioned there.
[00:40:41] Alex Sarlin: Yeah, and they have a sort of who's who in a lot of ways of educational investors. They have, uh, GSV and Rethink Charter School Growth Fund. Transcend, yeah. Ate, so there's clearly, I think a hunger from investing. And there, there, it was led by Range Ventures, a hunger from the investment community to find which of these.
Slew of AI tools actually are hitting the ground running and just as you said, Ben, the ones that are, are, at least in this case, one that's purely about teacher efficiency and that is really resonating with people it seems like and could, could start to really accelerate. Is also has a teacher and principal I.
Founder. Yeah. Which is always a really nice thing when you're thinking that
[00:41:18] Ben Kornell: way. So another round was creatively focused. Also a K 12 company, they raised $3 million and it's really focused on special education talent pipeline, both how do you upskill paraprofessionals into full-time special educators, but also how do you retain special educators from leaving, you know, keeping them from leaving the classroom.
It is really interesting, you know, the teacher crisis, like step one is stop the leaky bucket and step two is fill it up. And so very interesting company here, which is really around kind of the grow your own phenomenon. The other element that is interesting is it's not super like Tech First, it's people First and The Round was led by York, ie.
With Mayors and Power Venture Capital, groove Capital and Gopher Angels. So we see some new players dipping their toes into the K 12 education space, and I think this special ed market is actually becoming one of the hottest markets where small and mid-size companies are finding traction and real revenues.
You know, it's just talking to IEP and me. It's just very interesting how that particular niche is one typically very, very underserved, but two has the funds to really invest in new solutions.
[00:42:40] Alex Sarlin: That's true. That's a great point. But it also, you know, we've talked about IEPs as being legal contracts that have basically medical data in them.
It is high, is a little bit of a highwire act doing AI with IEPs. So I think both Magic School and IEP and me and a couple of others are trying to walk that high wire. 'cause you don't want that to become the first scandal of the IEDU era. Tell us about QM
[00:43:02] Ben Kornell: Math. Yeah, so QM A is a very, very popular Indian EdTech company and we don't have time to go into the kind of Indian EdTech scene, but we had another large layoff.
The CEO cited a bad macro situation. His name's Manan Kma and basically trying to get the company back to a path of profitability. They raised 57 million in the last year and Google is one of their big backers. So you've heard about Q Math from us. In the past, but you know, I'd also say in the scheme of things, laying off, uh, a hundred employees versus some of the layoffs that we've seen with thousand or even 10,000 employees is probably on the spectrum, relatively small, just more signal that the market in India is coming back down to Earth and companies are kind of moving from growth mode to profitability mode.
[00:44:01] Alex Sarlin: We also saw the COO of UN Academy leave that company, that's another big ed tech unicorn in India. So, you know, definitely some churn happening there. And you could argue, you know, that I think the by Jews affect the idea of like the sector people are feeling a little shaky about the sector and that probably, that definitely includes investors.
It might even include, you know, CEOs and people inside the company. But by Jews itself announced there will be no layoffs a few months ago. But I, I mean, you know, it's a very, very tricky moment in Indian ed tech right now, which has been, you know, have been booming and, and still, you know, very well made continue to boom.
But I think it's facing a dip in confidence.
[00:44:41] Ben Kornell: Yeah. Well, with all of these trends kind of converging, I think it's a great time for us to kind of get under the hood with the Prof gym team and dive into one of the ascendant companies that could be shaping the world for our kids and our investors and our space.
So let's dive in.
[00:45:00] Alex Sarlin: So as always, we've been talking so much about AI and how it's changing the educational landscape. Our guests today are doing this in a really interesting way. We have the co-founders of Prof Gym. This is an AI education company from Silicon Valley. We have talked to them in the past on the podcast, but they have some really interesting new products and we're really excited to talk to them today.
Dr. Deepak Shar and Maria Wally, co-founders of Prof Gym. Welcome to the pod.
[00:45:27] Maria Walley: Hi. Thanks
[00:45:27] Alex Sarlin: for having us, Alex. It's great to talk to both of you. So first off, for those listeners who don't know yet what Prof GM is, who may not have heard that previous episode, give us an overview of what it does and how it transforms content into interactive video.
[00:45:41] Dr. Deepak Sekar: So we use AI to make interactive video awesome and easy to do as well. So the way our technology works is you can feed a PDF of a textbook into our AI and it automatically creates a course. And the course would have not just teaching content, it would also have assessments, it would have interesting tidbits about the content whole lot more.
