← Edtech Insiders
Edtech Insiders
Week in Edtech 7/1/26: ETS Acquires ACT, Handshake Buys Uplimit, Google's Biggest ISTE AI Announcements, Microsoft AI Adoption Survey, and More! Feat. Karlo Young of Teachstone & Tom Sayer of Ello
01:19:32
Loading summary
Transcript91 lines
- [00:00]
A
This season of EdTech Insiders is brought to you by Starbridge. Every year, K12 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 and 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.
- [00:32]
B
The launch of Gemini Notebooks for school. This is actually something pedagogically as a format we've had in the past, a notebook where you take all your notes and then you reflect on it and that becomes your source of truth as you write a paper or do a science experiment or reflect on your learning. So the irony that those who brought you Notebook LLM would bring you gemiini Notebooks for Education, I think that's a really interesting development and it makes me think about students as they're collecting their learning, how they're creating repositories of that learning for each class so that they can draw on it. I think it's just a really well guided thing.
- [01:18]
A
Welcome to EdTech Insiders, the top podcast covering the education technology industry. From funding rounds to impact to AI developments across early childhood, K12, higher ed and work, you'll find it all here at EdTech Insiders.
- [01:34]
B
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 backchannel insights from Alex and Ben. Hope you enjoyed Today's podcast. Hello EdTech Insider listeners. We are back. Back in full effect. It is summer. It is Ben. It is Alex. It is all things EdTech. Welcome to your happy place. EdTech Insider speaking. EdTech back. Alex, did you miss me while I was gone on my RV vacation?
- [02:18]
A
I did. I did miss you. But I had some incredible guest hosts. It was so much fun to talk to Thomas Rogers from Whiteboard to talk to Gemma Lemowicz Picoult from samvid Ventures, who is a mutual friend. She's amazing. And Dani Pico from Forum Ventures, totally great guest host. But of course nobody has the chemistry. We've done this for so long. I'm very happy to have you back.
- [02:40]
B
We're back and there's so much going on in the edtech world. Normally summer is a quiet time. Not this summer you mentioned as we were prepping the screen free summer as the theme. So we'll get to that. We'll also talk about some major acquisitions that have come through. Like normally, summer is quiet time and there's a lot to catch you up on. And then of course, our recap of iste. But before we do that, what's coming up on the pod?
- [03:07]
A
Yeah, so we've been talking to incredible people recently. We had an interview with Sue Kim, who's the CEO of Brilliant, who I've actually never met before, and talked to her and it was the most fun conversation. I love talking to her. We talked to Julie Young, who is the founder of the Florida Virtual School and has done incredible work with the ASU Prep. She's like a pioneer of online learning before it was a thing. We talked to Antonio Pietri, the new CEO of PowerSchool, our old friend Tom Sayer from Ello. And I just had a great conversation yesterday with Michael Preston, Lisa Gack and Mehtari, who are doing all this amazing work with the Joan Ganz Cooney center about co design for edtech. And they're always amazing. And those are just the, those are just the shorter interviews. There's so much happening always. We're talking to Kavita Guy soon, as well as Aidan McDowell from UniClearn, which has been getting a lot of buzz in the edtech space. So always amazing conversations. Tune into those.
- [04:03]
B
Well, let's dive in. I think the world is buzzing right now with some major acquisitions we had. ETS acquiring act and Uplimit was acquired by Handshake. Which one do you want to jump into first?
- [04:16]
A
Let's talk about the ETS acquisition. I mean, these are two big names in standardized testing. Two big acronyms in standardized testing. And the ACT is being acquired as a for profit by ets, which is a nonprofit. You can unpack what exactly what that means, but it looks like ETS has also been, they basically have had been the administrator of the SAT, which is a College Board generated exam until 2024. And in the wake of that, of the changes where they are not the administration of the SAT anymore, they are now moving hard into the ACT to stay in the college admissions game. What do you make of this acquisition?
- [04:54]
B
Well, first, like I've been beating the drum of assessment, assessment, assessment. It really is one, people always pay for assessment. So from a business standpoint, it's a durable good. Number two is the, the potential with AI in assessment is quite profound. And third, there is a real value to like fully integrating the vertical. And so what this essentially does is it recreates what ETS had with sat, as you alluded to, by bringing in the act, in I think the questions that I have are ACT was a nonprofit. Then it was acquired by a private equity firm, take in for profit. Now it's back owned by a nonprofit. Is that because ACT is just not a profitable business on its own, and therefore it always should have been a nonprofit? And is this like a fire sale by the private equity folks, or is there some sort of like hybrid future that we're going to. I think we all remember a couple of the acquisitions. 2U acquired edX, that was a for profit acquiring a nonprofit in that situation, not the other way around. And hmh, Houghton Mifflin Haracore also acquired nwea, the assessment provider. Again, a for profit acquiring a nonprofit. In this case, it's reversed. So it just raises a lot of really interesting questions around what drove that decision and those dynamics. But you can understand from an ETS standpoint, if you already have toefl, you already have gre, you already have Praxis, which is a teacher certification platform. You've got to have an undergraduate assessment. And they've got something like 50 to 70 years of experience implementing these things. So College Board, take note, you've got a real competitor on your hands that used to be one of your deepest partners.
- [06:49]
A
Exactly. And this comes in sort of a larger atmosphere of many, many universities starting to go test optional over the last few years and then basically spinning back. And you've seen most of the Ivy League basically reinstitute standardized testing as part of the admissions process because the professor class basically complained that they were getting students who are far underprepared and they didn't think that the academic rigor of the incoming class was actually up to snuff. So we've seen a big reversal there. So you can imagine that if you're the College Board or ets, you're looking very closely at those trends and seeing is this, you know, the beginning of the end of standardized testing, which we might have said it was coming a few years ago, or this huge bounce back where they're realizing that the value of it and it becomes a really durable business, as you're saying, it'll be interesting to see. I mean, these companies are sort of borderline Edtech. I would call them Borderline EdTech. The SAT especially is administered digitally now. But they're really like pretty core education and assessment companies. They're at the very center of the admissions world and TOEFL and PRAXIS and APS and all of these important assessments. And it is a really good business. Let's talk about uplimited handshake, uplift. You know, we know the Uplimit folks really well. What did you make when you saw that news?
- [08:03]
B
There was a big question. So Handshake has been around for a long time, and originally it was connecting students with internships and jobs after graduation. And then they became one of the number one AI training platforms where basically they could connect college gig workers with the AI platforms to label data. And essentially the company blew up and we've seen a bunch of folks who've started in edtech who then really hit it in AI, kind of say goodbye to the sector. And so this really doubled down on the original mission of Handshake, which is really around providing connectivity between skills and careers and job opportunities. And I actually think it's a brilliant move because what Handshake has is a fundamental problem with customer acquisition versus lifetime value. They acquire college students, they get the value out of it, and then college student moves on. By being a kind of lifelong learning platform, they could provide enough learning to keep people engaged and then they can monetize on the back end all the AI training. So part of me wonders, is the education going to be the monetizable asset or is the education actually the content that keeps people engaged, engaged, that allows them to train and label data for AI? That will be a really curious thing to see. But Julia Siglitz continues to be one of the key entrepreneurs to watch. She was, of course, work with you at Coursera and kind of birthed this thing before everyone was thinking about fractional just in time learning opportunities to upskill. She was on it from the very beginning. And now this partnership with Handshake means her scale is truly unprecedented.
