Edtech Insiders
Edtech Insiders
Week in Edtech 12/17/25: Coursera–Udemy Merger Shakes Online Learning, AI Platform Wars Heat Up, Purdue Mandates AI Skills, Higher Ed Faces Consolidation, and More! Feat. Isabelle Hau of Stanford Accelerator for Learning!
Join hosts Alex Sarlin, Ben Kornell, Michael Horn and Dhawal Shah as they break down major moves in online learning, AI, and higher education shaping the end of 2025.
✨ Episode Highlights:
[00:00:00] Coursera and Udemy announce a $2.5B all-stock merger forming a 175M-learner platform
[00:00:30] Michael Horn on Coursera’s growing leverage over university partners
[00:02:08] Ben Kornell explores Coursera’s potential to become a global university
[00:05:40] Dhawal Shah explains the financial motivations behind the merger
[00:09:54] Michael Horn compares the deal to the 2U–edX acquisition
[00:11:54] The hosts discuss channel power and aggregation in edtech
[00:16:49] Debate on Coursera’s acquisition strategy and platform future
[00:21:43] Dhawal Shah on why these businesses may perform better as private companies
[00:22:33] Ben Kornell outlines Coursera’s two paths: efficiency or AI-led reinvention
[00:30:24] Gaps in online learning around mentorship and advanced skills
[00:35:29] OpenAI’s latest release and rising competition with Google Gemini
[00:38:55] Why content and IP still matter in the AI era
[00:48:36] Purdue introduces an AI competency requirement for graduates
Plus, special guests:
[00:52:54] Isabelle Hau, Executive Director of the Stanford Accelerator for Learning, on human-centered and social AI in education
😎 Stay updated with Edtech Insiders!
Follow us on our podcast, newsletter & LinkedIn here.
🎉 Presenting Sponsor/s:
Every year, K-12 districts and higher ed institutions spend over half a trillion dollars—but most sales teams miss the signals. Starbridge tracks early signs like board minutes, budget drafts, and strategic plans, then helps you turn them into personalized outreach—fast. Win the deal before it hits the RFP stage. That’s how top edtech teams stay ahead.
This season of Edtech Insiders is brought to you by Cooley LLP. Cooley is the go-to law firm for education and edtech innovators, offering industry-informed counsel across the 'pre-K to gray' spectrum. With a multidisciplinary approach and a powerful edtech ecosystem, Cooley helps shape the future of education.
Innovation in preK to gray learning is powered by exceptional people. For over 15 years, EdTech companies of all sizes and stages have trusted HireEducation to find the talent that drives impact. When specific skills and experiences are mission-critical, HireEducation is a partner that delivers. Offering permanent, fractional, and executive recruitment, HireEducation knows the go-to-market talent you need. Learn more at HireEdu.com.
As a tech-first company, Tuck Advisors has developed a suite of proprietary tools to serve its clients better. Tuck was the first firm in the world to launch a custom GPT around M&A. If you haven’t already, try our proprietary M&A Analyzer, which assesses fit between your company and a specific buyer. To explore this free tool and the rest of our technology, visit tuckadvisors.com.
[00:00:00] Alex Sarlin: Coursera and Udemy are merging in some ways it almost looks like a Coursera acquisition. They're coming together. This will happen late in 2026, but it's an all stock transaction valued at approximately $2.5 billion, and it's two giants of the online learning space coming together. They're gonna have over 175 million learners together.
And they mention a lot of interesting synergies that they are planning on putting together. These are companies, both have relatively new CEOs.
[00:00:30] Michael Horn: The really interesting news, I think a few months ago was Coursera starting to charge that fee on their university partners. And sure, none of the university partners loved it, but because the aggregator has become Coursera, not the universities, like right Coursera zoomed to the top because of the university brands.
Now Coursera is the aggregator. It is sort of like, well, sorry you don't like it, but we're still gonna charge it because they have the market power. And in some ways I see this as a consolidation of that.
[00:00:59] Dhawal Shah: They have $1.2 billion combined once they merge in cashflow and cashflow positive, I think they'll do a stock buy by program first.
But I think it's also difficult to do something as a public company. I think as a private company, if each of these individual companies are profitable, it's a good deal. But as a public company, they don't want the 2.5 billion to go to 1.5 billion. So they have to make moves. That has to be like, if they want to grow 10% year over year, just 10%, that's a hundred million dollars more a year.
And I think that's be going to be difficult as with the same product they have. But it's also going to be difficult no matter how many acquisitions you do. I think you use sort of kind of rolled the wave of acquisition trilogy get smarter and until it like hit headaches and then you know. I, I'm not sure like, yeah, I think Coursera might at some point, like I feel like there's another transition coming.
Maybe a bigger company eventually buys Coursera or Coursera goes, takes itself private. I think it'll something have to be seen. But I think they're like good solid businesses, just maybe not public companies.
[00:02:08] Ben Kornell: If you imagine like the United States does not have a national university globally, if you look at the growth that they've had in students from India, Brazil, like global south, if they get the cajones to do this, they could be the global university that people start choosing above universities.
Now, some of that's in their control and some of it, it matters what happens with our university sector, where we've seen like spiraling costs, lower value add, you know, their ROI decreasing. So you know, if Coursera goes with a cost cutting move, I think that's less likely. But if they go with a double down move.
It could be more likely. And just remember Andrew Ang being chairman. That's a huge signal. He's dialed in with all of the AI Illuminati, like 1.5 billion, 2.5 billion. This is like walk in the park, like chump change. You put 10 billion, 20 billion on this thing. You are talking transformation of higher ed.
[00:03:13] Alex Sarlin: Welcome to EdTech Insiders, the top podcast covering the education technology industry from funding rounds to impact to AI developments across early childhood K 12 higher ed and work. You'll find it all here at EdTech Insiders.
[00:03:29] 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, we have an awesome episode. To wrap up your 2025, we've got the one and only Michael Horn. A mentor, a visionary, somebody who's really pushed the disruptive thinking in higher education. Joining us here today to go and all the way through weekend EdTech. And then also we have some breaking news around Coursera and Udemy, and so we've got to Dhawal Shah here.
We'll do intros and all of that. But first, Alex, what's going on with the
[00:04:22] Alex Sarlin: pod? Well, we talked to Isabelle Hau from the Stanford Accelerator for Learning for this episode. Right at the end, she was just named for a World Education Prize. She has a book out, she's doing events, she's doing a challenge out of Stanford.
Always incredible things. And then as we end the year, we're publishing our amazing interviews with Rene Kizilcec from the National Tutoring Observatory and the Future of Learning Lab at Cornell, as well as Nolan Bushnell and Dr. Leah Hanes, founders of ExoDexa legendary game creators that are now moving into education and creating a virtual game.
We're also have recorded. 2020 predictions with people all over the ed tech field, including Jean-Claude Brizard from Digital Promise, Katie Kurtz from YouTube, and everybody in between. So keep an eye out because we're gonna have a whole bunch of amazing releases at the end of the year. Predictions about next year and coverage of this last year from people all throughout the EdTech sector.
I, I think we have 21 this year. I would, we've never had quite that many folks predicting, but they're fascinating predictions and lots of great
[00:05:25] Ben Kornell: themes. So speaking of great themes, we have a juicy topic to dive into. So welcome Michael. Good to see you. Good to be with you and welcome Dhawal. Let's dive in.
Maybe Alex, could you just set the stage, what's happening?
[00:05:40] Alex Sarlin: So, we woke up this morning to news. Uh, the, the ad Tech insiders WhatsApp community was all of us. We woke up to this morning to news that Coursera and Udemy are merging in some ways it almost looks like a Coursera acquisition. They are coming together.