The other option we provide. People can just indicate what topic they want and we can automatically create some teaching content on it. And the way we do that is by leveraging material on the internet, which is, uh, not copyrighted. And we leverage, uh, GPT-4, so you name any topic under sun, we can create a 15 minute video on it just at the push of a button, which is pretty remarkable, I believe.
And so we call these sorts of textbooks converted to videos or internet material connected, converted to videos. We call them real books. Mm-hmm. So that's real books. REEL. Like real of film, right? Yes, exactly. Yeah, exactly. And the reason why it's interesting is there's a lot of studies out there which indicates, uh, gen Z and Gen Alpha, they prefer videos compared to traditional textbooks.
And in fact, that's why TikTok is so popular, right? Among younger people, because people like video a lot. And because we convert these textbooks into videos, we are seeing learning outcomes improve quite significantly.
[00:47:12] Ben Kornell: So cinematic textbooks that have AI interactive avatars, I feel like I am, you know, launching forward into the year 2050.
The headline news from the K12 deep Dive was that you saw parody with engagement and outcomes with human instructors. Maria, can you just tell us a little bit more about the product, but also what you're seeing around this use case? And also I think for our listeners, when is this like a compelling use case and when are the human instructors compelling and how do you make sense of that?
I'd love to hear your perspective.
[00:47:51] Maria Walley: That's a really great question. So boils down to it is we are trying to make it easier for educators to bring their text to life. We want to make it easier for them to make really awesome, compelling videos. So this is a way for educators really to take the wheel and they don't have to, you know, be have all the studio equipment or sit down and have a bunch of time.
Because yes, there's so much going on in that space on YouTube and the like, but. That's time consuming. That's a career, really. And this is for your average teacher to go online and make a video using our technology. And it's also for textbooks to do this at scale as well, so they can, they have all this rich technology, but students don't like reading.
They're going on YouTube. So we've done some intriguing experiments. So one of them, we had 700 students and they spanned all different age groups in geographical origins. And they were split into two groups. So 350 of the students tackled an obscure topic through traditional textbook based lessons. And the other 350 had those same lessons, but using our technology, using real books and bringing it to life.
And the result was that our approach yielded significantly superior learning outcomes. And which is so important because like EdTech is super exciting, but if we don't have learning outcomes. What's the point? So again, this was at a very obscure topic, but the first group exposed solely just to the text.
They answered 34.6% of the questions accurately, and the people who use our technology responded to 47.3% of the text. Again, these were very obscure topics, but to be honest, that means that our real book users had about 36.8% better in terms of accuracy, which is I think a huge leap forward in learning efficacy.
[00:49:47] Ben Kornell: So is the vision for real books to be used in combination with the educator as an alternative to the educator, as an alternative to textbooks or all of the above? Like help us understand like in the wild. What's the recommended, you know, use context.
[00:50:05] Maria Walley: So we do not believe that the educator will ever be replaced simply because people are propelled by relationships, especially the younger students in K through 12.
However, we do see this being used for all kinds of scenarios. First of all, there's substitute teacher issues and there's a huge substitute teacher shortage right now. I mean, there teachers are leaving in droves post pandemic, but like substitute teachers that was already an issue before the pandemic.
So this is something that teachers, they don't have to feel guilty about taking a sick day and they can prep these in advance. So there's sub the substitute teacher bit that they can use. And there's also things like homework. This is something that, it's an interactive, interesting bit of homework or a student is sick, you know, essence of the lesson that the.
Teacher created and they don't have to grade this homework. And this is also something that we see textbooks using and creating their own courses, their own lessons, and really rich content that they can just really, I think this gives the power for textbooks to kind of solve the educational crisis in a way and meet.
Their users and very meet their consumers. Straightforward
[00:51:19] Dr. Deepak Sekar: and other use cases, we are seeing, there's a lot of people working in companies, but they wanna get, take courses on the site to learn about AI or some other new area upskilling employees. Uh, some of them do the master's degree on the site. And for those online learning lessons created, uh, using our content is great.
And one of the use cases we see a lot is if you look at Coursera, right, it's $2 billion in market cap and they have 7,000 courses. While if you look at a company like McGrath. They have 200,000 books in that catalog. So if you look at all the publishers out there, there's a huge opportunity for publishers to take that catalog and convert them into course marketplaces bigger than Coursera, and actually generate a whole lot of revenue.