- [09:54]
A
We'll be right back. Tuck Advisors was founded by entrepreneurs who built and sold their own companies. Frustrated by other M and A firms, they created the one they wished they could have hired but couldn't. Find one who understands what matters to founders and whose North Star KPI is the percentage of deals closed. If you're thinking of selling your edtech company or buying one, contact Tuck Advisors now. Yeah, all great points. Yeah. The three founders of Uplamit are all ex Coursera. They are. That's where they all knew each other. Julius Stiglitz, Jake Samuelson, and Saurabh. And they have some amazing engineers there as well. My read on this, and I'm not, you know, without knowing all the details of it, my read on this is, yes, it is a sense of handshake going sort of back into edtech rather than being more about sort of Being an AI services company or a gig company for, you know, for students or for grad students or PhDs that are looking to train AIs. But I think what's also interesting about this is Handshake, in making this pivot, has gotten really close to the Frontier Labs.
- [11:01]
C
Right.
- [11:02]
A
And the sort of surrounding pieces of this Frontier Labs. Because they are doing a lot of training in direct conjunction with Frontier Labs. They have this really interesting opportunity where they could start to become a sort of training arm of the Frontier Labs. I mean, you have OpenAI and Anthropic and Gemini all putting out their own courses. There's an OpenAI Academy, there's anthropic courses. They put courses out on Coursera, they put courses out in a variety of platforms. But I also think the idea Uplimit is in person, cohort based courses. And you can imagine there being a really interesting strategic melding of this idea of Handshake is a place where you can actually turn your expertise in any domain into AI expertise basically, or into AI related expertise. And while you're doing so, you can actually learn how AI is actually working. Upskill, reskill. You can imagine Handshake offering this to companies, B2B to companies, but basically being a go between between Frontier Labs and companies that are trying to upskill their own staff around AI skills uplift. That's exactly where they were going. B2B. AI upskilling, you know, mass upskilling, even though it's synchronous and I just think it's a really interesting I. It is possible that there may be a little behind the scenes venture matchmaking here as well. Upload's done really well, I think, but it's also like they may have been not moving as fast as they wanted to into B2B. And then handshake has set is flush with money because they've done so well recently with all this.
- [12:28]
B
Well. And both of them have like a GSV connection too.
- [12:32]
C
Yeah.
- [12:32]
A
So there may be a little bit of behind the scenes stuff happening there. This is not, it's just a guess, not coming from any inside information at all. But I do think that there's a really interesting synergy there because I think that Handshake has this amazing scale now and has access. It's basically becoming a go to place for people who want to enter the AI world. And then Uplimit is basically, okay, you want to enter the AI world. We have all of these really amazing trainings and ways to get you ready to be in that AI world. So that was really exciting. You mentioned the screen free summer, Ben. That is, this is sort of my, my internal branding. I just feel like we've gone back and forth on these screen policies. People are, I've started to ask a lot of guests about, you know, what do you make of these sort of screen policies? And I just stood out to me today we saw a really cool article. It was actually really nice to see in the New York Times, big sort of feature tech article about the company yoto, which is a really interesting edtech company I think out of England. Right. I believe that does this basically. I mean technically they have a screen, but it's like this tiny pixelated screen. It's basically an audio company, sells these cards where people can listen, create their own cards, do all these different audio based learning and entertainment. And they basically the article featured it as like a oasis of like goodwill and greatness in a, in a tech landscape that's like increasingly, you know, people are skeptical of tech companies, but this one is just so positive and exciting and beloved. And it was nice to see somebody making a distinction between edtech, I consider yoto, edtech and tech. Right. I was like, it's just nice to see that because I think a lot of ed tech is an exception to the big tech backlash. But it was good for people instead of conflating them, lumping them together, which we've seen a lot recently, starting to say, hey, look at an edtech company as an example of what tech can do that's positive.
- [14:18]
B
Yeah, these are the kinds of articles that really worry me with the kind of simplistic painting of good and bad and you know, screen bad, therefore audio good. 1. I've seen my own kids zonk out on podcasts before. So it's not like mindless doom scrolling versus a podcast. There are times where you can have just as unproductive audio technology experiences as others. I think the other is that what works in like a classroom with 35 kids can often be really challenging to make audio work. That goes for audio inputs and audio outputs. But I think the overall takeaway is that people are experimenting with form factor because they understand that AI is a powerful technology. They understand that digital learning is powerful, but having it all presented on a screen as the only mode of interactivity is probably short sighted. It harkens back to like a lot of what Sam Altman was talking about with how he imagines AI to be consumed. And I know you and I, we pace around and talk to our Claude, you know, I, I have a Claude. I do Claude for work. I do open AI and ChatGPT for travel and, and then Gemini for like fixing things. And it's like they're my friends that I call on but often it's audio and it drives Juliana nuts. But that's, it's a mode that can work for me that is very screen free. But I think in general I just worry about these like stark contrast picture painting, especially by general news media. And I worry that everybody's going to read that and there's going to be a bunch of district supes who are like, let's just get a yoto for every class.
- [16:13]
C
Boom.
- [16:15]
B
Like let's slow our roll on both sides. Like ed tech and screens and AI aren't the savior and they're also not the devil.
- [16:24]
A
A hundred I definitely agree with that. And yeah, you could make a pretty clear case that yoto is not even really ed tech. It's basically entertainment technology. It's like screen free audio tech. But I don't know, I get it both ways. I mean when I see a headline, the name, the headline of this is this music box is a ray of hope for a decadent tech industry. I'm like, that just make it is definitely a, you know, Manichean, you know, good versus evil stance. It's like tech is terrible. But you know what, this one company's doing a good thing. So I definitely get your point there. At the same time, just excited to. I mean he mentions halfway through the article that the Brian X. Chen is the consumer tech lead for the Times that he has one for his daughter. He got one for his daughter. His daughter loved it. He's using listening to the Beatles on it and he made him so happy. Paul McCartney is an investor in Yoto. He's like this clearly like a very personal piece of this to him. He's like, hey, people keep saying tech is really bad, but man, I love my yoto. My daughter does too. And I feel like a lot of people have stories like that, right? I mean they're not. There's a lot of zoning, there's a lot of brain deadness, but there's also a lot of like positive, exciting stories when it comes to personal tech or edtech. And anything right now where you see a headline, especially in the New York Times that is positive towards technology, I'm just, I'll take it. But I agree we're in a pretty weird, we're in a pretty weird moment for all of that. Let's talk about iste, let's talk about iste. So neither of us were at ISTE this year. You know, you're just getting back from this amazing trip and I am trying to do a lot of different things in my hometown. So we weren't actually over there. But there's a lot of interesting news coming out of there, a lot of announcements. ISTE did a mild rebrand. They rebranded themselves still as ISTE instead of ISTE plus ascd. It's still iste. I think they changed what the acronym stands for. Right, Ben? Slightly, but it's a little bit anticlimactic. But they did do that. They rolled out a new index. They're redesigning the index.
- [18:15]
C
Yeah.
- [18:15]
B
International Society for Transforming Education.
- [18:19]
C
There you go.
- [18:19]
A
International Society for Transform.
- [18:21]
B
Moving away from a technology focus towards an education transformation. Very on point with the times, but also we were really having fun with like what should the new brand name be? And it's still is these. So it was like, okay, yeah.
