Edtech Insiders
Edtech Insiders
Week in Edtech 2/12/2025: DeepSeek AI Shake-Up, OpenAI’s Deep Research, CSU’s ChatGPT Rollout, NAEP Learning Loss, Trump’s Ed Research Cuts, FEV Tutor Shutdown, European EdTech Funding Trends, and More!
This Week in EdTech, Ben Kornell and special guest co-host Libby Hills from the Jacobs Foundation and the Ed-Technical podcast break down a week filled with major AI breakthroughs, troubling education policy shifts, and key trends shaping the future of edtech.
✨ Episode Highlights:
[00:04:53] 🔍 OpenAI launches Deep Research, a $200/month AI research tool.
[00:12:06] 🐋 DeepSeek AI challenges US models, raising concerns about China’s progress.
[00:17:24] 🏫 CSU offers 500K students free ChatGPT Pro access.
[00:24:29] 📉 NAEP data shows continued COVID learning loss, with state disparities.
[00:29:23] ⚖️ Trump administration cuts federal education research contracts.
[00:37:19] 💸 FEV Tutor shuts down due to ESSER funding cliff.
[00:41:56] 💡 BrightEye Ventures’ report highlights European edtech funding trends.
😎 Stay updated with Edtech Insiders!
- Follow our Podcast on:
- Sign up for the Edtech Insiders newsletter.
- Follow Edtech Insiders on LinkedIn!
🎉 Presenting Sponsor:
This season of Edtech Insiders is once again brought to you by Tuck Advisors, the M&A firm for EdTech companies. Run by serial entrepreneurs with over 25 years of experience founding, investing in, and selling companies, Tuck believes you deserve M&A advisors who work as hard as you do.
[00:00:00] Libby Hills: I mean, I think making research more accessible to developers is a really important step, but it's not the only thing that needs to happen in order to make products more evidence based and more likely to be efficacious. So, yeah, a big thumbs up to anything that contributes to helping people better understand.
What the research is saying and what that might mean from a product design perspective, but that needs to be coupled with, okay, how, how are we actually going to integrate this into our product roadmap and build around this in a meaningful way on an ongoing basis?
[00:00:30] Ben Kornell: I think the real loser here in the DeepSeek race is Meta, who it's like, whoa, you got outperformed.
By someone spending a fraction of the money and they're doing it in an open source way too. So they're even kind of like circling around your entire strategy. That's got to be a shocker. And I'm sure there were some folks in the Facebook towers that were. Very upset with that.
[00:01:00] Alex Sarlin: Welcome to ed tech insiders, the top podcast covering the education technology industry. From funding rounds to impact AI developments across early childhood, K 12, higher ed and work. You'll find it all here at ed tech insiders.
[00:01:16] Ben Kornell: Remember to subscribe to the pod, check out our newsletter and also our event calendar and to go deeper, check out ed tech insiders.
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.
Welcome back EdTech Insider listeners. We have a special co host today. I am so excited to have Libby Hills here by day. She is. Leading the EdTech work at the Jacobs Foundation, and by night, she is a fellow podcast host of EdTechnical, covering everything that has to do with AI, education, and impact.
Welcome to the pod, Libby Hills.
[00:02:04] Libby Hills: Thanks, Ben. It's great to be here again. We had a lot of fun last time I was on at Tech Insiders, so yeah, excited to be here today and hopefully can bring a little bit of a European flavor to today's episode.
[00:02:15] Ben Kornell: Yeah, we're really excited to have you here and for those of you who want to listen, where can they find Ed Technical Podcasts?
[00:02:22] Libby Hills: Yeah. Thanks, Ben. You're always so good at helping us to promote the podcast better than we are ourselves. So yeah, at Technical, we're on all major podcast channels and also on LinkedIn. So do check us out and if you like what you hear, please do share, subscribe and help us share some of our content, which is all around speaking to experts in this space and hearing what they have to say about what's going on in AI at the moment.
[00:02:44] Ben Kornell: Speaking of podcasts, this week we have a bunch going on on our podcast. Tigran Sloyan, the CEO of CodeSignal. I love Tigran. He was actually an investor in one of my previous companies. They're basically building the pipeline for talent for coding, but also creating this incredible HR platform where people who are employing or searching for new technical leads can better assess.
The skills and abilities of potential employees. We've got Matt canard, CEO of better lesson. They just acquired Abel, which is Adam Pizzoni startup and better lesson has been one of the OGs in the ed tech space. Matt is a source of incredible wisdom. Then we've got Jeff Magian Calda, CEO emeritus of Coursera.
So literally the week that we were coming out, he announced his retirement. And he's just dropping wisdom all over the place in that pod. I encourage you to listen to it. And then coming up next week, we have Adam Franklin with StudyBuds and we have Pierre Dubac from Open Classrooms. It's going to be a very, very big week on the pod in terms of events.