And that revenue potential is not tapped by many publishers right now. So we are working with a number of publishers on things like this. So some of our customers are, one customer we've talked about publicly is engage. Another customer we've talked about publicly is LinkedIn Learning, which creates a lot of homeschooling and hybrid schooling, video content.
And there's many others whom we are not yet able to name. But we are seeing huge traction for this sort of video creation all over the industry.
[00:52:37] Alex Sarlin: One thing I find really interesting about these studies that you're doing, you know, you mentioned Maria, the first one is text versus, you know, auto-generated video.
And then you've done one which is, you know, online teacher, a seasoned teacher doing a screencast video versus an AI assisted video. And then you've done one to test sort of the delivery, or you have the exact same script and it's a human versus an avatar. And you know what I love is that you're really systematically trying to show.
That this is something that is viable, that gives students choice about what medium to use that gives publishers this way to very rapidly convert their content into a medium that students may prefer or teachers may prefer if it takes less time for them and they can do other things. I love the way you are sort of splitting up the problem and going at it with studies with 700, 200, you know, students at a time.
Tell us about that approach. You know, why did you decide to do these sort of mini studies inside the company? Is it because you're getting questions from your potential clients or is it you wanted to know yourself? What's the origin of that?
[00:53:38] Maria Walley: Well, I think how we continue to build the product, we need to build it on results and where do we come up with like the direction versus just, you know, good old fashioned scientific method, experimenting.
To that end, you talk about, you know, the teacher versus the avatar. One thing that's so interesting is a lot of teachers. Again, screencasting takes a really, really long time with one of the teachers in terms of like trying to find the material, the designing of the, the slides sitting down and like getting set up.
She spent a lot of time doing that. I believe it was eight hours. However, with our AI technology, that exact same lesson and dare I say, looked a lot more professional, took your. 30 minutes after she kind of used the AI and looked over it and made, you know, the necessary edits and made it her own. So that was comparing kind of her homemade lesson screen casted with the AI end to end.
And then the other one we studied just the avatar and what we found out is that our avatar is just as engaging in terms of keeping the student's attention as a very engaging presenter. So that was, that wasn't just your average presenter. That person was also very used to presenting online as well and that particular experiment.
So that was really exciting for us because, you know, these aren't. Real people, like how excited are people going to be, you know, listening to an avatar. And we've made our avatars built to teach. So they, our AI voices was Deepak can kind of explain, we've made them homemade because all the AI voices out there, they weren't up to our standards.
They were just a little monotonous. And so we've made, made our own voices to be just a lot more expressive. So it keeps the student interested.
[00:55:32] Dr. Deepak Sekar: I. And the way the technology has improved. Okay. So we benchmark the technology by looking at learning outcomes. And so in the beginning, six months back, the AI was not performing as well as the screen casters.
And then we kept improving the technology. Yeah, we kind of obsessive about it. So we came up with our own AI voice technology, which I believe is the best voice you'll find anywhere on the internet. And we had to come up with a whole bunch of innovation, five, five patents worth of technology for that.
And so the AI voices now are pretty expressive. And then on the video itself, we made the videos better and better. And now six months back, most people could detect that it was an avatar. It wasn't real. But now we fool two thirds of people out there. Maybe more we have people asking, Hey, is this real or is it an avatar?
Yeah.
[00:56:21] Maria Walley: Yesterday, someone was looking at our Lincoln lesson and they're like, did you have an actor here? Did you dress up like Lincoln? Where'd you get the clothes? And we're like, that is ai.
[00:56:33] Ben Kornell: I will just say I, I feel like there's a lot of innovation here, but it's also a very classic concept pedagogically that you're relying on, which is that one great content coupled with high engagement leads to great outcomes.
And the game-based learning for the last decade has really demonstrated that more engagement. With content in an interactive way increases, you know, uh, student performance. And so really you're just leveraging the newest technology to actually build on these foundational pedagogical principles. What I find really interesting too is this combo of, okay, now if we can extend the educator and extend the textbook, you know, in terms of the engagement layer, how might we rethink the capacity of those educators?
Because, you know, as you were mentioning Maria, that's actually our scarce resource is educator time and availability. So I think it raises systemic opportunities if we're able to, you know, build upon the great work you're doing. Alright, last question. So you're an education company, but you're also an AI company.