This will happen late in 2026, but it's an all stock transaction valued at approximately $2.5 billion. And it's two giants of the online learning space coming together. They're gonna have over 175 million learners together, and they mention a lot of interesting synergies that they are. Planning on putting together.
These are companies, both have relatively new CEOs in place, which I think is really interesting. And of course, so Dhawal Shah, if those who don't follow him, Class Central is absolutely the core news outlet for anything MOOC related or online learning in that world. He's been doing it since 2013, has been followed this incredibly closely.
So I wanted to start by throwing it to you. You've followed Coursera at Udemy since the beginning, seen all sorts of changes, all their expansions, all their product lines, all their paywall moves. What is your take on this news today?
[00:06:47] Dhawal Shah: Yeah, it's, it's kind of a dejavu like when do you acquired edX? Yeah.
It's like, it's that level of shocking event in online education and yeah, it's very surprising. And I was looking at the numbers and I think they are trying to phrase it as an ai, like it'll help with ai, but it's mostly like a financial stock price sort of decision. Both these companies have struggled this year.
I think they do really well in revenues, but not well in stock price. I think they're both good businesses, profitable businesses. Udemy actually turned profitable. Coursera is still cashflow positive. Lots of money in the bank and in a way it's complimentary because two third of Coursera's revenue is consumer udemy's.
Two third revenue is enterprise and combined they like, like I was just checking without the graph and it's like 50 50% business and consumer and Coursera for last few years have tried to do, like I always thought Coursera's biggest opportunity was kind of building a Udemy themselves. And they tried in different ways over the years.
Like there was the Coursera. Instructor program or something. They acquired a company a while back and then in the last few years I've seen them like basically hosting content from other partners. Like they have skill shares on Coursera now simply on Coursera. So they had been basically turned into an aggregator where they are more courses from like basically non university courses on their platform.
The job, the new courses are from these kinds of partners because they can just get a grab catalog. But I still don't think they were able to close that U had. So I think that's where I think this acquisition sort of, which is, as you mentioned, Coursera acquired Udemy. Andrew Wing was going to be the head of the board and the CEO is going to be the Coursera, CEO.
So I think it's basically, and I think it'll take a while. I don't know what product integrations they will do. But I think in both companies are struggling and especially in business to business too. I was looking at their net NR and both have like decreasing over the last years. So I think very similar problems.
Growth is slow, better in consumer. Udemy has struggled in consumer for a long, long time. So I think it's a very similar trajectory. I looked at the employee size, they basically have like at the end of 2024, the difference was only like 50 employees. So effectively like the same company but different variation of it.
That's right. And huge marketing platforms, both of them. So they can cross market. And so I think financially all it makes sense, but I think they're trying to, same with when I think two, you acquired edX, there was also like a product narrative, but it was more of a marketing decision where two, you thought they could monetize edX users better.
So I think it's similar here.
[00:09:42] Ben Kornell: This to me feels fundamentally different than the edX to you, but I'm curious, Michael, how do you take this news and then how will this radiate out into the broader higher ed ecosystem do you think?
[00:09:54] Michael Horn: So like Dhawal, I surprised by it. Didn't see it coming and it feels significant, obviously, and I agree.
Ben with where you're going, which is while it's like earth shattering of the sort of like the two U edX deal, right? Back in the day, they feel fundamentally different, right? In many ways, edX or the acquisition of edX for two U was its attempt to create sort of an aggregator type model for itself.
Whereas here you have two companies vying to be an aggregator. Coursera. I think doing it much better than Udemy on that front, and you see that in the consumer business being much bigger than Udemy's, and this feels like, okay, let's bring more into the tent. In some ways it's more consolidation Now at the top.
These companies all born roughly 13 to 15 years ago or so, right? We've sort of seen this game play itself out maybe, and we sort of have our clear winner, if you will, out of this particular epoch, if you will, or epic of this particular phrase of the massive online course providers. I don't, I've never loved the phrase MOOC for a variety of reasons, but we sort of see our culmination and, and I think Coursera, you know, there was Udacity, there was edX, and Udemy was never quite the same genre, but now it's sort of like we actually have more choices in, with our aggregating power, we can spread that across and probably monetize even better across the platform and an even bigger range of offerings.
Last really quick thought on it, the really interesting news I think a few months ago was Coursera is starting to charge that fee on their university partners, and sure none of the university partners loved it, but because the aggregator has become Coursera. Not the universities, like right Coursera zoomed to the top because of the university brands.
Now, Coursera's, the aggregator, it is sort of like, well, sorry you don't like it, but we're still gonna charge it because they have the market power. And in some ways I see this as a consolidation of that.
[00:11:54] Ben Kornell: Yeah. It just goes to a fundamental truth in our space, that channel is power and customer acquisition cost is the whole story of yes, education success.
If you have high cac, you've gotta have a high LTV model. If you have low cac, you can have a lower LTV model, but the scale required for any low LTV model has to be so massive. And what's great about this is it allows Coursera to play at a number of price points across a number of, you know, if you imagine as a double-sided marketplace, now the inventory is much better and the consolidation.
Of the channel is strong. I think the question that to you was trying to figure out, they were almost like spending so much on marketing and then their own partners started to become competitors with them. That the flywheel has to start with the consumers coming to you. Yeah. And then you ramp supply. So demand first, then supply.
And I'm actually kind of excited for the workforce learning products to improve with Coursera's leadership. 'cause you could imagine. Essentially an employer being able to climb the ladder to actual credentialed courses that can lead to degrees in a maybe a clearer path than with Udemy. But is this the final bell?
Like is the match called and we've got the boxer in the winner, or where does this go? It does seem to me that part of the context of this is that the game may be over, but we're onto a different game where you've got the Googles of the world aggregating content, and this is almost as much as this is like a winning move.
It's also a defensive move against where we're headed with AI and learning.
[00:13:41] Michael Horn: I wanna hear what Dhawal has to say on this one because he tracks the sector more. But I'll just say quickly, from a disruptive innovation perspective, right, that you tend to have, when these new disruptive innovations come into the market, you have lots of them.
And then when the new disruptive innovation underneath them comes out, you tend to see consolidation of the old, right? Mm-hmm. Because you're sort of harvesting cash, if you will, from the old game while the new game is being figured out. And I mean, it's still very early on the new game in in, in my judgment, but you could read this as sort of a classic sign of, Hey, it's time to start consolidating the market.
There's not really space for all these different players in there. And while you noted that there's been struggles in the enterprise segment, certainly for Coursera relative to where you would've said five years ago, you might've expected them to be, I think, and I think Udemy as a stronger enterprise, at least in terms of revenue, a stronger enterprise business than Coursera.
So I'll, I'll be curious like. Does the customer acquisition, like is that helpful to Udemy's business? I'm not sure. I'm not actually sure. You'll see as much uplift in enterprise from this as, as one might imagine.
[00:14:46] Dhawal Shah: I think one of the things they said was they'll save like one $50 million in cost and I think part of it is just combining the sales teams and
[00:14:56] Michael Horn: agreed,
[00:14:57] Dhawal Shah: like shrinking it out.
And maybe, maybe there's like, I think I heard during one conference, sort of one earnings call that Coursera usually goes like, it's a higher price product when it comes to enterprise and that's why the penetration is less. So you can imagine like they have one T in enterprise now, the Remy tier, of course, third tier.