We also have the Bay Area EdTech Summit. That's actually today. So if you're listening to this, you've probably already missed it, but we have about 150 folks coming to the Cooley offices in San Francisco. This is our third year of the Bay Area EdTech Summit. And if you missed it this year, make sure to join EdTech Insiders Plus membership so that you can be included in all future events.
All right. So much to talk about Libby and it's really unclear where to start because. I feel like the weeks are feeling more like years right now.
[00:04:31] Libby Hills: Yeah. Yeah. Ben, when I agreed to come on and co host, I was expecting we might, you know, be talking about a couple of cool new AI tools, maybe some interesting M& A deals, but wowzers.
Yeah. What a week for EdTech.
[00:04:44] Ben Kornell: Well, let's start with our around the world in AI and technology. What's catching your attention this week in AI?
[00:04:53] Libby Hills: So, one for me is the launch of my new best friend, Deep Research from OpenAI. So, Deep Research is a new AI agent from OpenAI that combines one of their latest reasoning models, 0.
3, with agentic properties. And Deep Research is this amazing tool to really help facilitate deep research, unsurprisingly. And the autonomous aspects mean that The agent will go off and will, you know, search and synthesize resources from the internet, which you can actually see happening in real time, which is fascinating to watch.
And then we'll produce a product that's been said to be kind of akin to the quality of PhD student piece of research. And so I think it's really interesting. I mean, professionally. It's an incredible work tool. It does come at a price. It's for their, you know, highest prescription level. So 200 bucks a month.
But, you know, for me thinking about implications for education, it's, you know, I think just furthers some of the concerns that we have around, you know, assessment, for example, making it even harder for educators to be able to discern what students work, what's AI work, you know, where are those dividing lines going to fall.
And then secondly, this debate that you and I have had in the past, Ben, about how to really make sure that these tools enhance and don't shortcut or undermine, you know, learning. So, um, you know, if you're thinking about a professional context, it's all around the end product, right? Like what the end result is if you're using a tool and that's really where the value sits.
But from an education perspective, it's all around, you know, the process, like how a student gets there. And so, you know, the more and more we have these kind of exciting tools. That shortcut, that process. And I think the kind of, you know, more and more important it is to really think about how can we combine these great innovations with, and what we know works when it comes to actual learning improvement and learning benefits.
So really exciting, great new tool, but I think kind of compounds some of the questions that the education space has around how to best navigate this to ensure learning's happening.
[00:06:54] Ben Kornell: As I've seen this rolling out, it's been interesting to see everyone's reactions from their different corner of the universe.
We hear from, you know, data privacy and safety people. Oh my gosh, this creates new IP issues. And then others saying, Oh, well now everything will be cited and have footnotes, that's actually an improvement. You hear from people that are thinking about job markets. Oh my gosh, this is going to eliminate jobs.
And then other people saying, wow, you know, most companies don't have a budget for a data researcher if they had one person who part of their job is using this now you're going to see research infused and so it feels like every single piece of this is a double sided coin and. When it comes to higher education, for education in general, how do we want to represent what is authentic intellectual work, and how do we want to also acknowledge that authentic intellectual work could be inclusive of leveraging AI technology and tools, and so one of the big challenges with all of this is Not everybody has 200 bucks a month to access this.
And so for a period of time, there is a real haves versus have nots element of this. And I think the longer that that break occurs, that will actually fuel more of the tension. Once it's universal, like many of these other pieces, I feel like the kind of sense of one group getting a competitive advantage over another will die down and then we'll have more of our.
Typical moral, ethical, practical dilemmas with all of this. Having tried it myself, what I will say is one, it requires more patience, more think time. This is not search. This isn't a search replacement. And second, you know, the ability to prompt, re prompt and multi prompt. You really have to have a deep question that you're trying to explore.
And you're looking for a pithy answer, this is not the solution. It really, to me, this feels more like a research assistant that would be working for a researcher, helping surface ideas, but it never struck me as like a full replacement of a human research student. Or something I would want to turn into my professor.
[00:09:14] Libby Hills: No, no, no. Because I mean, as you said, Ben, like the ability to make the most of it requires quite a bit of skill on the prompting side. And I think it's actually kind of in the short term, I can see folks who might make benefit the most from it is being folks who are experienced with giving kind of clear instructions in the workplace.
So potentially more, you know, senior experience members of the team in the short term who are more familiar with, Hey, how do I. Give really clear instructions to more junior people on my team and then how do I iterate on the product with them to get us to a good end result. So I see, yeah, in the short term, that skill sitting more with folks who might be like running teams versus folks who might not have yet had that professional experience.
[00:09:55] Ben Kornell: So given your focus at Yacob's Foundation on research around EdTech efficacy, do you think that this might be an opportunity for EdTech companies to use? These kinds of programs, whether it's the OpenAI one or another one, to actually build more solid connection to research based practices and implementation?