Who do you see coming from the generalized AI space that is a competitor or they're doing things that you think are really interesting? Because one, one thing Alex and I are watching is, you know, generalized consumer products kind of dipping their toes into the education space. And I think that that is also, you know, something we're gonna have to watch.
You know, Google is the largest s tech company in the world. It just turns out that it's a rounding error for them. And so purpose-built products versus AI consumer products. Who are you looking at and how do you think about these spaces Melding.
[00:58:27] Dr. Deepak Sekar: Yeah, so that's one thing about ai. You know, it's, you see hundreds of AI companies, but when I look at some of those companies, I'm like, Hey, that's your technical work is something I can do in one afternoon.
Now how are you differentiated? Uh, and I think about that with like half a dozen companies a week. There's a lot of people just playing with GPT and getting the output. So for me, as a technologist, you know, being differentiated is pretty important. So the US Patent Office actually gave us, uh, a pretty fundamental pattern a few months back on converting textbooks into avatar videos automatically.
And so. If someone else tries to take textbooks, full textbooks and converts them into avatar videos, they're likely infringing our patterns. So that's one way we are differentiated. And the, the reason why we got these patterns approved is we didn't start working on it after chat. GPT took off. We've been working on it since 2020 before chat g PT was a thing.
And so we find a lot of our patents back then, so that helped. As far as the whom I consider challengers on the generalized side of things, there's a number of these companies like Synthesia. Which make these videos automatically. The main challenge with those approaches, and Synthesia has a number of clones as well.
There's like every month I see a new clone show up. And so the main challenge with those folks is they can scan textbooks. You need to enter text line by line. And if you're going to create education content at scale, it's pretty hard to enter things one by one and create slide decks and stuff. It's good for one minute marketing videos or you know, shorter courses, but you can't create content at scale.
And then there's a whole bunch of startups which just take GPT and claim. They do personalized learning and all, but most of them just create text output videos where today's consumers are. And we like to think we are fairly differentiated in terms of creating education videos. Uh, and we're kind of focused on that.
And so if you look at our content, it's not just videos in the middle. We've got assessments, they've got personalized learning path. We actually have games in the middle of our lessons as well. So the student has a full experience. And I like seeing we have some of the best, uh, instructors in the world, uh, because no one else has Lincoln and, and Jane Austin teaching for them and the kids get a kick out, uh, seeing all these people come to life.
I'm
[01:00:53] Alex Sarlin: still waiting for my virtual avatar.
Yeah, we've had so much work we. But it's really fascinating and we, we will link in the show notes to this episode to some of your most recent videos, including this sort of Lincoln Explains the Gettysburg address. You can see the quality of the visuals, the quality of the avatar are just continuing to improve and improve and improve.
And it's just getting to a point where, yeah, it sounds like people are getting already sort of having that you're beyond the uncanny valley. People are already like that. That seems like a person. It's a really interesting moment in education and I love that you're really, you know, working to accelerate, you know, to free up teachers instead of eight hours doing these screencasts, which they were all doing during the pandemic, especially really short time, but also leverage these huge catalogs of text material that you know have, are sitting in book stores and warehouses and places and we're entering a whole new era.
So I'm really a fan of what you guys are doing at Prof Gym. Maria Wally, Dr. Deepak Sekar, co-founders of Prof Gym. I. Check it out. Really, really interesting stuff. Thanks for being here with us today.
[01:02:00] Dr. Deepak Sekar: Thank you.
[01:02:01] Alex Sarlin: Thank
[01:02:02] Dr. Deepak Sekar: you for having us.
[01:02:04] Ben Kornell: Okay, so Alex, that was a fascinating interview and I would be remiss if we didn't just digest that a little bit.
So the headline is like AI Achieving parody with educators. And you and I both know Prof. Jim has been like at the forefront of leveraging AI for interactive content for a long time. Pre-chat GBT. I'm like processing from the interview and I'm like, what does this mean? Right? And is this real? Is this not real?
How do you make sense of all of this?
[01:02:35] Alex Sarlin: I think there has been a story, and I'm calling it a story. It is a little bit of a narrative, but I, I also do think it's a true story for generations of ed tech saying, you know, every time something comes out, people say, wait, is this gonna replace teachers? It was true of radio is true of television.