So I think there will be some benefits of, you know, selling higher and lower. But yeah, I think as it's like, you know, like I started out class until 2011, then 2012, Udacity. Coursera came out. Remy also came out similar times, and I think all the major players like Code Academy, all of that came out in that few years.
And since then, especially in the online education world, like I haven't seen something new come up that is at the same level of revenues or brand recognition. So I now wonder, does it even open up floodgate for Coursera? Like let's acquire skill share, let's acquire like, like instead of just like getting the catalog and let's get, you know, let's build this giant subscription model where like one price you, you basically like, it's like, you know, Netflix buying more browser.
But I dunno, is it similar to that? I wonder, I dunno what the future will tell, but I want for them.
[00:16:23] Michael Horn: Yeah. I'm curious what you guys think about this. If I can jump in with a question, but it's not my podcast, but I'm super curious, like, just based on what you just said there, Dewal. QU School of Business and Technology, for example, I think is like the coolest slickest M-B-I-M-B-A out there, completely mobile based, not a huge business in its own right.
Could something like that come into the Coursera fold? Based on what you just said, I'm curious what the three of you guys think about that.
[00:16:49] Alex Sarlin: Yeah, I have, I have, I have deep and personal thoughts about all this. I did the QU MBAI, I literally did it on the entire thing online within a year. It was fascinating experience.
I learned a huge amount. I also was an early Coursera employee. And you know, I think just to speak quickly, to your point before about Udemy being a different genre, and when you said dewal too about their complimentary sort of skillset, I think Yes, sure. Answer to your question, Michael. Yes. I think it would be very smart for Coursera to do some consolidation.
I mean, we're mentioning the two U edX purchase. The two U edX purchase did not go the way they expected it to go. It really basically bankrupted to you. I don't think they got the synergies they wanted to get. Totally agree. Right? Definitely. So we don't, we don't want that to happen again here, but I think that these two are looking at each other and saying this kind of a blue ocean right here.
Surprisingly, I mean, yes, there's all these newcomers in terms in terms of open AI and Google, and there's a whole new world of of ai, but in terms of sort of being the go-to online content platform that can sell into enterprise, that can sell to consumer, that can uplevel and do that ladder from a five minute video all the way up to a degree, there's not that many people left in that space.
I think your, your point, Michael, about them sort of being the last, last man standing from that era is definitely, definitely true, is also really interesting because Coursera, back in the day. When we were looking at Udemy, we were always jealous of Udemy in certain ways because even though Udemy's product is much cheaper, it's much more agile, you know, it's much more user generated content.
They were incredibly fast to do things that were very high need for learners, and it was a very complimentary catalog because it was Udemy would, as soon as a new tech product came out, I remember at the time it was Swift, right? There was like the Apple programming language. They would have 200 courses on Swift in that week, and then the best ones would rise to the top and they'd have all those great, you know, YouTube like catalog dynamics, Coursera, because of its university partners, eventually it's technical partners.
It was very constrained to long form. Very college-like courses that were pretty intense that could be transformative. And there was a lot of great things to say about them, but it was not very agile. Udemy also has 250,000 courses on the platform. I mean, again, back in the day of Coursera, we were just trying to build a catalog because at the original version of Coursera, there was no catalog.
It was people would come and do a course on Coursera and then it would end and they'd never do it again. And it was like, it was like an event platform. So building a catalog was the original sort of impetus for going to a on demand model. It was a lot of the goal of those early years was to make a Coursera catalog that was meaningful and exciting.
And meanwhile, Udemy was lapping Coursera over and over getting tens of thousands of courses, very cutting edge courses, very technical, very relevant. They've made a lot of big mistakes. And adu, I'd love to hear you talk about some of the mistakes that Udemy has made, because I think they've stumbled a lot in their journey.
So does Coursera, but I think Udemy especially has, their stock prices have reflected that, but I think Coursera is the last man standing there. I think they've already made a deal with OpenAI to sort of bake themselves into the OpenAI app platform as a very initial partner. I'm hoping that they're making these moves and trying to stay at the very top of the relevance scale as this AI world comes to fruition.
This sort of AI workforce, this upskilling need that is becoming sort of universally acknowledged. But I, I've been surprised at how low their stock prices stayed all over, you know, over these years it's just gone pretty much continuously down since the IPO during the pandemic, which is, you know, despite continually growing and being cashflow positive and all those things.
So that's my long form take on your question. They should consolidate. I always thought they should buy Data Camp. I think they could buy Skillshare. I don't know how well masterclass is doing, but man, wouldn't that be great? Coursera could just pull the masterclass catalog in. They could sweep the board here, and that would be a very interesting future.
They could also go bankrupt doing it.
[00:20:34] Dhawal Shah: They have $1.2 billion combined once they merge in cashflow and cashflow positive, I think they'll do a stock buyback program first. But I think it's also difficult to do something as a public company. I think as a private company, if each of these individual companies are profitable, it's a good deal.
But as a public company, they don't want the 2.5 billion to go to 1.5 billion. So they have to make moves. That has to be like, if they want to grow 10% year over year, just 10%, that's a hundred million dollars more a year. And I think that's going to be difficult as with the same product they have. But it's also going to be difficult no matter how many acquisitions you do.
I think you use sort of kind of rolled the wave of acquisition trilogy get smarter and until it like hit headaches and then you know. I, I'm not sure, like, yeah, I think Coursera might at some point, like I feel like there's another transition coming. Maybe a bigger company eventually buys Coursera or Coursera goes, takes itself private.
I think it'll something have to be seen, but I think they're like good, solid businesses, just maybe not public companies.
[00:21:43] Ben Kornell: I think the question of where this all goes, it just depends on the scope of ambition. I mean, I think you could see a private equity firm really liking cash flows here, and one way that AI could be helpful is just cutting costs.
So if you imagine like a refactoring of their cost structure, this becomes a much, much higher margin business from a delivery standpoint because you've got this like massive amount of data, content, everything. It should be right for like the AI revolution. So one bet could be that way on cost structure.
You know, I think there is a risk on the employer space of coursera's higher prices getting diluted down by Udemy price. That's what I was
[00:22:26] Michael Horn: wondering if they're, do you think they'll lose business? Yeah.
[00:22:28] Ben Kornell: Yeah. I mean they may gain in volume, but lose in average customer value.
[00:22:33] Alex Sarlin: Yeah.
[00:22:33] Ben Kornell: The other way to go if we're really home run hitting here is they become the new university.
And by the way, like that's not out of scope of the ambition of the origin story, but if you imagine like the United States does not have a national university globally, if you look at the growth that they've had in students from India, Brazil, like global south, if they get the cajones to do this, they could be the global university that people start choosing above universities.
Now, some of that's in their control and some of it, it matters what happens with our university sector, where we've seen like spiraling costs, lower value add, you know, their ROI decreasing. So you know, if Coursera goes with a cost cutting move, I think that's less likely. But if they go with a double down move, it could be more likely.
And just remember Andrew Ang being chairman, that's a huge signal. Yeah. He's dialed in with all of the AI Illuminati. Like 1.5 billion, 2.5 billion. This is like a walk in the park. Like chump change. You put 10 billion, 20 billion on this thing. You are talking transformation of higher ed. I, I'm, I'm curious what you guys think, 'cause I know this has been in the water for like 15 years, but it is actually like, that could be the play and the AI innovation by the way, you know, in terms of disruption, once you aggregate, there's a weird way in which defensibility in AI has not proven to be, there's no moats.
So actually channel is defensibility. Like aggregation is defensibility. So they might actually be primed for this next revolution rather than disrupted. I
[00:24:14] Alex Sarlin: mean, we've seen all these universities rush to do AI majors and, and put AI in their, you know, it's like the fastest growing major in, in the schools that are doing it, MIT and, and
[00:24:23] Ben Kornell: others.