Or do you think this is going to kind of whitewash the space where everyone can kind of run their quick query, say it's researched back because they ran it through an OpenAI system? Like, how do you think this plays out specifically in edtech?
[00:10:30] Libby Hills: Yeah, that's a super interesting question. I mean, I think making research more accessible to developers is a really important step, but it's not the only thing that needs to happen in order to make products more evidence based and more likely to be efficacious.
So, yeah, a big thumbs up to anything that contributes to helping people better understand What the research is saying and what that might mean from a product design perspective, but that needs to be coupled with, okay, how, how are we actually going to integrate this into our product roadmap and build around this in a meaningful way on an ongoing basis.
So yeah, great step forward, but more work to do to make sure that we're building evidence based efficacious products.
[00:11:09] Ben Kornell: Yeah, I would love to see a funder step in and say, okay, we want to make this platform available to more ed tech companies, to their chief product officers or something. My article two years ago was like donating 1 percent of computes to social impact purposes.
This would be another example where it's like, man, it would be great to get some free. licenses. Speaking of cost,
[00:11:34] Libby Hills: is that a hint, Ben? Is that a hint? Yeah, very subtle. Yeah. No, it's cool. I mean, and just to build on that, I think there is some really cool organizations who are, you know, really thinking about how like a chat interface can be used on top of like high quality evidence repositories and sources to tackle exactly this problem, right?
So I know like There's an organization in the UK, EEF, who are looking at this, you know, I think folks at Stanford are thinking about this. So I think that's a really great, great development from the space.
[00:12:06] Ben Kornell: Yeah, well, at the same time, we're seeing new models come to the market and DeepSeek really shook things over this last month.
We saw a huge stock market run on many of the AI invested companies because there was a real fear that this Chinese developed AI model was competing and potentially even out competing. The U. S. Based systems. I think things have settled down with regards to the competition. But what it brought up were questions around who wins in the A.
I. Landscape. And how does all of this play out and kind of number one was. This idea of fast following, it continues to show that no matter how far ahead you push it, you're only six months ahead of your competitor because it's very easy to reverse engineer what you must have done to achieve your breakthrough results.
Second is lack of competitive moats. And so this idea of switching costs between one or the other, it's just incredibly easy to switch and so hard to build in value. And then I think the third one is around, This cost versus compute power, like change, which is, you know, as previously thought, this is going to be hundreds of billions, if not trillions of dollars to build AGI or a supercomputer.
And what we're finding, and there's a lot of debate over what the true cost in China was to build this. Right. An order of magnitude smaller than OpenAI, I think is fair. This idea that we could actually achieve greater. AI results by focusing a little bit more training it on more focused things. And then by using a labor force, that's lower costs.
And what this leads you to is there are likely to be more global entrance into the AI space race here over the next five years. And ed tech companies have to figure out what to do with all of these. And I've been talking to a bunch of ed tech companies as they've seen deep They're like, basically the idea of having an.
AI or an LLM underneath their product. That's from China is just a non starter in the U S market. What's your read on how this impacts ed tech and how we should be thinking about this space evolving going forward?
[00:14:24] Libby Hills: Yeah. Great question. I mean, I think for me, one of my takeaways was this is like a more.
Shift towards large language models becoming commodities and that there's gonna be, you know, less and less kind of difference in terms of performance in the models, less and less difference, you know, in cost over time. So I think for me, that was a kind of, it was another signal that we're on that trajectory, you know, over the coming years.
I think, you know, hearing similar to you that For lots of companies, you know, using a model from China, just a non starter, you know, you see countries banning DeepSeek, concerns around, you know, data privacy, how that's going to be used. So I think that, you know, that there's a question there from models outside of people's home turf, that are those questions going to continue to be really salient around, you know, trust and privacy, et cetera.
But I completely agree with you that, as well, that we will see that growth, you know, internationally, that, you know, this also shows us that it's possible for non US, non European companies to fast follow, and that we're seeing a real reduction in that time. Melanie, you said six months, but I would say, you know, a few months, and that we're probably going to see that potentially shrink even more.
I mean, I think, you know, taking a step back as well, I think that some of these developments will maybe stop making headlines as frequently as they have been that they're actually, you know, when there is a new model, unless there is a significant technological development represented by it, that Perhaps it'll attract less and less attention moving forward.
You know, given that some of their recent releases, we haven't actually seen a big step forward in terms of the tech, you know, I think people would argue that that is the case with deep seek as well, that actually there's not much that's new there. And so perhaps it hasn't quite warranted all of the sort of hype and concern and stuff that occurred during the elite release.