It was true of, you know, the internet. It was true of Google. And I really think that when I see something like this happen. I think what the big innovation here is, is not really about competing with teachers. I think it's about competing with written text. I don't think it's a coincidence that the biggest of the three studies they did was actually comparing written text, like text directly out of textbooks or out of websites versus their, you know, interactive multimedia.
Because I think if you think about their business model, they're selling to publishers, they wanna tell McGraw-Hill and Cengage and all them, Hey, you have 200,000 textbooks. They're sitting in a pile and yes, you'll continuing to sell 'em, you know, bit by bit. But this, we can turn 'em into something that would, you know, light kids up.
I think that's the core message here. I have never seen anything truly replace teachers. Ed tech has been around a long time. I still don't think it's gonna happen here. And, and when I've talked to tutoring companies, they say the same thing that we just heard from them today. Their relationship teacher does way more than convey information.
They're about, you know, mentoring, they're about modeling, they're about actually knowing kids, personalities, personalizing. I just see this as a, a huge tool in teacher's toolkit rather than a replacement. But I, again, I'm maybe naively optimistic about this. Others totally disagree. What do you think?
[01:04:08] Ben Kornell: Three things standing out to me.
One is, I love your point on the content side. Educators have been using YouTube, TikTok, and other short form video for years now to make learning material come to life. And more often than not, the textbook is either a prelude or it might even be a door stop. And I think the fear that I, as an educator, I, as a parent, I, as a school board member would have, is are we building the endurance and like studying capability of students to dive deep into texts that may be dry and still learn and extract value from it?
Or are we making everything so convenient, fast food based, like fast knowledge that we're creating a, like a superficial culture that only has like a, you know, 22nd attention span and you know, self. So it's like, do you lean into where the kids are? Do you lean into where the technology is going because that's how you reach them with a like cognitive load of, you know, bite-sized chunks to kind of build up their concepts Or is really teaching, learning about a process of, you know, it's like a distance runner needs to build up that endurance to go intellectually deep.
Second point, which I loved was defensibility. I mean, man, I asked the question about competitors and Deepak immediately goes to why are we defensible B2B AI companies that have product defensibility are seemed to me to be the best bets in the space. Whereas consumer ai, there is no lock-in. There's just like, it's so easy to switch out.
And other than Chad, GBT and some of these main interfaces, it's gonna be really hard. I mean, I was thinking character ai, you know, I'm having Lincoln do the Gettysburg address for me when I go to like character ai, it is so superficial. They're building real depth and then they're selling it to the textbook companies because those are the people with the content that need to come alive.
So I wonder, I. You know, in the AI space, is it going to be the B2B models and the B2B platforms that essentially, you know, unlock AI capability for other businesses selling to schools and selling to consumers that are gonna win. And the third thing is, you know, I think it's also still early. You know, the studies are like 350 people.
Sure. 700 people. Like a short
[01:06:47] Alex Sarlin: of course, right?
[01:06:48] Ben Kornell: This is like the canary in the coal mine moment where you're like, here's where we are today. In five years, in 10 years, where is this technology going? You know, I'm an educator. Like, I don't think we're gonna replace the relational pieces of educator. But can we replace some of the instructional pieces of an educator?
The controversial statement here, but I think it's coming, and this means that the educator role is more of like a coach, guide, enabler of class culture, you know, supporting learners. It's a social, emotional community, interpersonal role, which it is now today. But the instructional components I could totally see kind of coming off teacher's plates, especially if you go to like elementary and middle where you have multi-subject teaching where it's very hard to be like an expert in that space.
So, man, oh man. Does this feel, like I said, I feel like it's like a little glimpse into 2050 when we all have our robo, you know, instructor avatar, and then we have a teacher who's checking in on us who's more like a social worker care co coordinator, like group organizer. Fige, you know, all those things.
So, wow.
[01:08:05] Alex Sarlin: Think about community college. Think about that model. Yeah. I mean, with the instructor already, it is not, you know, always leaned on to be a hyper pedagogical expert or anything. And one more quick point is that Deepak and Maria are thinking about this in so many different ways, and they've been leaning into the, like you said, the character based stuff, right?
The like, oh, it's Lincoln and it's Jane Austen and it's Grace Hopper, and things like that. And how cool. That said, they can make avatars of any person. You could have the same avatar teach you every subject, or you could make the avatar your own race and gender if you prefer. Or you can make the avatar a AI influencer who you love, or you could make them, you know, Superman.