- [18:38]
A
But yeah, transformation. I mean, I guess there is a, there is some subtext there. Right. They're taking the word technology out of it. So it does not seem specifically as a tech nonprofit and a tech organization. And they're adding transformation, which can be a little, you know, transformation can come from any direction. And because they do so much professional development with teachers, they can say transformation comes from within, transformation comes from, from teachers. And nobody can disagree with that. The new EdTech index that they're launching is also in collaboration with both state education agencies and some vendor partners, which is interesting. So they have four Clever, LightSpeed, ClassLink and School Day are sort of their four leading sponsors, I think vendor partners of the new index, which we can go look at it. It implies that there's probably some kinds of changes coming to the index. They say it's now aimed at about saving many, many hours a year for districts on selecting apps, which is kind of what it's always been about. But they're obviously, you know, doubling down a little bit on that and changing the strategy there. But there were other announcements from other companies as well. What jumped out to you, Ben, from. We saw Google, we saw Magic School. You know, what were the ones that sort of caught your eye?
- [19:41]
B
Yeah, I mean, on the big tech side, no one takes education, especially K12, as seriously as Google does. And they put together like a 30 page PowerPoint of all the new updates that they're rolling out for educators and students. Which one? I just appreciate the thoughtfulness that you're not just like you'll figure out as a consumer all the good things we're doing and then you teacher or school will kind of translate it. You and I both got an early conversation with Steven of NotebookLM and we saw the power of basically building a knowledge base and the ability to play with that. Once you kind of load your knowledge base, the ability to play with that and create outputs like the podcast, which was the kind of killer feature, but also insights and chat experiences and all of that. And I think the launch of Gemini Notebooks for School, this is actually something pedagogically as a format we've had in the past, a notebook where you take all your notes and then you reflect on it and that becomes your source of truth as you write a paper or do a science experiment or reflect on your learning. So the irony that those who brought you Notebook LLM would bring you Gemini Notebooks for Education, I think that's a really interesting development and it makes me think about students as they're collecting their learning, how they're creating repositories of that learning for each class so that they can draw on it. I think it's just a really well guided thing. I personally create projects in Claude or I create projects in ChatGPT and upload source documents, but they've really flipped the model and just said, actually that's the core part of the user experience, not the chatbot itself, but the notebook. So that was exciting. And then Magic School had a big release of administrator tools. So that's always interesting to see people climb the ladder when they're going direct to teacher. Eventually they start rolling out administrator tools where the core value proposition I can see is compliance and the ability to shut things off and on. They also rolled out templates, so the ability to kind of mass produce your lesson plans and your worksheets and all those things in a way that could be governed by your administrators. It shows a maturing company that's doing more enterprise sales. But it also, you know, I like the cowboy style, a little bit of the, like going direct to teachers. So whenever you launch those administrator functions, you're always like, okay, is a little bit of the magic going to come out of Magic School? Let's see, on brisk, they launched basically a new build of their platform. You know, Armand claims they built it from scratch, kind of a V2. But the core builds on what they announced in January, which is the curriculum intelligence. This idea that AI doesn't live separately from your curriculum, but it lives in an integrated space. So it does feel like the AI everything, platform wars. I mean, if you think about Google And Google Classroom. They almost represent the old school LMSs evolving into new school LMSs. These are the AI native platforms, magical and brisk, that are coming into iste with lots and lots of new features, trying to win the hearts and minds. And what they've got to be thinking about is these next two years are make or break for Brisk or Magic School. If they become a permanent part of the ecosystem they're in, there's lots that they can do. If they fade, then it's over. And that'll be a really interesting story to watch for the rest of this year.
- [23:31]
A
Yeah, great points. And I think one thing that is worth calling out that sort of spans both Google and the Brisk and to a lesser extent, the Magic School one. But is this move to sort of really integrate the AI and the academic workflows together so that everything is really smooth. So like Brisk's new platform is they're moving from a Chrome extension to a full web platform, which it was this really original and sort of unique go to market, becoming market as a Chrome extension. But they have now rebuilt the whole platform as a native web app, which is going to be more comfortable for certainly for some people. And this curriculum intelligence sort of grounds everything together and would. And part of what Google announced. They announced a lot of things, as always. They did 50 things last year at Iste and probably not that different this year. We could do a whole episode about many of the things they announced, but there is sort of an integration of Gemini with Google Classroom so that you can have Google Classroom data and use Gemini Power to analyze it or to give feedback or to support or to adjust assignment difficulty or all these different things. So I feel like they're sort of headed in the same direction in some ways the way that Brisk is going to the curriculum Intelligence, where the AI is really grounded in the core curriculum that's already being used. Google is saying, well, the Google Classroom is the sort of core of what Google teachers are using, and if we have AI be able to support them in that context and based on the assignments they've already built, the study tools they've already built, it just makes it more integrated and more powerful. A couple of other things that Google announced that I think are worth calling out are sort of some edtech control tools. So they have like a focus mode. That sounds interesting that basically it allows teachers to sort of manage where students go. They can keep students inside NotebookLM. They can keep students inside. This is interesting. Keep students inside the Google Guided Learning Environment rather than Gemini's natural environment so that people can say, okay, guided learning is a better learning experience than Google Gemini. It's not going to give answers as readily. It's trained to be active learning and metacognitive. I only want my students to be using guided learning. And that's important because we know from what we've seen in the past that students don't always actively choose guided learning. But you can do that. So you can direct. Basically a guided learning control will direct students directly into structured learning mode and reduce access to unrelated content, which is a big complaint. And the last one is I don't fully have my head around this yet, but they are developing a model context protocol for Classroom which is like a technical, you know, model context protocol. We know the big one comes from anthropic, but it's basically like an API. It allows systems to talk to each other and connect to external systems. So the idea that Classroom might have a model context protocol that can allow it to connect to external AI tools is really interesting because maybe then we're going to see the ultimate LMS where you have its classroom and it can go out and speak to the brisks and the magic schools or it can go out to speak to Quill for writing assignments. That's really intriguing. I don't know exactly what it means yet, but I think we should keep
- [26:26]
D
an eye on it.
- [26:27]
B
I mean, look, if they make a software developer kit too, it could allow the curriculum companies to build directly the connections and feature sets within classroom too and might disintermediate risk. Like we have tectonic shifts here. And I think in the lms, the AI LMS space, there's far more open questions than there is certainty. And like where in the value chain will the value accrue from a business standpoint? But also like what does the future of interoperability look like? These are jump balls and it's very exciting.
- [27:04]
A
Exactly. Jump ball is such a good metaphor for that. And you know Ben, you've talked about this for years. When we first met there was like one of your main theses at the time, if I remember correctly. Feel free to correct me if I'm getting it wrong, but is that one of the big advantages that certain ed tech companies have is they're install based, they're really deeply already built into schools, especially places. Things like Google Classroom and some of the big LMSS and EdTech companies have the potential to very quickly grow their distribution and grow their user base if they sort of integrate really smoothly into Instructure, into Google classroom. So it's interesting to see maybe there's going to be, you know, yet another generation of sort of integrated marketplace apps inside Google classroom, but really with AI at the center of it, so that actually you can send data between different tools. You're right. It could disintermediate people. Or it could be the sort of way that some of these small edtech companies get noticed is that this is the new place to build, is using the Google Classroom mcp. Because then people can use your, whatever really interesting idea you're using for AI can go right to classroom teachers and they can recommend it to each other and all those things. It's going to be interesting. One other thing that stood out, this was not technically part of iste, but it came out right in, right before it, a new survey from Microsoft about AI in education. And what stood out to me is that these are actually the highest numbers I've seen of any survey in terms of usage. They basically said that 92% of students by their survey and 88% of educators are already using AI for school related purposes. And I was just looking at some of the Rand studies and the Gallup studies and even the Digital Education Council 1, which was in the 80s. It was like the highest I've seen. This is even higher. 92% of students. That is a crazy number. I mean, remember this technology is still only four years old. Did that number surprise you, Ben, or is that just like. Yep, yep, that's. It's the up and to the right.