[00:16:07] Ben Kornell: Yeah, I think it's not so much the implications to us as consumers of the A. I. It's more implications of Investor side, which is it's really hard to differentiate in that point you made about commoditization and if deep seek can do this What's to prevent your competitors? I think the real loser here in the deep seek Grace is Meta who it's like, Whoa, you got outperformed by someone spending a fraction of the money and they're doing it in an open source way too.
So they're even kind of like circling around your entire strategy. That's got to be a shocker. And I'm sure there were some folks in the Facebook towers that were very upset with that. On the implementation side, we've seen a bunch of attention grabbing headlines around AI in education. The big one this week was open AI launching with half a million students at California state university systems.
Basically ensuring that everyone has. a pro version of GPT. Is this a meaningful development? Is this like just window dressing? Is it something that you think is going to be a common trend across universities? What's your read on the higher ed space and relationship to AI?
[00:17:24] Libby Hills: Yeah, I think it is the start of a trend.
I think universities, you know, higher ed are under pressure to respond to, you know, changes in the workforce that are occurring due to AI. So I'm not sure we'll see as many such kind of bold and ambitious, you know, announcements and plans as we saw with CSU, but I do think it signals, you know, a wider trend about universities wanting and needing to be seen to respond to, um, appetite for a kind of AI ready workforce.
So I think we'll definitely see more and more equivalent projects, even if not on quite the same scale. I think it's, you know, if I put on my kind of impact and evidence and like educator hat I think there are some questions around how much, you know, increasing access to a tool is actually gonna lead to better outcomes and, you know, achieve some of the goals that I think the project has.
And I think there are slight shades of kind of similar announcements in the past or, you know, attempts to expand access, you know, like MOOCs, you know, anyone remember MOOCs? But I think unless that's coupled with, Hey, here are some of the. Subsequent changes to like curriculum and assessment that need to be made to actually embed use of these tools in a sustainable way.
You know, I think there are some question marks there about how much it's going to be able to deliver kind of improved outcomes. So maybe that is part of the plan. It just hasn't made the headlines, but I think that was definitely a question for me when I was digging into the story.
[00:18:48] Ben Kornell: Yeah, no, I think that's the headline.
You hit it, which is, is this really going to change learning outcomes on what I think is the most important part of this announcement is that teachers and professors at CS use can now universally expect that every student has access. To an AI system and that it may be as much a shift in mindset from the educator side, which, you know, they should be changing their assignments so that they're not easily done just by the AI and they could even assign things to use AI as part of the process.
So while I think that from a student standpoint, one, like many students already had access to these tools. To yeah, I don't think that we've really set up really thoughtful pedagogical use cases. Yeah,
[00:19:44] Libby Hills: exactly
[00:19:45] Ben Kornell: But in terms of changing the like thinking of the professorial class This could be a really important move Where they can essentially build that into their course materials.
They're thinking their logic and that's to be honest, like in the change that we're going through from a non AI to AI world, it's actually the adaptive change that's going to be the hardest part. Cause the technical change is far outpacing our human ability to catch up.
[00:20:14] Libby Hills: Yeah, I mean, one of the other questions that I had interested in your thoughts on this, Ben, is like, you know, what's the impact of kind of moves like this on, you know, startup ecosystem?
You know, if there are similar kind of big shifts and big partnerships with big tech companies, but if that particular type of partnership is something that we're going to see more and more of, what does that mean for startups who are Building kind of narrower, but you're narrow tools that might be useful for either administrators or professors or students, but aren't offering the same, you know, kind of full package that someone like an open AI can potentially.
So yeah, not sure if that's been on your mind at all, but that was definitely on mine when I was reading the news.
[00:20:53] Ben Kornell: Yeah, it's, I mean, it's going to be harder to sell features that are chat GPT wrapper. Kind of products into universities when they're like, well, we already have this universal access, and it's still relatively underappreciated, the kind of concept of building your own GPT, but the ability, once all these students have pro models, they'll be able to take all of their papers and, and so on, train their own version of the t.
And so it's going to be really difficult for an off the shelf ed tech vendor to compete with something like that. So baking that into your ed tech startups, thinking that everyone has this, like, how do we actually scale it? It makes me also think there's a professional development or like services industry that could pop up here.
Which is really around helping students and professors leverage the AI that they already have access to. I know that's not venture backable, but what a great way to build a consulting business if you're supporting CSU students to leverage AI to improve their outcomes.
[00:22:03] Libby Hills: Yeah, I think that's that's super cool.
I've been, I've been thinking about the same a lot and I think there was some really interesting examples again, not venture backed examples, but I know like Stanford have what they call an AI tinkery, like a kind of maker space to help educators figure out how to build their own tools. And there are like similar initiatives, which I think are super interesting development.
I think anyone who's like had a go at building their own GPT or something like that. Building their own app using some of the no code tools that are out there at the moment, it can feel really empowering and you know, you can build something completely bespoke and I can really see motivated tech engaged administrators, school leaders doing the same.