I mean, there is a whole world inside that that we haven't even yet unpacked. You know, they're doing the historical side, so is to our companies like Hello History, but the idea of being able to have any simulated being become an instructor and then be able to pull all the textbooks and journals by the way of the world into that person's delivery, you know, Corpus, it's pretty mind blowing and it really does, it feels like outta Starship Troopers or one of these crazy kind of wacky futures, you know?
[01:09:17] Ben Kornell: So what do you think who's gonna win in that space? Because you know, for example, professor Jim is building off prof, Jim, they're building off this huge corpus of textbooks as kind of a means of bringing learning to life. And then, you know, the avatar layers on the video, layers on as engagement mechanisms.
But if they had, let's say, $500 million in cash and like infinite, you know, engineering capacity, they could essentially morph into this is your personal tutor, you know, super fast. But that's what everybody's going at. Like that's everybody's answer to if you had unlimited resources. So from a market force standpoint, I'm like looking at is there a double-sided marketplace here where at a certain point you get enough users and enough, maybe it's content providers that basically your quality or your program surpasses others.
I'm not sure that Prof, Jim, or anybody is quite there yet, but, and what I'm struggling with is, given that AI is kind of peanut butter as a capability across the space, what are the business models that are gonna create the, the scalability on demand and supply side that creates the defensibility more as a business model and not just technically.
[01:10:38] Alex Sarlin: It's obviously a wildly open question, but I agree with you that what I think is very clever about their B2B model, that working with publishers is one, the publishers have all the content. This is a very easy thing for them to test. It's a very easy thing for them to try, and if it works from a sales perspective, from an outcomes perspective, they're gonna be all in.
They can work with multiple publishers and they're white labeled, of course. So there's no, you know, and. The publishers have relationships with all the schools, right? I mean, so it's not just that you're getting into the hands of EdTech companies. You're getting into the hands of the EdTech companies that are still the most entrenched in schools throughout, at least the US if not the world.
If you, you know, you go to Oxford Publishing, you go to all the big textbook companies in India and India, Indonesia, like that to me is really interesting. They're sort of injecting AI into a pipeline that already exists in education. Instead of having to create a whole new set of relationships and convince teachers individually or convince districts that you know, Hey, you should be using ai, they're saying, Hey, it's gonna come in through your textbook company, your your same publisher that you've bought things from.
And that's true for higher ed too, for, you know, decades. That's interesting to me.
[01:11:48] Ben Kornell: I mentioned it in our interview that Google is the largest a tech company in the world. There's a way in which the Google strategy. The Microsoft strategy is get lock in from early learners into your ecosystem, and then they just think, I'm gonna make a document.
Oh, I'll do a Google Doc, or, oh, I'm gonna open Microsoft Word.
[01:12:08] Alex Sarlin: Right?
[01:12:08] Ben Kornell: There's a way in which the big home run dream is becoming the personalized assistant for everyone in the world. You know, that's the ai, like Holy grail is there's a company that everybody, it just like learns me and is my AI personal assistant.
As I navigate the world, schools as a channel to acquire customers into the AI personal assistant is a top of funnel strategy that eventually is going to play out. And so could Prof Jim, who, by the way, nothing in the, that interview said that's where they're going, but could that, that kind of distribution strategy, like.
Leveraging textbooks and building an installed base, could that ultimately pivot into, you know, the upstream version of a personalized assistant? It could, and this is where I think we as education industry people need to understand that we are part of a larger business system for ai, personal assistance, where people will be looking at our space as, you know, essentially a loss leader for.
Like getting user installs that then fuel a lifelong, you know, avatar relationship, you know, AI assistant relationship for all their purchases, et cetera, where all the money is,
[01:13:35] Alex Sarlin: I mean, so many ideas here. I know where, you know, once you have a system like that in place, the same way that in, you know, apple, you can buy sticker packs or you know, or Google, you can buy, you know, add-ons.
You'll be able to buy, you know, the Pokemon. Avatar pack from Nintendo added onto that system. It'll be a platform, right? You'll be able to do all sorts of crazy things. And he mentioned Synthesia and all the clones, you know, that are coming out. And I think, you know, I love your question about this, and I think it is not a given that you know that a Prof gym versus a Synthesia would be ahead.