- [28:57]
C
Yeah.
- [28:57]
B
I mean, part of me, anytime I see a Microsoft survey, I'm like, of course they're using AI because they're stuck with Microsoft products. So they've got to find a workaround. No offense to my Microsoft friends out there, but man, I feel like we always talk about at the beginning the big AI companies and we never talk about Microsoft. And so if I'm sitting at Microsoft Education, I'm worried because people are using these tools and we're not a winner in this. And Copilot has a real opportunity in the education space to be a winner, but they really need to be doing. They need to pull some lessons from the Google Playbook and be much more aggressive. None of the reports are surprising to me. I guess I always go the for what? And if you're using AI to like fill out an email or to write a short paragraph on an assignment, that's a totally different use case than fundamentally reimagining your teacher workflow or students dramatically transforming the products that they create. And tools they create or cheating on everything and using it to eliminate all cognitive load. So it's just like I feel like this binary of using, not using is now irrelevant and now it's really, we've got to get into the what for. And I also will say Microsoft, they are a B2B company. They actually have a quite large install base in K12 education and in higher ed. I think people would be shocked to know how many higher ed. You know, part of it is also the security of systems that are Microsoft are favored by more conservative IT leadership. And so especially higher ed, we just got lots of Microsoft in there that's a sleeping giant that could be awakened to really reshape how AI is used. It just, it's so lagging that I feel like that's all I have to say about it right now. I wish I had more.
- [30:57]
A
I believe he's still the head of AI at Microsoft, Mustafa Solomon, who is, you know, co founder of DeepMind with Demis Hasabas. I mean real, real super duper superstar in the AI space. His incredible LinkedIn by the way, if you ever follow his stuff. He's just like always has these amazing ideas about how to use AI. But yeah, you're right. It's interesting how little Microsoft is sort of part of the conversation when it comes to AI and education, even though they're, I mean it's one of the biggest tech companies in the world and it has so much impact in educational environments of all kinds. I totally agree, but they're, they're studying it, right? I mean they're putting, they're obviously looking at it. So there's got to be some strategy coming from somewhere. Well, we'll be reporting on it when it comes out. Maybe they said something that we've missed, but they do have a big opportunity. I agree with you. And it's just about when they actually take advantage of it. I mean we have some great guests on this episode, so I want to make sure we have time for them. Do you have anything else you'd like to cover today, Ben?
- [31:53]
B
No, I mean we've got some great events coming up this fall, so stay tuned. Big conferences coming up this fall will be in at New York Ed Tech Week in October. In November there's going to be the Stanford Summit again and so check out our website. We're going to be launching our, our page with all of the upcoming events and our webinars will get started back up in August. We've got some great stuff coming there. Matthew Raskoff's also going to be doing a Future of Higher Ed series and we've got the end of our efficacy and AI newsletter series wrapping up. So whether it's on the podcast, whether it's on the newsletter, or whether it's our virtual or in person events, we hope you'll keep connected with us all summer. And if it happens in EdTech, you'll hear about it here on EdTech Insiders. Talk to you all soon.
- [32:43]
A
Welcome to our deep dive for this week in EdTech from EdTech Insiders. We are here with a great guest, Carlo Young, CEO of teachstone, which is focused on improving classroom quality through research based solutions. Carlo served as President and COO at Concentric and an SVP GM General Manager at 2U, overseeing over $250 million of revenue and 500 employees across global education partnerships with over 20 years in Ed tech consulting and leadership. He specializes in scaling organizations, digital transformation and educational impact. He holds an MBA from Northwestern, Kellogg and a Bachelor's from Drexel University. Carlo young, welcome to EdTech Insiders.
- [33:29]
C
Alex, thanks for having me. Excited to be here this morning.
- [33:32]
A
I am excited to talk to you. So first off, let's talk about what teachstone is for people who don't yet know about it. You focus on the science of interaction. What does that mean? And why does human to human interaction and relational intelligence so important for learning, especially in the age of AI?
- [33:49]
C
Absolutely, Alex. So you think about it. When we're born, our brains are wired for connection. You know, you want mom's touch, dad's touch, kind of out of the gate. You know, you listen to the tone of voice, it regulates the nervous system. I don't have to be an FBI interrogator to tell you that eye contact matters relative to building trust. And then you think about language development, right? You know, if mom speaks native language at home, you're learning a different language in school, that language development happens kind of in real time in those back and forth kind of volley. So that is the thing that kind of grounds us is those interaction, that particular connection. You know, whether you're a kid or adult, you know, you want to feel like you're being seen or heard or understood kind of in the classroom. And that really is kind of the foundational way I think about it. I'm super pro, you know, kind of, I, I don't know that I can do all of those things right. So I would juxtapose like the importance here. You think about a lot of the solutions in classrooms for kids. There's some awesome Tools that are out there driving, kind of predictive, you know, kind of learning or helping to either speed you up or slow you down. But the thing about this, like, as a, as a first grader, you go out to recess, you get into a fight with your best friend, you know, on the playground, and you come back into the classroom and you're sitting back in front of your AI, and your AI is just going to say, hey, you're struggling in these subtraction problems, right? You're struggling. So let's dial the distraction problems back. We won't do regrouping right now. We'll move to a similar form of subtraction. But is that really the problem? No. You need that teacher that comes over and says, hey, what's going on, Carlo? Something doesn't seem right relative to what's happening. And you say, hey, I had a little scuffle with Wayne on the playground. And then your teacher kind of can help get you back to kind of regulate your nervous system, get you back to kind of growth mindset at that point. So that's the huge importance of, like, it is the first thing. It's not a secondary thing. It is the first thing. Getting those interaction, that relationship down is the most important thing that kind of drives learning.
- [35:38]
A
And tell us how teachstone gets to that, because I totally agree. But sometimes edtech field, as you say, doesn't have access. It's delivered through a screen or a device. And it doesn't always know the context. It doesn't know that the student is wrestling with something at home or just had a fight with their friend. How does teachstone take an approach that actually incorporates that human intelligence into it?