Then they've got total control over their own system. So I think that would be another interesting trend and development in the space over the next few years.
[00:22:48] Ben Kornell: Well, as it develops, we're going to be talking about it here on Tech Insiders and Ed Technical. You know, we've kind of done the around the world in AI.
We've also jumped into higher ed. Let's talk a little bit about K 12 before we go to the big news in U. S. Department of Education. On the K 12 front, we have really sobering reports from the NAPE The NAEP score really is the kind of pinnacle of performance data year over year because they bring things up from an individual state level all the way to a national level and looking at those assessments.
And I know you've been focused on global achievement levels, but that NAEP scores recently have just been shocking. They've essentially shown. That we haven't closed the COVID learning gaps, and in many ways, we're seeing a precipitous decline in mathematical skills and abilities. And so, you know, getting into the report just a little bit, states like Kentucky, Louisiana, Mississippi, Tennessee, and West Virginia, all red states, did show same or better improvement, but Arizona, California, Florida, Massachusetts, Oregon, and Washington, a mix, but largely blue states, actually showed huge lost ground in both fourth grade reading and math and eighth grade reading and math, even as investments improvements.
So what's your, I mean, help us at the high level, you're so deep in this like efficacy world, high level, what should everyone's takeaway be? And then as you're looking at this as someone who cares deeply about research and connecting that to impact on the ground, what are you thinking about?
[00:24:29] Libby Hills: So firstly, I'd say you're not alone, that internationally results, we're also seeing declines around the world.
So it's not a US like specific trend. Some of these declines have been happening for a while pre COVID, like expedited by, you know, COVID, but it's not just a US specific issue. I don't think that anyone, there's no one thing, right? I think it's like a collection of things that are contributing to some of this, you know, some policy, some socio economic, you know, increasing inequality, you know, COVID, digitality, more screen time.
I don't think any one of those things, I think, you know, people's best guess is that it's likely a collection of different things. What I would say in terms of how to respond, what next? I mean, I thought it was really interesting in some of the reports that there, as you're alluding to, there are a couple of bright spots.
I think, you know, Mississippi seems to have seen in the U. S. like significant gains and then Louisiana. So I think from areas that are doing well from states and places that are doing well, like what can we learn from there that could be useful elsewhere? And, you know, as always in education, that's likely to be pretty context specific, but.
I think one of the common threads and that this is also true globally is that, you know, getting those strong foundations in early around literacy and numeracy is, is like pretty critical. And so places that do that really well and have figured out the best approach to do that, you know, tend to be better placed than others to improve some of those, you know, really key early results.
[00:25:58] Ben Kornell: Yeah. I feel like the challenge that we're seeing also is that there is no clear silver bullet. And so While everyone is kind of lamenting these lower scores, we've seen, okay, throwing money isn't working. What is a way to, I think there was an edunomics where they said we need a full reset. Like, what is the way we do a full reset, and how do we do that in knowing that schools are now dealing with Healthcare provision to students.
They're dealing with behavioral and social, emotional challenges. They're working across special ed. They're working across high flying students and students that are struggling. And we really have all of this adaptive technology sitting up here. And I keep feeling like, gosh, there's gotta be a connection point, but in our, like.
Technology skepticism cycle. We are at high skepticism right now, I think, in our space. So, you know, as much as we like to bring our ed tech lens to the solution and, you know, we've got the hammer and everything looks like a nail, it does feel like there's even a research community. Resetting of like, how do we measure what success looks like?
How do we even look at pedagogical practices and kind of getting back to the basics of teaching, learning, and then building back up from there. That seems like a daunting task to me. I, how do you make sense of that? Where do we start?
[00:27:29] Libby Hills: I'm less daunted by that because I feel we know so much more now than we ever have done about the kind of science of teaching, the science of learning.
And, you know, if we are starting with, Hey, let's define what good teaching and learning looks like, as you said, and which I agree is a good starting point. We know a lot about what good looks like now in terms of good teaching and learning. So let's really like, you know, help people. Engage with that and kind of build out from there so that there's more and more like good teaching and learning happening as much as possible in, in as many classrooms as possible, you know, in the US, but also in other places around the world.
So that does give, make me feel more optimistic, you know, about our ability to tackle the challenge. I would say, you know, I'm for me, you know, I'm techno optimist slash techno realist. Like I think you are better. And for me, of course, technology is a really powerful tool to help make that happen in lots of different ways.
I would say, as you alluded to, it feels like there's so much expectation on schools, you know, at the moment. And I think the tougher the world gets outside of school, it feels like the more and more, like schools have to try and deal with. And I think, you know, there's only so much that a school can do really to try and tackle, like, you know, all the other problems that are happening.
And so, you know, how can we Just, you know, really help schools to figure out what they're really well placed to help with and just help them really excel at that and not put too much on poor old, you know, teachers and schools in terms of our expectations of them.