And who would you bet on to have a better AI major than a mid-tier university or Coursera that you can get in your living room at a fraction of the cost?
[00:24:33] Alex Sarlin: I mean that literally had Jeff Hinton and Andrew and Daphne Kohler and all of these people on the platform from the very beginning.
[00:24:41] Michael Horn: I do think, I'm curious, you're, I mean the platform's obviously evolved significantly, Alex, from the days when Oh yeah.
Like I would stop by at lunches and back then and say like, ah, I think you're gonna have a little bit of challenge. 'cause the pedagogy of the, of your partner sort of sucks. But, um, I do wonder, like I, I don't know the rights, intellectual rights, IP rights on a lot of the courses and stuff like that, but it does feel to me then like you would've to do a dual transformation play, maybe where you spun up a new platform that could harvest all the courses and resources from the existing business to sort of create something that was much more interactive and handholding with the AI as part of it.
That would be a swing for the fences move. But I would think it would have to look, you know, you'd borrow the aggregator position, but actually have to separate yourself from a lot of the existing legacy technology would be my guess.
[00:25:33] Alex Sarlin: Yeah. I mean the good news there is that as they went to degrees, they did layer on a number of different customer success and coaching, you know?
Right. That's a good point.
[00:25:44] Michael Horn: Yeah. It's
[00:25:44] Alex Sarlin: the original pedagogy of just, which is the pedagogy and the support systems or, you know, the wraparound services around the original just courses was negligible. But you know, as they went to Degrees, which was a number of years ago now, they started to lean into that.
As they went into certificates, they started to lean somewhat more into that. They've outsourced some of it to nonprofits and others, but I agree completely that if they were to, I mean, you could in, in a world where Coursera put out, you know, an AI degree that was trying to be the world's standard and AI degree, maybe in with a partner like a Stanford or a Dartmouth or in anybody like that, they already have Anthropic on the platform.
They already have open AI on the platform and Google. All three providers already have courses on the Coursera platform. It's a pretty unique position to be in, to be able to combine the LLM Frontier Labs and the best universities in many, many countries and regions. That said, I don't know why they hadn't done it already, if that was the plan or, or if they have, why it hasn't made it down to our radar, but it's definitely really interesting potential there.
And I mean, Michael, you put out this incredible article recently about all of the financial struggles of so many different universities in the Northeast. I mean, yeah, we were in a very strange moment, to your point, Ben, about what, depending on what universities do, but I'd love to hear you talk about what is happening in the traditional university space that sort of potentially opens up the disruption in this way.
[00:27:01] Michael Horn: Yeah. Well, I mean, again, higher ed's always been funny, right? Because you have at the top universities, we'll call them businesses for a second, but whose value has been derived from whom they exclude, not the volume that they include, right? And so to the extent that there would be disruption, from my perspective, it always had to be in the middle tiers of the market, if you will, and sort of an absolute, and I think we're now seeing that possibility actually come to roost because to your point, Alex, like.
These business models. Look, I've been predicting for a little over 10 years that they were gonna drive off a cliff, and now we're really seeing evidence that that's accelerating. I think, and it's not just the ones that I expected, the small 1000 sub colleges and universities and the regional publics and sort of the for-profits with shaky value propositions.
Those are definitely in a world of hurt and are continuing. So it's stretching into the mid-tier though private tuition dependent schools that are not research based, by and large, and there's like 500 of those across the country. Our analysis went deep on New England, but we've been starting to go through west coast schools in Midwest.
I'll tell you the mid Midwest are like a little bit more responsibly managed on the cost side, so they may be a little less vulnerable. But I think the point holds, which is if you start to say customers have clearly become much more price sensitive in the higher ed segment, we see that there's a demographic cliff.
Every single university thinks that they're gonna grow into this market. Their growth, I think, was frankly gonna be on online master's degrees, targeted adult learners. That market is now done, I think, with the one big beautiful bill act like that's not free money anymore. And so I don't see how any of these institutions frankly grow.
They might be lucky to hold on to market share if they're lucky, but most of them are gonna lose enrollment, I think in the next 10, 15 years. And so a bunch of them are gonna consolidate. Like it's just we have something like 8 million excess seats as it is in higher ed. Over what the demand is like I think we have enough seats for like 23 million kids and we have about 15 million students enrolled.
Like that's not gonna work those numbers. Right. And so to your point, there's a pretty big runway you don't have to attack Harvard, which is having its own share of challenges to create a disruptive business, I guess is the point in this ecosystem. There's a lot of room for this and you know, look, I've been an, I forgot that you had gone to qu, but I've been an, you know, an advisor to them since relatively early on.
And I wasn't bullish initially on them going into the BA market. What they nailed was, for most MBAs that are largely undifferentiated, there's a huge opportunity to disrupt there.
[00:29:39] Alex Sarlin: Exactly. There's
[00:29:40] Michael Horn: a lot of other places you can drive a wedge through right now, I think is the point. And there's gonna be a lot of schools that go under and, and take advantage of individuals and families that are asking for return on investment and get the eye part of that equation down.
[00:29:54] Alex Sarlin: A hundred percent Dhawal. I wanna give you the last word on this story because it's just, you know, you, you've been following that narrative that Michael just brought up, that I, you know, that's gone up and down over the last few years about disruption in the higher ed space from the technology sector or from innovative models.
What do you think, if you zoom forward and say five years from now, let's say three years from now, and Coursera, OMI are merged, they've come together and they've made some really smart moves and have become sort of the winner, what would that look like?
[00:30:24] Dhawal Shah: I think it'll be more of the same. If you look at the last two, three years, I think they haven't done any strong product moves until sort of the AI came in.
I know you were talking about degrees, and Coursera is actually stepping away from degrees. They merged it into their consumer segment and even the students are increasing, their revenues haven't been increasing. So I think it will be more of the same. I haven't seen like a strong product DNA in these companies for a while.
I think they're more better at partnerships and marketing than at a product right now. Like in the early days, of course, you know, when you were there, you are launching new products, you know, you're trying new things, you're failing, but you're trying. And in the recent months, you know, like until you had something like the specializations and eventually professional certificates, right?
In recent months, recent years also, there hasn't been like, they have a single model and they're trying to fit everything into that model. So what I would like to see is maybe, you know, like the new platform or a new, like, I feel like we should get more from the $1 billion in cash they have and not just like more of the sea.
[00:31:31] Alex Sarlin: I hear you. So what would it look like? It would be the whole catalog would create this incredible training set. You'd walk in, they would have this amazing AI coach take you through the whole experience. Go, you go do a five minute video or a week long course or a weekend course ramp up to a degree. You could have an interdisciplinary degree, you could use the Udemy user-generated content to clarify the things that the professors from, you know, are not quite as good as explaining like, is that the vision that you know, if they get that product story together, is that the kind of thing we could expect to see?
[00:32:01] Dhawal Shah: I feel like it's like trying to buy a new computer or something. You have Ram, you have hard desk, you have, you know, so like having similar like mentorship, you know, is one of these how much mentorship community, how much, you know, like you have these kinds of like individual can like, oh, I can learn Python myself, just gimme a course.
I'll do it myself, but hey, I need mentorship. Yep. Or I mean like group community, I, I think you could figure out these other parameters. AI is one of them. Like you could sort of configure something that suits you or suits you in your journey right now. And that might change. Like I might need more help in Python, but less in something else.