- [35:58]
C
Yeah, we, we look at and, you know, kind of really explore, you know, kind of those interactions is the first thing. And I think about, like the, the moments in our lives where we got unstuck kind of in a classroom or on the field. It was usually a teacher or a coach to help get that unlock. Right? So we really focus kind of on that. That element is kind of the most important. So we're looking at those daily interactions and measuring that, you know, kind of along the way. And there's decades of research to kind of support this work. Right. You know, and it's saying that it's the strongest predictor of a children's success are these interactions. So we took all this research and drove it to practice. A lot of solutions in education don't necessarily take evidence and kind of drive it into practice. A lot of great research that's been Done over the years we've taken that and operationalized that into what we call the classroom assessment scoring system. The acronym for is class Ha ha. Where the goal standard relative to measuring those interactions. I think about my own experience stepping into high school, went to public middle school, went to a top performing kind of private high school. And I was writing book reports in middle school in English, ela, writing book reports. I get to high school and I'm now being asked to be charged with writing these essays. The first book that I started reading was the Odyssey. A complex epic poem. Is not the book you reading to start to learn how to drive essay writing. So needless to say, I go from like, you know, straight A's kind of in middle school to like struggling, you know, kind of in English in high school. But I had a teacher who said, hey Carlo. I went back after class, said hey, I need work on these papers. And he kind of worked with me. Fast forward out. A few months later, to my surprise, I was nominated by that teacher to be a writing center counselor. So I went, I knew that I could write. I just didn't know this format of writing. And once I kind of unlocked that, he saw potential in me to kind of go and become a writing center because I went on to like do technology, you know, kind of made engineering kind of in undergrad and clearly could write, you know, kind of compared to a lot of my friends, I could write pretty well. Right. You know, so that, that is those types of teachers unlock the genius that is kind of in all students. So that's our thing is we want to make sure that we level the playing field for all students relative to these great interactions.
- [38:11]
A
We'll be right back. Innovation in pre K to grade learning is powered by exceptional people. For over 15 years, EdTech companies of all sizes and stages have trusted higher education to find the talent that drives impact. When specific skills and experiences are mission critical, Higher education is a partner that delivers offering permanent fractional and executive recruitment. Higher education knows the go to market talent you need. Learn more@higheredu.com that's H I R E edu.com yeah, it's really powerful. And as you say, there's lots of evidence that that relationship, that trusted relationship between a student and their teacher, whether the student feels like the teacher cares about them, looking out for them is one of the most predictive things at both both high school and higher ed for academic success. But yeah, sometimes that feels out of the purview of ed tech of lots of the regular edtech world because we are sort of in parallel often to that human relationship. So when you think about, you know, the quality of teacher student interactions, this is something you really focus on at teachstone. How do you measure it and how do you make sure that people are incorporating it and really thinking about the quality of relationship as a top order metric for what a classroom is doing? It's not just about grades, it's not just about attendance, it's about actually relationship building. How do you support that philosophy?
- [39:41]
C
Yeah, so we do a lot of, we train and build capacity relative to like being able to observe teachers in practice. You have to get into the classroom and really help drive that. The research behind this and is one big body of research relative to the Gates foundation, the measures of effective teaching. And it, it showed that we call it a class classroom where the class is kind of the foundational kind of framework for the classroom. For students that were in a classroom where the teacher was driving kind of high class scores in a year's time, these students produced three more months of math outcomes compared to folks that were in lower quartile class classrooms. Right. You play that out and we all think about the, you hear all the folks on Wall street, we talk about the effects of compounding that happens in learning as well. Right. You think about being kind of a kindergarten student and you're producing three more months of math outcomes and you accumulate that over nine years. So you go from kindergarten to eighth grade and you've, every year you've produced three more months than a peer group. By the time you get to ninth grade, you have accumulated 27 more months of math, you know, kind of genius. So you show up as a 9th grader, but you're at a 12th grade level, right? Where someone else maybe showing up at night, some folks are actually behind and not showing up as a ninth grade. So that is a huge way in which we think about that. So that research points to that. And then also it's not just academic, we believe. And the research points of this, it's academic, it's social, it's emotional. I'll give you a physiology kind of example. We measure, kind of research measures cortisol levels in students, right? So cortisol is a stress hormone. You think about a kid that goes to the first part of their day and kids are kind of maybe normalized, so to speak. And then you see a cortisol spike in the second half of the day. They might have had a tough night at home and it starts to show up sometime during the day. So Teachers are trained to kind of navigate these moments in a class classroom. We tend to have. It drives the cortisol level down. So you get high emotional support. When you see high emotional support throughout the day, it also adds to that academic learning kind of the second half of the day, the social learning second half of the day, and the emotional support in the second half of the day. So those are other ways when we want to see kind of that research kind of show up in practice to drive the student outcomes.
- [41:51]
A
Yeah, that's really powerful. And it's interesting. You know, we're in this really interesting age of ed tech and AI and everybody's trying to, trying to sort of figure out what the right approach is. And I think that one of the things that's really interesting about what you're, what you're talking about here and what you're doing at teachstone is as you say, you have academic performance as sort of a north Star, but you also have social behavior, emotional stability, you know, cortisol level staying down. And I think, you know, as the sort of vision of, in the edtech field, we think about what the classroom of the future looks like. I don't think we think often enough about the relationships, about how people actually feel, how calm they are, how confident they are, whether they feel like they've, you know, made progress because the teacher is feeling really excited. Even though we know that that's like core to what the educational psychology says. Like how can we as an ed tech field, I think we often are sort of driven to focus on outcomes where say, you know, okay, how many months of learning, how many, what are the standard deviations, what are the positive efficacy results? But how would you recommend we as a, an ed tech field start incorporating some of these other aspects of well being into our tools and into our measurements.
- [42:58]
C
What I love about again, I'm super excited about AI as like one in some ways level in the playing field, but also really for teachers. Teachers have been burdened by a lot of administrative tasks over time. Just to be very clear about that. What I love about AI is if you can use AI to automate or strip out some of the administrative. I know a lot of teachers don't know any of them that love grading, grading papers, grading kind of assignments. If you can automate some of those things, I think that's helpful. There are some that want to copilot relative lesson planning, some really enjoy lesson planning, some want to copilot relative to that. But if you take that out of the equation, you build this Environment where teachers become CEOs of classroom chief Emotional Officer is where I think of it.
- [43:40]
A
I like that.
- [43:41]
C
And that's the title that I give them. Because now you can get from moving, just spending time in front of the chalkboard. Right. Or in front of the smart board, to really helping students work on these things that you don't get in other facets. Right. So how do you think about a teacher helping with that? I gave the playground example. Help them with that. But also executive functioning. A lot of kids who are not. You're not getting executive functioning by spending time on the iPad all day long. And I don't mean that just in school. I mean that at home. A lot of parents have said, hey, look, you know, the kid gets quiet when they waiting to watch YouTube. I struggle with that even in my household. And you want to make sure that you're helping to build these, these executive functioning kind of skills, you know, focus, resilience, etc. I think teachers being able to spend more time on that, helping students get through a problem that they're navigating, that to me is like the combination, the best combination of AI and kind of the human touch of a teacher kind of in the future.
- [44:32]
A
Yeah, I love that phrase. The Chief Emotional Officer that teachers can really. And this is sort of the dream. And I think we all want this to be true and we're working on making it true. That idea that AI and some of the other really powerful ed tech tools that raise efficiency, that raise productivity, that take some of the administrative burden off of the teachers, give them back the time and energy and focus and reduce burnout and allow them to actually, you know, raise the care of their students, observe and catch some of those, you know, the fights or the problems or the social emotional stress. And I hope that we can get there. I really, I believe we can get there, and I hope we can get there. Tell us a little bit about how Teachstone, like if Teachstone sort of knocks it out of the park, if you get all what you need to do, what does a classroom look like that has that level of emotional support and students are feeling resilient, they are feeling, they are raising their cognitive functionality. How would that affect the education system at large? And if teachstone does exactly what it's designed to do?