[00:28:55] Ben Kornell: Yeah, well, this is one where, you know, shining a bright light on what's working, I think, is going to be really critical for you and for us at EdTech Insiders.
Because I do think there's this spiraling sense of lack of hope, but I love your optimism. And I also think like never before have we had the kind of access to data that we can and should have today. That is coupled with the announcements at the department of education.
[00:29:23] Libby Hills: Oh no, we should have ended on the optimistic note.
And now with this guy. Yeah.
[00:29:29] Ben Kornell: I'm actually so glad to have you on the pod and I know like. We wanted to bring a global perspective to today's news, but this is a very U. S. centric topic. So the Trump administration has been working through their efficiency agenda, quote unquote, to look at department by department and education department appears to be next.
And one of the first actions that they've taken is terminating contracts that involve research and data, you know, the education community, it's sparked an outcry. I think there's. You know, the countervailing perspective is we've been publishing data on this for a long, long time, and the results of that publishing of data hasn't actually translated to any real impact on student outcomes.
So there's a debate around even what's the purpose of national education data and national education research programs. And it does feel like the state governments would have to step in if the federal government steps back. As somebody who cares about this phase, what's your read on the current situation?
[00:30:38] Libby Hills: Yeah, I mean, and I hope I can do this justice, you know, not being in the U. S., um, although I have lots of partners and connections to the U. S. I mean, it's just, as someone who is, you know, works for a foundation that focuses on evidence and research, it's really hard for me to, to understand, you know, if what you want to do is be able to make decisions about what's working and, like, focus your resources on those things, which, you know, I think, Some of the new administration are claiming to want to do, how can you do that without having data, you know, without knowing what's actually working.
So kind of cutting the programs that are going to give you that information to help you make decisions is hard to fathom. I think there's a real, you know, what we've seen is more of an outsider perspective is that some of the You know, national frameworks that connect to evidence in the U. S. So, there's one called, you know, Every Child Succeeds Act, that ESSA Tears of Evidence, which folks may be familiar with, which is a, you know, a federal framework, has actually been, you know, really helpful in providing clarity and making it easier for, You know, um, vendors to be able to efficiently work across states because there's, you know, one standard that is applicable across different states.
And it is actually something that we've, you know, held up as, as a kind of exemplar for other parts of the world is how to provide clarity around, you know, what, what good evidence looks like. I think there's a question when it comes to, you know, such potentially significant cuts on, on the research side, you know, as someone who works in, for, in, for a funder, what does that mean for, uh, You know, for philanthropy and, and, you know, how much can philanthropy, you know, step in to try and potentially support on some of the significant potential cuts on the research side.
I think, you know, the reality is the numbers are just so big when it comes to, you know, the size of some of the funding that the federal government can offer. There's probably a limit to how much philanthropy can do, but I think that that's a question on some folks mind.
[00:32:33] Ben Kornell: I think the challenge that we have is we don't have a great track record of seeking.
The work that's been done at the federal level and moving the behaviors at the state level. And this is where you have a challenge in the U. S. where there's local governance of school boards, then there's state oversight of state education departments, and then federal oversight. Whose job is it to sponsor research?
Whose job is it to create the content around which programs are working and which aren't? And generally speaking, the view has been that the federal government is best positioned to do that because they can look across the different variables of state and local. And so, you know, there's a funding question, there's an implementation question, and then there's a governance or responsibility question.
I think what makes this particularly challenging is that Congress has earmarked this as a federal responsibility, and then you've got the executive office saying, no, we're cutting this. So, back to, like, what do we do with this? I feel like everybody in the education space, you've got to kind of Keep one foot in front of the other, keep plowing forward, and know that this is going to all get taken to court.
It's all going to be hashed out. And just like those contracts that they froze, like a month ago, that then there was a court order on his day, and then there was lack of payment. This story hasn't been finished yet. So I feel like, the last thing I would just say is, One of the political elements of this is organizations like the Gates Foundation have also been pretty big sponsors of federal or national and global research.
And there is a kind of like, who are the billionaires who are in? Who are the billionaires who are out of this administration? And so like there's a gap here where the philanthropic community led by Gates and led by some of those others could and should step up to really clearly define their commitment to research and really define their vision for how this works, because I think we're experienced a little bit of a gap of leadership from the other side of the debate, and that's left a bunch of education organizations and institutions.
[00:34:56] Libby Hills: Yeah, it's really tough, you know, it's really tough seeing the news. Yeah, it's really tough seeing the news. And yeah, I know there are conversations happening about how philanthropy can respond potentially, but also, as you say, Ben, also important to note, there's a lot of uncertainty right around how this is going to play out.