Like accept, like I just need a little bit. So, so trying to have this product and trying to have this content that matches where, you know, you are in your journey in education and also being able to go from, I think the biggest thing, like self placed learning missing is like going from like beginner to even intermediate.
I think it goes from, you become like still an early beginner, but getting to intermediate advance is very tricky if you already know something, these online courses are more of an hobby and they don't really teach you anything and you have to go outside. Look at very advanced articles or listen to podcasts.
Like you have to piece together your own knowledge. And I think, I feel like either hopefully we'll do it or somehow combining it into a single platform, or I think that would be the,
[00:33:26] Alex Sarlin: I can do that too. AI could pull in, you know, the most recent research articles about the topic that you happen to be watching about right now, or the most recent news.
I like that roadmap, right? The ability to ramp up mentorship, the ability to ramp up community at will, depending on where you're in your journey. And bringing in cutting edge, modern resources for the intermediates to advance learners to make sure that they actually could get what's happening in the field and nothing gets dated.
I think that's pretty powerful. Combine it with all the translation they're already doing. Combine it with all the Udemy content, 250,000 courses. You know, I'm always optimistic, but there is a path for them from a product perspective. But you're not wrong. They have not done a whole lot innovation for a while since I think the guided learning guided projects.
[00:34:07] Michael Horn: Do you guys think it was a mistake for them to, like, on the one hand, integrating into the chat GPT. Sort of consumer inter like right interface was seen as a big move, right? Like they're one of the early apps there, but they sort of give up their aggregator position by doing that. Was that a
[00:34:21] Alex Sarlin: mistake? I think they thought of it as a new way into the Coursera world.
Whether that's working or not might be a little bit hard to tell. Yeah, I mean, it's a great question. I was talking to somebody this week and their, one of their predictions was that SaaS is in trouble. The entire SaaS model is in trouble because of the aggregator position that these frontier labs are starting to try to take.
That if you go into chat BT and ask it to do something and it pulls up Canva and does it for you from Canva, then why are you subscribing to Canva? Right? I mean, it's a pretty wacky ecosystem. I don't know. It's too early I think to tell. I think it hasn't made as much of a splash as I would've expected.
It depends on, I think how both companies, what you can do on Coursera that you couldn't do with Open AI's Coursera app, which I think is still quite a lot. And while how much open AI continues to build onto their app store and promote things and sort of try to actually pursue the Apple ish and Android app strategy versus just having it as one of 50 press releases this year, which is I think where we're at right now.
But it's a really good question. I time will tell. I'm not sure. Dhawal, are you okay closing out now? Yeah, I'm fine. Yeah. Okay. Thank you for having, thanks so much
[00:35:29] Ben Kornell: Dhawal.
[00:35:29] Alex Sarlin: So much. Thanks for joining us. Good to see you. Speaking of Frontier Labs, there are many different things happening across the ecosystem because we have the great Michael Horn here.
I definitely wanna get to the higher ed section quickly, but why don't we just do a little bit of around the world in the AI space. We saw after the code red announcement from Sam Altman last week, OpenAI launched GBT 5.2 as a way to sort of counter Google's Gemini 3G 5.2 is about math and science work.
It's increases their reasoning. It's obviously performing better on the benchmarks. What do you two make of the OpenAI announcement? We clearly have a, a Battle of the Titans here.
[00:36:08] Michael Horn: Well, I, my quick take is that I'm sort of in the Ben Thompson camp here, right? Which is, you have this funny position right where Google, the ultimate consumer company is like running with Gemini without like a clear business model on that at the moment and chat GPT is trying to position itself as the consumer but doesn't have an ad model yet on it.
And still needs to be a little bit of sorting out I think on on in terms consistency of business models to make this really work. But I would be worried if I were open AI as well. I think Google's in a stronger place than I predicted two years ago when I thought this would disrupt their search business completely.
[00:36:39] Ben Kornell: Yeah, I mean I think that Code Red might be an apt statement for all of the people in this space, just because there's been such a rush to release new features and capabilities. And I think we often this year have talked about AI as the subject, not the verb or not the object of the sentence. And what they're starting to realize is the use cases really matter.
And you know, Michael, you're a huge scholar of jobs to be done and the work to be done. And I think what we found is a technology that's a hammer and it's been looking for nails. And now I think people are wrapping their heads around the use cases. And this is where I think Google is ahead because can I discern what's better about 5.2 versus 5.1 or five?
Sure, I can if I, if that's my objective. But other than that, I care more about is it a convenient enablement where I am with what I'm doing. And I think the other thing is AgTech AI has, while I think there's a lot of excitement about the potential in the buzz, I don't feel like we've gotten, because you can't trust each agent, each node with full accuracy or full reliability as you string together agents, the predictability and accuracy of your outcome.
Declines by a larger factor of the amplitude.
[00:38:08] Alex Sarlin: Yeah.
[00:38:09] Ben Kornell: And so Mirati is working on some of this work to replicate AI queries where it's consistent over time. You know, if you think about Six Sigma processes where they were on the Toyota Simply Line, and they're just weeding out the variance. We need a Six Sigma for ai, but everybody realizes that they've got to consolidate wins.
And this is where also the Disney branding and just like we just talked about Coursera, like where does content fit into this ecosystem? Right? I mean, Michael, when you were an advisor to my company 10 years ago. We talked about how the value of content is approaching zero because of open education resources and so on.
I couldn't have been more wrong at that time. Yeah, it's
[00:38:55] Alex Sarlin: totally wrong. Yeah. Yeah, we didn't know it and
[00:38:57] Ben Kornell: so it, it may turn out that content is hugely valuable here. 'cause my excitement about playing with an AI image model goes up dramatically if I've got Disney characters in it and Marvel and Star Wars.
Yeah, exactly. So maybe we're in the third inning of the game here, but it is a moment where we're seeing a horizontal spread of AI rather than a vertical one. And I think that will be really interesting to follow in the year to come.
[00:39:25] Michael Horn: Yeah, it's a really good point. I also think stories that we connect with still really, really matter and.
The reason the content of Disney's Library is so valuable, right? Is these stories that have decades and decades of history that we can connect to in different ways. And I don't know about you guys, some like the, some of the different VR ed tech stuff that I've seen over the years. The ones where I'm like, oh, that was cool.
Have nothing to do with the VR and everything to do with, wow, that was a really cool narrative to learn X through, right? And totally the VR was like almost a distraction from my perspective.
[00:39:59] Alex Sarlin: Totally. I think character and narrative and story and content writ large, which has sort of combined some of those things is a huge part of this story.
I mean, to your question, Michael, about Coursera and, and OpenAI, I think one of the really wacky things that's happening right now is that these AI tools have have appeared on the scene, and they are both platforms and underlying technology that can be put into anyone else's platform. And what you're seeing this very strange back and forth where you know Disney, I mean that shocked me when I saw the Disney thing.
'cause I had written the whole article about exactly this. There are companies with lots of great IP are going to start licensing them to doing AI versions of them. But I didn't expect it to be on open AI's terms. I thought it would be Disney saying, you know what we're gonna do virtual Mar Marvel characters and you can go talk to Spider-Man and Black Panther and OpenAI will be happening under the hood, but it'll be under the Disney brand.
Of course, on the Disney site, of course, in the Disney toys, of course. And that's not what they did at all. They're saying, Hey, go into Sora and you can make Spider-Man fighting with the Hulk on the moon. Like what does that do for Disney? I mean, it's a billion dollars, right? There's some other things going on, but.