- [45:30]
C
So our focus, as you kind of picked up, you know, we have teachers at the center to drive outcomes with students. Right. So it's equipping teachers with the right sets of tools to help them move forward. One of those things we're Building one of our tools, Instant Insight is being able to give real time coaching kind of to educators kind of in the moment. You're able to, in a few minutes, grab kind of a clip of yourself in practice, in the room, upload that, and we have a set of tools that help show you, hey, if you want to improve on this area, this is what you would get kind of in return and being able to just do that on a regular. Most teachers are used to having observations out of someone come and observe their classroom once or twice a year, Right. You know, if they're lucky or if they're lucky, unless you're just on a more negative plan, you may get more observations. But in any sense, giving the teachers kind of the tools they need to drive those high powered and high quality interactions with all students. Because sometimes what happens is you have great teachers that still kind of work on the bell curve, right? They're working on. These students are right in the middle, students that are super locked in, not the students that are necessarily kind of on the outliers. Right. And we want to make sure we're loving the play field for all students. So one of the things is just making sure that, you know, for a teacher who even is exceptional in many ways or believes they're exceptional, hey, you missed, you know, Jessica was over there and she had, you kind of shut her down, you know, you didn't give her a chance to follow through with her thought. Right. You know, being able to kind of show that in real time it will actually raise the bar for all students in the classroom and not make, you know, make sure that the Jessicas of the world aren't left feeling discouraged and not wanting to raise their hand again because the teacher doesn't want to call them and doesn't want to probe kind of their ideas. So in our environment, you know, or our kind of ideal state in the future, I don't care what zip code you're in, I don't care who your parents are. I don't care. Every student has high quality teachers and has a high quality interaction with their teacher to promote again, as I said, the academic, the emotional, the social kind of outcomes that we all want to see for our children. So it's very personal to me. I'm a father, right. You know, and I mentor a lot of folks. It's really personal to me to be able to kind of drive this forward again. Don't care where you are, don't. And this is global, right? We have partners that are not. And this isn't just a us thing. This is, I think around the world we're all thinking about how do we move this forward. So we're fortunate to have great partners kind of all over the world to help move the partnership practice forward and also move the research forward.
- [47:50]
A
I love that. So putting together some of the things I'm hearing you say it's really an exciting vision, right? It's like every student, regardless of where they land on the academic bell curve, is getting supported, is getting dedicated attention from their teacher. Teachers are getting real time insight. So every teacher, regardless of their emotional intelligence or regardless of, is getting real time support in a meaningful way. And then people are getting these academic outcomes, these three months a year, so they're getting the equivalent of years of extra academic outcome. I mean, that, that is a pretty exciting vision of how a school could work. And you know, I also hear, I hear so much of like the joy in your style and your voice. And I feel like it's another piece that's key to this. It's like, how can school, even in the era of AI, even in the era of academic outcomes being paramount, be joyful, be fun? I'm curious how you think about student to student interactions and if that's something that is relevant to Teach Zone's approach.
- [48:43]
C
It is spot on. And you kind of hear again, even in the playground example, it's like if you're just glued to a singular device all day and you're getting kind of the interaction, the back and forth volley with your co pilot as a student, that is great, right? It's going to produce some academic outcomes. It'll kind of move with you. But this is really important. Your network is important, right? You know, and I think teachers showing up and I'll tie this all together. Teachers showing up as mentors and coaches are going to be important and they're going to be able to kind of show why these networks matter, right? If you think about a lot of work kind of up and down the education continuum from early childhood to workforce, you know, development. Alex, and what I would tell you, a lot of kids are struggling even to find like an internship, right? AI may tell you, hey, based on what I know about you, engineering is the right path for you. A teacher or mentor or coach, I think teacher will have those roles kind of codified, kind of in their DNA. They'll be able to say, hey, you should go talk to Alex. You should go meet this person. You should go to meet this person. Because those three people might be able to unlock either an Internship for you, you know, kind of going forward or help you see another path that you didn't really know. Right. So that's a huge element. Or they'll tell you, hey, the two of you both have similar interests. Two students in the same classroom both have a similar interest. You also pair together and one plus one equals three. Right. Relative to those pairings. So I think the teacher is still kind of at the center, but they can help connect those dots, help regulate some of the social and emotional kind of components across the board. But if you go back to AI, the neural networks. Right. We're talking about networks here. Right. And that's a, that's a key ingredient to everything that we're kind of describing here. So I think there's a, there's a huge element of making sure that this is not a one to one relationship. This is a one to few, one to many, and we all kind of move forward and lift one another up. And that's the joy that I get in doing this work.
- [50:33]
A
Yeah, I mean, the idea that interactions between humans, teachers and students, students and students, administrators even and teachers like making the observation feel more joyful and more frequent is such a powerful. I feel like it's a element of the edtech world that, I mean, I do a lot of these interviews and we don't talk about it nearly enough, especially in this sort of AI era. I think you're mentioning going from one to one to one to many or one to few. And that one to one paradigm is something that I think we all naturally sort of fall into because people tend to work on devices. To one, we think, oh, one to one, we're going to personalize education for every individual student and it's going to be one to one. But I think this approach of no, a school is a network. All the students are networked together. The teachers are networked. The teachers and the students are networked. They all can help each other, bring different resources, bring support, bring emotional support. I really like that approach and I think it's really powerful. It's a powerful direction, I think, for our whole field to think about how to develop tools and thinking that doesn't isolate people, doesn't make people feel like they're atomized and each individually sort of having to grapple with everything that happens on their own.
- [51:38]
C
Yeah, you got it. Yeah, you got it. You, you know.
- [51:42]
A
Yeah, no, it's exciting to think about it that way because I feel like in the edtech field sometimes we just end up going that way. And I don't think we always want to go that way, but we end up sort of heading that way because it's just the nature of the connection. But the teacher is the. Is the key to the classroom and the. And the class itself is so connected.
- [51:58]
C
That's right. It can be seductive. I mean, they're like, you know, as a business leader, there are margins and all things you're kind of thinking about, but what's best for the learner, what's best for the teacher. And not just in the immediate. Right over again, compounding what plays out over a period of time. That's why I love that we're focused. Our core focus is early childhood space, like get them while they're young, birth to five, birth to eight, and produce those, stack those outcomes kind of over time. That's where we're trying to launch a bunch of amazing students into some of our best schools, kind of in this nation and other nations as well. But then also helping to think through this at, you know, the secondary level and looking at outcomes kind of post secondary as well.
- [52:34]
A
Yes. Because as you mentioned with the English teacher example from your personal experience of having, you know, when students hit these sort of academic roadblocks, there's this huge confidence issue. And they say, oh, am I not a writer? Am I not a math person? Am I not a student? And if they end up on the wrong side of that, that can really take their entire confidence out of them for many years and even into adulthood. And then flip side. Right. People who have great mentors, people who have teachers that change their lives, they will say that into their careers. They say, I took this career because of my fourth grade teacher who helped me see that I could do this. And, you know, I think we all know that that's the sort of dream of the educational experience. But again, it doesn't feed into ed tech as much as it could, I think, because it's just sort of not where we're focusing our energy. But I love that that's so core to what you're doing at teachstone. If I have.
- [53:20]
C
If I could just share one last please here.
- [53:22]
A
Yeah.