And what the specific impacts are going to be some of the existing programs. I would say one glimmer of hope was that the small business innovation research program, I think is one that seems to be potentially safe. And I know that from a lot of our partners, that's been a really great source of research funding for kind of early stage innovation and ed tech.
So I'm trying to find a small glimmer of hope here, but that's one that I'll point to.
[00:35:36] Ben Kornell: Yeah. So we've talked a little bit about these cuts. I think the other thing that the industry is grappling with are also the Essar cuts. It's actually a cliff. And that was part of the funding of COVID. And those Essar cliffs have taken another victim.
In this case, it was a massive kind of overnight shutdown of FEV tutor. FEV tutor, uh, essentially was a tutoring program that existed far before COVID. The founders basically were able to work with overseas tutors. Who would zoom in and kind of tutor us based students thereby cutting the price to serve in tutoring sessions, online tutoring sessions, and then they experienced massive growth as well as they got acquired by Alpine capital private equity group, kind of at the peak of the market and with the Esser Cliff.
Coming and funds naturally, this is not really in the purview of the Trump administration with them kind of wrapping up, they were experiencing really big misses in terms of their growth projections, as well as losses in their existing. This kind of follows what I would say was the meteoric rise and fall of paper.
And so there is a question around this tutoring space and where it is headed. A bunch of educators have also talked about FE tutors. Challenge was it's not in person tutoring and it's not fully automated AI tutoring. So they were caught in the middle. What do you think the story is here? Is it that the Essercliff is driving tutoring businesses out?
Is it around the future of tutoring? Is it around like capital structure and, you know, private equity companies buying ed techs? What, what do you think is the story here?
[00:37:19] Libby Hills: So. I'm intrigued by the story because it was so sudden, right? It was so abrupt that the closure and you know, we like you speak to a lot of companies and other investors in the space and a lot of people have been, you know, planning for a while for the shift in funding.
And so I think, you know, I'm kind of intrigued by, by kind of sudden events like this one, when it's not a, it's not a shock that ESSA has been, you know, ESSA funding is going to be winding down. I think there's a really For me, what the story is like, what next for tutoring and what does like steady state business as usual look like for tutoring?
We know that, you know, high dosage, high quality tutoring can be a really great driver of like better outcomes for students, but it's often seen as like an add on rather than like integrated into business as usual within schools. And so, you know, with the end of a massive funding, I think now we need to like figure that one out.
Like how much is it and should it be integrated into, you know, normal school curriculum, normal school business. And if so, like, how is that paid for? What's the sustainable model? I think there's an interesting example in the UK. There was a national tutoring program that was like set up to provide, you know, catch up support after COVID.
And there was, it was a big national program initially. And that basically didn't work. It didn't, it didn't, it sort of failed and the government decided to redesign the program and make it school led. And so schools were able, funding went directly to schools and so like they were able to arrange their own provision.
And so I, I think perhaps there's something here around, you know, a more school led, school driven model and way of designing tutoring that perhaps is more sustainable versus one that comes from states and districts.
[00:38:59] Ben Kornell: Yeah, I feel like there's a search for a sustainable model, and there's still a lack of evidence that full AI tutoring really works well.
And so, by the way, even if there were, then the question would be, is that actually commoditized down to zero because of what the AIs can or will be able to do? It feels to me like this is a space where the efficacy is being questioned. Because of this, the name score results. The future of the staffing and business model is in question and the funding is ultimately in question.
And so it's a, I think it's a proceed with caution space for most educators. That said, whenever you see a space like this, it is ripe for innovation. From an entrepreneurial standpoint. And so I do think there's a couple of companies that I'm watching, and some of them have actually grown up in more of the early literacy space using AI.
Or some who've been long time solvers. So like Ello coming up from the reading AI assisted standpoint, they're doing an AI tutor that I think. Could be massive at a cost that is accessible to the developing world. Like that's really exciting. Then you also have third space learning, which has like 10 years of overseas.
Chat and they're in the UK and they're basically building an AI guided tutor on that. So I actually think this makes this space even more exciting because you, some of the big players are going to have to disrupt themselves or die.
[00:40:39] Libby Hills: Yeah, it's a really exciting space. Lots of ongoing innovation. I mean, I think not to get too meta, but like, I think there's this, you know, enduring question, like, what is tutoring these days?
Like, what do we mean when we're talking about tutoring? So I think that one needs to get kind of nailed down a little bit, I think sometimes.
[00:40:54] Ben Kornell: Yeah, totally. And when we have people saying this is high dosage tutoring. It's got to be better than a search bar, you know, basically,
[00:41:01] Libby Hills: yeah, exactly. Yeah, yeah.
That's not a tutor. Yeah.
[00:41:05] Ben Kornell: Okay. Well, and by the way, you actually have a great podcast episode on this one. So people who want to dive deeper into tutoring and AI tutoring, check out a technical.