I don't think it's a great move for Disney. I think it's a, a really interesting and exciting move for the space, for all the reasons you're saying, but I don't think it's the right move for Disney. I think they should have said, we're the platform, we're gonna make a deal with OpenAI, where OpenAI fuels our stuff, which is part of the deal.
But giving away your, your whole library of characters is like, it's kind your whole thing of your Disney. So I was shocked, but I'm also excited for EdTech. I think, you know, the idea of being able to do Spider-Man teaches, you know, teaches what, what's entomology? Who's your teacher today? Kid. Spider-Man.
Exactly.
[00:41:41] Michael Horn: Yeah. But I guess I have a slightly different take, which is I think maybe they learned the lesson of like say Sony pictures, right? Which is, you could spend billions and billions of dollars trying to build a streaming business that is still losing money. Or just recognize that, hey, Netflix and YouTube.
They're the aggregators. So like rather than try to compete on their terms, take your strategic point of differentiation and monetize it across all the platforms. And that's sort of how I saw the Disney news now surprising because they've been one of the ones spending billions and billions on a streaming business that has been a money loser.
Right. But I was wondering, hey actually maybe they played it exactly right here.
[00:42:20] Ben Kornell: Yeah. I mean, where else will it go? You know, Michael, this is actually an interesting question for universities then that are sitting out here where they have a very, very strong ability to deliver on one channel, which is in-person country club style, like physical brick and mortar, university, full stack education.
And it's full stack. Not only in that they own the buildings and the teachers and the curriculum and the content and so on, but they also own the places where people sleep and where they eat. And full on. But in a digital delivery model. Only a few have shown that they've been able to be nimble and let's look at Arizona State, probably the best example.
They still get so much heat from faculty and alumni and all of these folks, even though they've been by any measure, like incredibly successful at reaching more students. Unlocking economic value, unlocking educational value. So if I'm Harvard and I'm sitting there looking at that, or maybe that's a bad example.
'cause sometimes when we use an extremist, but if I'm maybe Indiana University's Kelley School of Business and I'm seeing this deal, the question is do I build a platform or do I build, do I continue to invest in my own or do I go to aggregators like Coursera or to an OpenAI and do I do, there's a first mover advantage to be the first one to do the business school deal because pretty soon it is all commoditized.
[00:43:52] Alex Sarlin: Yeah.
[00:43:52] Ben Kornell: And you know, you basically get this revenue stream that is more of a low tier or mid-tier revenue stream for your ip, and then you can still hold this premium experience, which is the full encapsulated, you know, university experience.
[00:44:08] Michael Horn: Yeah, and I think my quick take on it is I think most of the value of these institutions is actually not in the content, right?
And the teaching, it's the social network and social capital and et cetera, et cetera, that comes out of it. And that is intensely regional still, right? Despite the virtual work, there's still huge advantages, uh, in the moment in the job market to regionalization. And so the value of those institutions comes from their place in the region.
And so I guess. I could argue on either side. I'm not sure how valuable that content is if they brought it to an aggregator. The flip side, I think right Ben would be, well, you might as well 'cause it actually lifts your brand within your region and allows you to actually become even stronger. So you might as well, like if that's the commodity, leverage that to drive value to where you're really able to win.
That's right.
[00:44:53] Ben Kornell: Like how many accounting one oh ones do we really need? We don't need that many. But the credentialing of it, the brand signaling of it is differential value. And I think there could be a race on who's gonna be the first ones to go ahead and get that stamp in. And if you're thinking of a Google as an aggregator or Coursera, you have this interesting question, do I try to be friends with all and have a panorama of brands or do I try to make a few bets?
And if you have a panorama of brands, you could imagine it like. All the big 10 schools, and I mean, I'm thinking football conferences, you know, all the SEC schools. Here's an SEC degree, which, you know, choose your logo that you want on it. It's the Georgia logo, it's the Alabama logo. I mean, all of these things, the disaggregation potential is there.
I'm not
[00:45:42] Michael Horn: sure how valuable that is in its own Right. Right. And so the question is what would the aggregators be getting out of it, I think is the question on, on, on that one, right. That I think we'd have to work our way through a little bit. 'cause I, I do think, again, the connection, the, like, what I think works about Quant is not just that the mobile interface is really cool, but that they built, you know, regional teams get together and solve a problem together and do all these projects together.
That creates some sort of value. Not to say that these aggregators couldn't do that. I think the interesting thing on Google. I'm biased on this. A couple of students in my class wrote their paper on Google's career certificates and, and the partnership with Coursera. And their big argument was that Google for cultural internal reasons, actually has made somewhat of a mistake trying to make them portable into higher ed or stackable into higher ed and should have just run the table basically with a separate unit that goes direct to individuals and it makes a more concerted effort to make those certificates valuable in the marketplace in their own right and sort of go up market, if you will, over that.
And leveraging the fact that you said Ben, right, that they can lower cost of acquisition much more than a lot of these other online platforms can. And so if I'm a university against that background, do I want to feed that or do I wanna avoid it and maybe just double down on my own fiefdom, if you will.
Um, I, I have to think that one through. I'm curious your takes though.
[00:47:07] Alex Sarlin: I mean, I would just say that I think one of the, the differential values of universities, as you say, it's regional, it's, it's the social network, but it's also just experience. I mean, it's experiential learning. I mean, universities as even community colleges are a encompassing, you know, they sort of take over your life in the time that you're in them.
You have experiences, you meet new people. You may live away from home, A lot of people don't, but you might. And I feel like that's a piece of it that is really undervalued as well as the social network. But just that idea of like, as you go to digital learning, as you go to disaggregated learning, there is a lot of potential for disruption.
But IJI, I just still think we don't always, especially because this multi-generational, right, parents want their kids to have a college experience. I think they still do, even in this era of enormous doubts about the ROI and they don't wanna pay a lot for it. But I think there's still a real interest in, hey you, you meet people, you date people, you join teams, you join clubs, you have buildings you go to, you actually like, I feel like that always gets lost in the discussion.
So I just wanted to introduce that back in as another differential advantage of regional schools, of almost any in-person college. I wanted to bring one more headline into the four, just because I think it's relevant here. I hear, I think, Ben, you wanna jump in, but Purdue made news this week by basically announcing an AI competency requirement for undergraduates.
They now have to have the AI competent to graduate, and that I find is an interesting attempt at differentiation as well. I'm curious what either of you think of that, but Ben, I think you're jumping in.
[00:48:36] Ben Kornell: I mean, I would want to see what those are. I haven't had time to dig into it, but this idea that AI is the new literacy or that there's some sort of element of technological baseline that everyone needs in order to do any function, just tells me that they're recognizing that we're going to live in an AI everywhere future, where whether I'm creating or consuming AI is going to be part of that experience.
And what I love about this is, this goes back to why higher ed and informed citizenry. That's one argument. One is to get jobs and economic mobility, but in informed citizenry we need people who are AI literate. And my read on this one, it was not that you've gotta become technically proficient at like building an LLM, but it really is about like a core skill that will be invaluable to people as a citizen and as a human being and professional.
Michael, what, what are your thoughts?
[00:49:32] Michael Horn: So agree. First that I wanna look into it more, but my thoughts are Purdue is reading the market spot on, right? Which is that parents are freaking out about entry level jobs role of ai. Like they are really, there's a lot of anxiety right now and Purdue is leaning into it and saying like, we'll take that one off the table for you, right?
Like totally everyone coming outta here, they're gonna know what they need to know about AI to be okay. And whereas I think most colleges and universities are having a conversation of like. Cheating, et cetera, et cetera, like how, you know, blue Books, et cetera, et cetera. They're saying like, no, we're gonna meet you where you are, where your worry is.