- [53:23]
C
I did a keynote speech at my alma mater, my K through kind of eight middle school. I did a few weeks ago their graduation promotion ceremony for the eighth graders. And one of my guests was my assistant principal. So we've kept in contact over been a couple decades, we've kept in contact, but she's one of the first mentors, Ms. Williams, one of the first mentors that I remember having. Right. And we've just stayed in contact. But it wasn't for her identify me for kind of gifted and talented and Johns Hopkins has a center for talent at youth. It wasn't for her identifying me for those types of programs back then. I don't know that I'd be CEO of teacher today. Right. You know, but it was a huge confidence boost for somebody to say, I see something in you and I want to make sure that we keep you kind of on the straight path. Right. So that is just played out. But I think it's important on the other side of this equation is like respecting kind of admiring kind of our teachers, the folks that have really helped help shape us. Right. And I think a lot of people do that. A lot of people forget their teachers. Right. You know, but I think it's important to remember the ones that really helped move you along kind of in your. In your life.
- [54:20]
A
Yes. And for the EdTech founders and the EdTech operators of the world, it's important to support those teachers, make more of those teachers, allow those teachers to do their best work and change lives in the way that that's why they got into education in the first place. I feel like we want to enable that and empower them to do that more and more and more because that's what actually matters in the world. It's really, it's exciting to think about. I wish we had more time. We will talk more in the future. But this has been so much fun. Carlo Young is CEO of teachstone, focusing on improving classroom quality through research based solutions. Really focusing on human to human interaction. Thank you so much for being here with us on EdTech Insiders.
- [54:57]
C
Alex, thanks so much for having us. It's a lot of fun and look forward to seeing you soon.
- [55:02]
A
We have an amazing guest for The Week in EdTech. This week we are here with the great Tom Sayer, co founder and CEO of Ello, an AI tutor that gives young children the personalized one on one learning usually reserved for the lucky few. Ello began as an AI reading coach and now spans far more for ages 4 to 9. He's a Stanford GSE and GSB alumni. And Tom founded ELLO to close the early learning gap. It's backed by Y Combinator, Goodwater and others. Ella's been amazing. Tom sayer, welcome to EdTech Insiders.
- [55:37]
D
Hey Alex, thanks so much for having me back.
- [55:39]
A
I always enjoy talking to you. You are such a thoughtful person in the space about AI, about tutoring, about reading, and now other subjects as well. Tell us you just launched a product that goes beyond reading support. What problem are you most excited to solve now? What is the expansion of ello's mission to change the equation for learners?
- [56:00]
D
Yeah, it's a very exciting moment for the company. We've kind of been in our cave for the last best part of 18 months now, working on the new version of the product. And we set ourselves the challenge and said, like, given the shifts in technology, let's start by saying, how can we use that to build experiences that are much closer to the magic that happens when a great teacher is working one on one with a child? And so that's what we did. And we've been harnessing that and through that, because we started with those first principles, as you kind of mentioned, it means that we can actually go broader than reading, which is super exciting. Reading is still our primary focus in the first instance, but we can use this technology because a great teacher, a great elementary school teacher can teach reading, but can also teach math and can also teach English language learning. So that's really exciting. But I think the most exciting part is what's actually happening and how that's happening. And so for a lot of kind of adaptive ed tech, it's kind of adapting the content. It's like, what content comes next? What's the next question? What's the next video that gets played? But what we've built is really adapting the teaching itself. And so it's really asking those fine green questions that a great teacher asks. How is this kid doing? What do they need next? What's the next learning activity I should provide? Should I go a little bit quicker? Have they grasped this concept and maybe just to kind of really quickly go under the hood of what that looks like? Obviously there are LLMs at play in here, but there's not just one LLM, it's not just a ChatGPT wrapper at the moment. We've got four running. When you're interacting with hello. And so one is you kind of imagine one is the teacher that sat next to the child thinking about, how's this conversation going? What should I do next? What's the next learning activity? The other is kind of like the master teacher that sat next to 10ft away observing everything and then course correcting, like letting the teacher know, like, oh, I think actually this child may have missed this concept. You might want to revisit. The third is the curriculum planner who's looking out over the timescale of days and weeks of saying, exactly where is this child? Given the learning objectives we have and what do we need to do tomorrow in order to maximize their learning? And then the fourth is the safety system looking at every single income interaction and saying, is this appropriate? Is there anything we need to be doing here? And so it's really the second by second judgment of a great teacher. And it's super excited because I know Bloom studies getting a bit of heat recently, but one on one, teaching is still the most powerful thing we can do for a child. And it's, as we all know, it's expensive, it's not scalable, it's the reserve of the privileged few. And so I think if we get this right, we really manages to give access to that high quality experience to many more children.
- [58:44]
A
That's incredible. And as my instructional design mind is getting very tickled and very excited by this concept of embedding, you know, what some people call pedagogical content knowledge, or obviously, you know, teacher moves teacher training into a tutoring model so that it's not just about planning curriculum, it's not just about delivering questions or adapting feedback, even though those are all very powerful. It's actually acting like a teacher and a teacher coach at the same time. That is really exciting to hear. So this is happening in a really interesting context. And you know, I have to bring this up because it's just so top of mind for me and I think for a lot of our listeners, which is we are in the middle of this real confusion, I think in schools and in families where they're trying to figure out what role AI plays in education, whether it's a positive role or a negative role or how it can be optimized. And it feels like you've been so at the forefront of this, of thinking about how you can ensure with product design, with pedagogical design, that it is a positive role. How have you responded and how do you feel about this moment in edtech? And how do you think Ello can sort of change the perception or continue to change the perception that AI can be a really positive force for learning?
- [59:54]
D
Yeah, it's such an interesting question, an interesting time that we live in and obviously ASU GSV this year, it's top of mind for everybody. And I think it's. The backlash is somewhat justified. I think what's happening though is we're bundling a bunch of things together and we risk throwing the baby out with the bathwater. And so there's the, obviously lots of pretty awful things that have happened because of social media to young, young people's minds bundled with the Somewhat, maybe oversold and under proven aspects of some technology in the classroom. And then the generalized fear of this scary transformative technology AI, and they've kind of all been smushed together into one. But I do think it's really critical we don't throw the baby out with the bathwater. We need to be very cautious when it comes to AI because of how powerful it is. But at the same time, these kids need to be ready for the world that we're entering. And if we didn't get to use this powerful technology for their benefit, I think we'd just be missing some much. And so I think it's really as all good ed tech practitioners do, it's just really thinking about what those experiences are. Are they passive or are they active? Are they really building learning experiences that push the learner or are they designed to purely maximize engagement? A good test is whether I think the technology's goal is for you to solely spend more time with it or whether it's focused actually on the learning outcomes. And so I think I'm nervous. I don't know what the answer is. I think we have to. The edtech industry has to earn the trust back. Not that it was the edtech industry's fault that.
- [61:32]
A
No, I mean, somewhat, but I think the completion is real.
- [61:37]
D
Yeah. But I think there is an opportunity too with AI, because I think I can create. It gives us the tools to create the learning experiences that show that technology can have a really important role in a kids.
- [61:49]
A
Letting that answer resonates enormously with me. I actually use the exact same metaphor of throwing the baby out with the bathwater. As I was talking about this earlier this week. You know, in the UK they just announced over the last two weeks two interesting pieces of upcoming legislation around teenagers. One is a huge social media ban for students under 16, sort of similar to what they did in Australia. And the other is an expansion of AI tutoring, especially for underserved students. And I found that to be a really interesting combination because it tries to go against exactly that kind of conflation you're naming. Right. It's like social media. We have a lot of evidence of the negative effects of it. So let's use the power of the state to try to protect students from it. Education technology is not the same as social media. Yes, there have definitely been malpractices. There are tools that are designed for engagement. They're designed to keep students going through game after game after game. And they may not have the proof that it's worth that it's actually going to have the outcomes. But as a field, I think conflating education technology and social media and sort of screen time at large is a huge mistake. One thing that is so exciting about what I'm hearing you say about the new ELLO system is you obviously have safety just baked into it. I mean, not only are you thinking about pedagogical design and curriculum design, but you say there's a whole model just designed all around safety, privacy protection. Tell us why you're thinking like that and what you think other edtech companies have to do to earn that trust back in terms of safety.