He's right that if they're being built for sort of mass adoption, like getting whole corpus of textbooks in, that's a differentiator for schools. But Synthesia model is going through corporates, right? It's saying, oh, any company, if you're gonna start making communication videos or training videos, or any kind of marketing video, anything you want in a company, we're the people who make the person and you give them the voice and you go from there.
That could also be another way to get into the work stream
[01:14:33] Ben Kornell: of, and the big question in all of this is. Who owns the data? Who gets the data. And as a case wealth person, that is a very, very like hard and fraught conversation around child data. And you know, prof, Jim being built for education context with publishers is gonna have a far more sophisticated and.
Protectionist view of the data that it's getting back and forth from the users, and like you said, it could be even white labeled with the platform. Sure, sure. At what point though, does the user sign off or, you know, maybe it's when they're over 18. At what point in the top of funnel? Does the learner become linked to the avatar such that they want to continue with that avatar as a lifelong companion?
That's really what's like, you know, my brain is a little splattered against the wall here thinking about that stuff, but community colleges, by the way, Alex, like, what a great idea. Like that is the entry point. Totally. That could be huge. Scalable corporate, I don't think, because corporations are gonna want to keep the data and restrict it.
The avatar from pivoting out. Into like your lifelong avatar.
[01:15:51] Alex Sarlin: Such an interesting set of thoughts there. I mean, the other thing that you can imagine, I mean, I don't know how Prof GM's agreements with the, the likes of the cengage's of the world work, but my guess would be that Prof GM offers them a tool which allows them to turn their content into these suite of interactives as well as make new content.
You know, he, he mentioned that sort of in passing, but it's a big deal. I think the data is still probably owned and protected by the textbook company right now. I, I might be wrong. I would think that if they want, it gets into some very interesting territory if that's not how it works. So that's my guess about how it works right now.
And textbook companies have decades of, you know, protection for FERPA and COPA and making sure that they're, they're compliant. So that's another reason why it's sort of, I think, a clever strategy is that you going on the backs of people who have gone through all of the hurdles of working in this, you know, highly regulated environment.
[01:16:43] Ben Kornell: And so let's take Cengage as an example. Private equity owned. Basically they have verticals of like workforce learning verticals that they've been really doubling down on. So, you know, they have some K 12 career path line, they have some university programs, but their big new thing is really around this connection from, you know, almost like 16 to like 35 or 50, where you're basically upskilling.
And what I think is brilliant about their model is they basically like time to credential is a way they think about, you know, chunking the value. So it's like, I'm gonna learn these things and achieve this credential that allows me to get a promotion and a job or be eligible for a new job. That is a vertical that is super fragmented in the rest of education, but they're stacking that up.
And then you lever, you know, AI and data across a user's arc. And by the way, they're like doing corporate partnerships, university partnerships, school district partnerships, consumer. There's a way in which if they're able to attach the data pack to the, the learner as they, you know, go across all these different environments, they too could be positioned to be like the lifelong advisor, avatar, things like that.
So, you know, I think from a listener standpoint, you're getting a little bit of Alex and I's water cooler talk. You know, this is Alex and I hang out. This is what we talk about listeners. I think it's really watching. The key here is. Let's watch who aggregates and how, like user AI assistance and how does that set them up to pivot into broad mass market scale.
That could be the next, like not just billion dollar, but like trillion dollar company. And it could be, you know, we're like at the birth of the river up here in education. Oh yeah. We could be incredibly pivotal to that happening.
[01:18:45] Alex Sarlin: And just last thought, Deepak has 192 patents, maybe more like this is not a person who is just, you know, just out of college starting to try, you know this thing when he says, here's our strategy, we've already patented this whole thing.
It's not like that's coincidence
[01:19:01] Ben Kornell: for founders out there. You should be thinking about applying for patents. That's like a great piece of advice from Deepak. I.
[01:19:08] Alex Sarlin: Oh yeah, they got a headstart on it because they've been working on it for a few years before sort of everybody lept in. But that matters. And I mean, for better or worse it matters.
I think that is also going to be an interesting potential for a moat or a defensibility that we haven't talked about entirely. Right,
[01:19:23] Ben Kornell: right. Well, that's a wrap for our show. It's been great to water cooler talk with you, Alex. Listeners, this is one where we want to hear what you think. Please comment in on our LinkedIn posts or comment in to us directly.
It's EdTech insiders.org. If it's happening in EdTech, you're gonna hear about it here on the weekend. EdTech from EdTech Insiders.
[01:19:46] Alex Sarlin: 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.