And don't worry, we've got the students just fine. They're gonna leave with an AI competency and be ready for the world.
[00:50:14] Alex Sarlin: And the way it's the, just to fill in an interesting detail, you know, it's an AI competency that's being done in the context of individual programs. So it's not that everybody has to take, you know, a course on basic ai.
They're doing smart. Which is smart, right? Yes. Yeah. It's smart. There will be projects in their major, in their classes that have AI in them, so that they come out with some AI skills. I think that's, yeah, we
[00:50:33] Michael Horn: don't need, we don't need keyboarding.
[00:50:35] Ben Kornell: Again, I think this is, this is exactly like we talked about internet companies and non-internet companies in the nineties we're talking about AI versus non-AI.
Today it's just all gonna be blended to everything will have it, and it won't be as distinct. Purdue's reading the market, like Michael says, as a Hoosier, myself, I'm thrilled for our football team, but I wish the university would take some calls from Purdue University 'cause they're really crushing it.
The boilers are,
[00:51:06] Michael Horn: we're
[00:51:06] Ben Kornell: gonna how to get that one in. There're
[00:51:08] Michael Horn: we're gonna blast. We're gonna blast that one out
[00:51:10] Ben Kornell: afterwards. Yeah,
[00:51:10] Michael Horn: yeah,
[00:51:11] Ben Kornell: yeah. Well, it is time for us to welcome our amazing guest, Isabelle Hau. I can't say enough what a milestone it feels to have you here, Michael. You've really created almost the education media space where these kind of intellectual conversations can happen and people can not only think and dream about the future of education, but also understand and critique where we are today.
So thank you so much for all you've done for us in the space. Thanks for joining us today, and thank you EdTech Insider listeners for an amazing 2025. If it happens in EdTech, you'll hear about it here on EdTech Insiders.
[00:51:50] Alex Sarlin: For this week in EdTech, we have an absolutely amazing guest. Isabelle Hau is the executive director of the Stanford Accelerator for Learning, where she leads efforts to create innovative inclusive education solutions.
She's really at the absolute center of the AI and education world. She's the author of Love to Learn, which emphasizes the power of care and connection in early education. And just now, Isabelle Hau was announced as a finalist for the 2025 World Education Medals, founded by hp, which recognizes leaders who are driving change through AI in education.
Isabelle Hau, welcome to EdTech Insiders.
[00:52:29] Isabelle Hau: Thank you,
[00:52:29] Alex Sarlin: Alex,
[00:52:30] Isabelle Hau: for having me. A
[00:52:30] Alex Sarlin: pleasure. I'm so happy to be here with you today. You have so many different things going on out of the Stanford Accelerator for Learning, but the one I wanna start with is brand new. It just came out last week, which is that you are named as a finalist for the 2025 World Education Medal for Leaders funded by hp.
What does this recognition mean for you and for your mission at the Stanford Accelerator for Learning?
[00:52:54] Isabelle Hau: Yeah, thank you, Alex. And I'm so honored and humbled by this recognition. It's incredibly meaningful for me. It's almost like an affirmation of a belief that I've had for long, long time, but has guided my entire career that this future of education, future of learning, needs to really be at this intersection of technology advancements and humanity.
And for the Stanford Accelerator for Learning. I believe that this recognition reflects, you know, this collective work that we do to connect research, practice, and then impact and centers those learners who are generally overlooked, especially those with learning differences. And so to me, this award is a signal that the world is ready for that conversation.
[00:53:46] Alex Sarlin: A hundred percent what's so, I think so unique about the Stanford Accelerator for Learning is that you're right there at Stanford, it's absolutely, you know, the heart of Silicon Valley, it's the heart of where the business innovation is happening. You're down the road from a lot of these, these big ed tech companies, and at the same time, you're so research driven.
It's obviously, you know, a world leader university and you in particular because of your leadership, bring this humanity really to the forefront. So it's sort of the research meets the real humanity embedded into ai. Meets the technology. It's not like you're, you know, across the world saying, Hey, hold on, we gotta make this more human.
Let's not go so fast. You're absolutely at the center of, of the speed and yet also doing research to make sure that we really get it right. And I think it's really a unique role to play. And you know, these medals are spotlighting leaders as only five finalists in the entire world. You're one of two in the US that are designed to use AI to close learning gaps.
And you just mentioned, you know, learners who are overlooked. Let's talk about that a little bit. You know, how is AI currently being used to help us understand and support learners with special needs or learning differences or neurodivergence? I feel like this is such a important issue right now, and AI really could help if we get it right.
[00:55:00] Isabelle Hau: Absolutely. So AI is a technology that. Suited to address this question of learning differences and what we call learning differences. At Stanford, by the way, is learners with disability, whether visible or invisible disability. So that includes neuro, that includes physical differences or other forms of learning differences.
So it's a pretty broad definition, and AI is a technology that's particularly well suited for differentiation. So that really gives me a lot of hope about how can we serve learners with visible and invisible disability to thrive and learn in those systems of education. That for the most part, were not built for them at the center.
[00:55:57] Alex Sarlin: That differentiation is so core to what the potential of AI is. I mean, we've been talking to founders, we've been talking to investors for years now about ai, and it feels like the absolute through line of what everybody is doing is differentiation. So much so that when we made our market map, we actually went out of our way to not have differentiation as a use case because it is central to every use case.
No matter how you're using ai, differentiation and personalization are sort of baked in. And I, I couldn't agree more. It's a really important moment, I think, and we could get it right or we could get it wrong. I mean, let's be real here. It's not a done deal that, uh, AI is just gonna automatically serve these students, you know, the way we want them to be served, the way they need to be served.
But we have a potential here. One aspect of your research that I find very interesting is this, you know, and again, this idea of centering humanity and also social intelligence collaboration, like not having AI isolate us more, but having AI actually. Allow us to build what you call relational intelligence, understanding each other, understanding how to operate in the world, not getting even more sort of absorbed into technology.
Tell us about what you mean by relational intelligence and why it matters now more than ever.
[00:57:11] Isabelle Hau: Yeah, and I would say for a long, long time we have focused on human intelligence as primarily cognitive intelligence. So what many of us think about iq, which is a, a very, very cognitive way of thinking about our very own human intelligence, and then increasingly there's been a rise of emotional intelligence, eq, which a lot of workplaces and increasingly schools have been adopting over the past two, three decades.
My belief is that with AI rising, what will make us uniquely human is this capacity to relate to each other. So this what I called relational intelligence. So I gave it a little name like IQ and eq. What if we had AQ so that we can define this next evolution of human intelligence that I believe will continue, will need to rise in this AI world.
[00:58:10] Alex Sarlin: So le let's dig a little deeper into this because I, I wanna understand this concept even better and I'm sure our listeners do as well. When you talk about relational intelligence, especially with regards to ai, how are you envisioning AI being able to enhance people's relational intelligence, or at least not decrease it and make us feel even more unable to connect with one another?
[00:58:32] Isabelle Hau: So that's a really big question, Alex, and one that I would say has also, you know, raises concerns on my end for sure. Where most technology, and certainly the first inception of AI as we know it today, is mostly through chat bots. And most chat bots tend to be one human to a machine. So essentially, those chatbots run the risk of isolating us further by having more and more of our children and adults and time in front of those machines.
However, this technology could be used in very different ways, and with a lot more intentionality, we could have AI really connect further and help us with this concept of relational intelligence. For example, we could think about, and we are working on a big project at Stanford on this particular topic.