[00:41:15] Libby Hills: Ben, are you talking about the episode that you came on? Are you bringing up your own episode? That's right. Yeah, yeah.
Yeah. Love it. Love it. Awesome. Yeah. Great.
[00:41:28] Ben Kornell: So, last, we're going to go fastball down the middle of your sweet spot, European edtech marketplace. Our great, great friends at bright eye. We always really appreciate their analyses. They have a tech market report. I think there's a bunch of kind of new data coming out about 2024, but also looking prospectively, I'd be curious, what are your main takeaways?
As you look at the European ed tech landscape and. You know, what have we learned and where is it headed?
[00:41:56] Libby Hills: Yeah, great. And thanks to Brighto for another great report. I love that the annual funding report is awesome. I mean, I think the note that the report strikes, which I agree with, is kind of cautious optimism about, you know, where the funding market is, both globally and in Europe.
So, I think there was, you know, global funding increase from 2023 2024, which was great. Increased deal count, which is all positive signs, driven a lot by the US. I think one of the points that they make, which stood out to me there, is that some of this is driven by generalist activity and not just specialists.
So that's a kind of good signal around, you know, the kind of potential of edtech, which is on lots of people's minds given where we're in still. The edtech winter, you know, where funding's been tough. I think, interestingly, Europe was the only region where, in terms of the annual volume, there was a bit of a drop, um, between 23 and 24, but actually Q4 was, was, was the highest quarter.
So yeah, fingers crossed, that's a good sign for the year ahead. And I think one of the super interesting framings for the report to take a bit of a step back, which is a trend we see a lot in the space as a fund investor. is this kind of redefinition of EdTech. So interestingly, the report has two halves, one using their more traditional definition of EdTech and all the analysis using that, and then a second where they use their kind of newer definition for EdTech, which is a broader, more expansive vision, kind of EdTech 2.
0. And I think this is a really interesting trend that we're seeing across the space at the moment, across lots of different funds. You are expanding out of more kind of traditional ed tech, like formal learning, K 12, higher ed, and pushing into more and more kind of expansive versions of corporate learning, you know, encompassing HR tech, productivity.
And then others who are kind of expanding more into different verticals, quite explicitly in some cases, you know, into health, FinTech even. So it's really interesting working a lot with funds and talking to them about the evolution of their thinking and the rationale for that. I think some is kind of thesis driven and some may be a bit more pragmatic at times around, you know, where they're seeing deals that they're excited by, but certainly super interesting trend and was, was cool to see BrightEye like.
Provide both analysis using, you know, using more traditional and that older, and then a newer definition of what counts as EdTech for them. I would say like, I'm all for like progress and all for like, you know, expansive thinking. I think there was definitely like, you know, a few companies that have kind of raised my eyebrow here around like, you know, EdTech, even if we are being more expansive.
You know, helping gamification on TikTok or like relationship apps, I'm like struggling to see even with a board of thesis, like how we can squeeze those ones in.
[00:44:32] Ben Kornell: There's like two elements of this. One is, is edtech a vertical or is edtech actually a horizontal across a bunch of vertical? And I think that argument, I think is really powerful that learning.
It has actually become a distributed ownership Ross, all the verticals. Now, when I create an ed tech fund and I want a generalist investor to say, Oh, I want to invest in your fund. Do am I reframing a dating app so that it fits into my thesis? That I'm not so on board with, you know, a company like SANA AI is really an interesting one where it's like unlocking your company's knowledge base, like your shared knowledge through AI and then creating trainings and upskilling the main trend that really captured me in the bright eye report was around.
This idea of like micro learning, micro credentialing, you know, we kind of have these big blocks that are now just getting broken down at a time and again, down to the molecular level. And it's very hard for like high compliance systems to manage this, but if you have learner directed, and in this case, most of Bright Eyes investments are with adult learners.
If the learner can kind of go grab what they need just in time, right? They need it and is less concerned about Compliance with this credentialing system. It's actually a really exciting time So if you combine that with horizontal, you know ed tech is a horizontal not a vertical I think that like that really opens the Ecosystem that we could be supporting.
And by the way, these other specialties where they're diving into learning, they need our pedagogical expertise or they're not going to get the kind of learning that they want.
[00:46:18] Libby Hills: Yeah.
[00:46:19] Ben Kornell: Well, I'm going to wrap us up on that one. I just want to thank you, Libby from ed technical at by night. It's like you're.
You're like Batman at night and Bruce Wayne in her Bruce Wayne is leading the YACUBS Foundation, investing in impact evidence and research in ed tech. And at night she is running ed technical top on my Apple podcast list. So thank you so much for joining us, Libby, and thank you all ed tech insiders, listeners.
If it happens in EdTech, you'll hear about it here on EdTech Insiders.
[00:46:53] Libby Hills: Thanks so much for having me, Ben.
[00:46:55] 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.