We could think about AI agents in classrooms that support collaboration and peer to peer interactions or small group instruction. AI is really, really well suited to do this. Or we have a beautiful project in our early childhood team at Stanford called Find that's a video coaching tool for human interactions that helps with best practices on how to interact between a parent and a young child or a caregiver and a young child.
So very, very focused on human interactions. This tool is now increasingly powered by AI with some human in the loop, uh, to ensure that the recommendations that are provided to the humans on human connections are appropriate and are based in class from a research perspective. But this is an example where AI doesn't replace human connections, actually fosters and enhances human connections.
[01:00:29] Alex Sarlin: That's exciting to hear that there are some projects really focused on that. We, we interviewed Dr. Soen. Ro another amazing standard.
[01:00:37] Isabelle Hau: Yeah, no Alex. So Soren's project will be another beautiful example of the use of AI to connect humans. Thank you Alex, for giving that example. A
[01:00:48] Alex Sarlin: hundred percent with peer Teach uses AI to match peers and to get them to be able to tutor each other in really meaningful ways and track their progress.
It's really, really exciting work and all, all of the work coming outta Stanford is exciting. I also feel like it's very groundbreaking in that many people's paradigm for ai, as you say, is chatbots our AI companions or AI assistance or you know, there's all these different models that people have been come, or even AI agents that tend to be very isolating.
They're sort of mechanisms by which an individual might use AI to be more productive or to get questions answered or to get a tutor, which is a great thing, but it's still one-on-one or sort of isolated is one person alone. Some of the projects you're saying that, you know, the early childhood peer teach are, are just this totally different paradigm that I think my prediction I, this was my major prediction for next year, I think next year is gonna be the year of social ai.
I think it's gonna be the year we sort of collectively realize some of the things you just said, that it can foster and enhance human connection, not just be, you know, another device that we use or another interface that we use individually.
[01:01:53] Isabelle Hau: I love s Alex, may I add? Yes. 2026 should be the year of social ai.
Thank you.
[01:02:00] Alex Sarlin: Yeah, let's, let's, let's make it so, I mean, there's been these bright spots of people doing this. OCO Labs does amazing work. Breakout learning does does, yeah. The higher education peer teach human to human AI is a, is a company. There's a few companies that have like really focused on this.
Swivel is an interesting company that does AI bots in classrooms, that foster group work. But I can name them all in, in, in a few minutes and that's not what we want. We want there to be an explosion of people starting to think about, Hey, AI can be something really different. It's really exciting.
[01:02:30] Isabelle Hau: And I would, I would add also we want not only tools and applications that are pro-social, ideally we would also have all.
Large language model organizations focus on driving models for social behaviors. And that's actually a discussion that I'm leading with a number of those leading and cutting edge model companies. 'cause it should be not only the application layer, but also at the source of those models in the way they are behaving And fine tuned.
[01:03:04] Alex Sarlin: Hundred percent. I think OpenAI sort of quietly launched group chat a few months ago as a sort of experimental model or experimental interface. And I've been using it with my friends. I've actually been doing, you know, weekly group chats with OpenAI. We've been playing games, we've been doing conversations, and we have chate in the room with us and it synthesizes or it comes up or it hosts or it facilitates.
And I have found it incredibly rewarding. And it's funny, it's like, I think it's one of the least covered innovations in the LLM space. Everybody, 5.2, 3.0, new models, they can agents. But the social innovation is a big deal and I'm really excited to see Google responds to this as well, as well as philanthropic.
This is another example of why your work at Stanford, being right at the center of the tech universe is so amazing that you actually have direct influence and access over these LLMs. So I can't wait to see what, what happens next year in that mold. And then of course, people can build on top of it, right?
If we have a group mechanism or a facilitative interface within these LLMs, then EdTech companies can use that and build around it and build on top of it, and build use cases for early childhood or build use cases for tutoring, peer tutoring, or for teacher student interaction. It's really exciting and I mean, and I feel like this year we're gonna see innovations.
I'm really, really pumped about them. I wanna give you space to talk about some other incredible things coming outta the Stanford Accelerator for Learning. You just launched this Create AI challenge, create plus AI challenge with support from google.org, which is about this contest to start to get ideas from all sorts of people about how AI can augment human potential.
Tell us about what this means.
[01:04:43] Isabelle Hau: Yeah, so we are striving to have solutions that are augmenting human potential in terms in instead of replacing humans per, you know, per discussion. So we would like to see innovations in free areas continu rising. One is augmenting learning, and especially in this area of social learnings, which we just discussed.
Another area is augmenting teaching, and also similarly having classrooms that are collaborative but also focused on children that are very inclusive of all children. And then third area where we would love to see also better solutions is augmenting carrier pathways, especially in the area of early careers, which we know is being impacted.
So how do we think about also better solutions there? So those are the three areas, and we are soliciting applications. The deadline is generally, if, so it's coming up right after the holiday break, and we are welcoming early ideas. From a wide range of people, hopefully, and, uh, excited about seeing what will come out from that challenge.
[01:05:58] Alex Sarlin: Yes. So if you're doing any of those three areas, and these are grants of up to $50,000, I believe, to help support making these ideas, accelerating these ideas, making them a reality. Is that, is that right?
[01:06:11] Isabelle Hau: Exactly. So we'll have six prizes of 50,000 each, and then we'll have multiple smaller prices. And all of us is in the form of non dilutive coin capital.
So very beneficial to a wide range of players, whether they come from the for-profit world, nonprofit research, and more.
[01:06:33] Alex Sarlin: Amazing. And then you also do this amazing symposium, this event in February that I think is also very central to the AI and learning world. Tell us about your Stanford event that the Stanford Accelerator for Learning is putting on this February.
[01:06:48] Isabelle Hau: So it'll be our fourth Stanford AI and Education Summit. It's an event that we co-lead at the Stanford Accelerator for Learning with our friends at Stanford, HAI, which is, which is the Human Centered Artificial Intelligence Institute, and we are going to bring a few hundred people on campus for this event and then also broadcast for videos, this event for even greater participation.
This event this year, we have decided that the theme of the event this year would be on what we call an inflection point. On how AI is changing. Not only how we learn, but also what we learn and even maybe why we learn. So it's really, really framing some big questions and hopefully also highlighting and outlining some of the great work from many researchers across many disciplines within the Stanford ecosystem and beyond.
[01:07:51] Alex Sarlin: I love that, that not only how we learn, but what we learn and even why we learn. I totally agree. There's been some studies coming out recently that, you know, generation Z says that they're learning a new thing with AI every month, and they're doing it totally voluntarily. They're using it and some of the big models have said it's the number one use case that they see.
Philanthropic and open AI have both said that as they analyze tons of conversations, what people are using AI for is, is to learn and to build skills and to sort of make sense of the world around them. I think that's a change in, in why we learn. It's a tool that it's very different than traditional education formal and it's very, very relevant.
It's sort of by definition people are trying to learn exactly what they think they need to know at that moment. It's a really interesting moment. This has been so much fun. I, I'd love to have you on for a longer time. We could talk about the book. We haven't even mentioned it here, but you have so many amazing things going on in the AI and learning space.
It's just a absolute pleasure. Isabelle Hau is the executive director of the Stanford Accelerator for Learning, where she leads efforts to create innovative, inclusive education solutions. We didn't even talk about tutoring. Another amazing focus of the Stanford Learning Team. It's been a real pleasure.
Let's talk again soon, Isabelle.
[01:09:05] Isabelle Hau: Thank you, Alex, for having me. A pleasure.
[01:09:08] Alex Sarlin: Thanks 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.