- [63:13]
D
Yeah, so across the board, so we allo is for kids kind of age 4 to 9. So across the board we've really tried to take on the hardest challenges when it comes to AI for kids from how they speak. Like a 4 year old is much harder to understand than a 12 year old and it has to be through voice. And so our speech technology that we the powered the old version of Allo is critical again. But safety is one of those, like when you're dealing with kids as young as 4, like safety is critical. And if we want to prove to parents and to schools and to society that this technology can be transformative, we have to earn that trust. Parents should be super cautious around this technology because we, I don't know, we could spend the whole time talking about social media. We should learn lessons from social media. It's like we want parents to be cautious and thoughtful and it's our job to earn their trust. And so we take safety so seriously. The principles that we've always applied to privacy I think apply here. Transparency and control were always our two pillars, like don't want to hide anything. Parents should know exactly what's going on and they should be able to control it as much as possible. They should have a say in how things operate. We're applying that here as well. But I think in terms of the safety system, yeah, we have. My co founder Elizabeth is a clinical child psychologist. She, she thinks so deeply about this, about what is safe and appropriate. Obviously we do the thing like our AI tells the child that it's not real. We have a human in the loop system so that if things get flagged we can look at them and say like what's going on here? If we need to, we can alert the parent if there are issues. So we're trying to give insight to the parents and transparency to the parents. But maybe just to give an anecdote of something we saw a couple of weeks ago, which I think really brings in to life. A child was talking to the ello and said something along the lines of like, yeah, the kids at school make me feel really stupid. It wasn't like the, the scariest thing, but it was a clearly a deeply emotional moment for that child. And what the ello did was, was fantastic. First of all, it acknowledged the emotion. So if you're dealing with old school technology, you can cut it off or you can keep going. But the ello was like, that sounds really hard. And it acknowledged the emotion to meet the child where they are. But then it immediately went to, but I'm not real, I'm a computer, so you can't talk to me about that. And so constantly reinforcing the difference between a human and technology, which is such a critical skill that kids are going to have to learn. And then it redirected, it said, find an adult that you can talk to about this. Maybe it's your parent, maybe it's a teacher. And so it didn't want to just leave the child in that moment and say, well, no, not talking about that, that's, that's not my job. I'm here to teach you how to read, or I'm here to teach you math. It actually kind of guided the child through that moment to a more positive outcome. And so I think that's also, that's the exciting thing about the technology. We can see safety not just as a harm to be protected against, it has to be, but also as an opportunity to support the whole child 100%.
- [66:14]
A
And you know, it strikes me, as I hear you talk about that story and about safety and in the context of ello and these sort of four models you're naming, when we think about what a teacher provides to a student, an elementary school teacher, yes, they teach reading. Yes, they teach other subjects as well. Yes, they design worksheets and activities and curriculum and hands on activities and projects. Yes, they also protect the child emotionally, they respond emotionally to it, and they alert parents when the child is having any kind of issue that needs to be addressed. And it's like, it feels like increasingly the education technology products have to sort of recognize that as we're supplementing the teacher with some really powerful tutoring techniques, it has to understand some of the many roles that a teacher plays, some of the many skills that a teacher has, and not just sort of pick and choose, like you said, not just sort of leave a child hanging or just provide one thing and then not be there. Because this is a conversational technology, it becomes closer to having to be adaptive and respond in real time. And I think this is a real opportunity. I think some people are scared of that. And you can understand, especially for elementary students, you understand that that sounds scary, but when it's done thoughtfully and it's done really carefully, it is incredibly powerful. And it's an opportunity to expand the scope of what educational technology can take on and support children through. It's a wild moment.
- [67:38]
D
It is. I think we do need to be really careful or clear, though, because I think that's what's so exciting is that technology can do more. It concentrates the responsibility of the teacher into the things that only they can do.
- [67:51]
A
Right.
- [67:51]
D
And so making sure we're super clear about that line and then empowering the teacher for that stuff that we don't want technology to be doing and that technology can't do, I think that's what's exciting. It's like. And the challenge is figuring out those boundaries 100%.
- [68:06]
A
Yeah. I realize, as I'm saying that it almost feels like it's an argument for, hey, I can do all the things a teacher could do. That's not what I'm trying to imply at all. It's just trying to say that it sort of has to recognize that a teacher has such a wide variety of skills that they're offering a student at any given time. And a technology has to sort of understand how to handle scenarios and then when to bring in teachers, when to bring in parents, when to redirect, when to be age appropriate or safe or. You know, there's so many things that have to be packed into these AI ed tech tools that we have to use those responsibilities to really earn trust. And I think Ella's been amazing at that.
- [68:42]
D
Yeah, I love that because it just brought something to mind. And it's like the technology has to be aware of the entire ecosystem that the child sits in. And the teacher part, like, just like a teacher needs to know, hey, I've got my school counselor here. The teacher needs to know their boundaries of, like, here's where I need to take it to. And then I need to know exactly the counselor's capabilities and the role that they play in this ecosystem. And so it's almost part of a. An infrastructural team around the child. And the technology needs to know its place.
- [69:12]
A
Exactly. The technology needs to know its place and its role within the team. I think that's a fantastic way to put it. You know, one thing that stands out in your bio is that Ello is designed to be personalized, one on one learning usually reserved for the lucky few. Right. I mean, tutoring, reading coaches, home tutoring is an expensive proposition. And I think that part of the excitement about AI tutors is that it allows equity and access for many families that wouldn't have had it otherwise. I'd love to hear you talk about what that looks like through the ELLA lens. We talk about the 5% problem in ed tech, about, you know, only 5% of students are using technology the way it's sort of designed to be used, or that through the dosage or through the engagement. And many others are left behind. How are you making sure you don't fall into the 5% and that Ella really fulfills that goal of providing what would, you know, really, really high quality personalized tutoring to students even if they come from underprivileged backgrounds or even if they don't come in with a ton of cultural literacy or engagement. Like how, how is Ello meet every student where they are?
- [70:12]
D
Yeah, it's a critical again to earning back trust. It's a critical challenge to think about. So our mission at Ello is to maximize every child's potential, regardless of resources. That hasn't changed. So we've been around since before ChatGPT, that same mission word for word since the beginning. But we're in an incredibly privileged position right now. We actually have been funded to provide ELO for free in Africa forever, which given inference cost around AI is no mean feat. And it also funds our ability to scale it in Africa and so not just switch on the button on the app store saying it's there and it's free, but actually really ensure that it works and that it's being used at scale in Africa. Just as a side point that's really meaningful for us. Our first engineer was in Kenya. We have a small office, we've always had a small office in Nairobi. So the fact that we get to do this there and Africa, Africa as a continent is the fastest growing continent. It's going to have a huge role in the world. And so if we can ensure that AI's gains don't just concentrate to the wealthiest, which they absolutely could, and you see SpaceX's IPO and you can believe that AI is just going to continue that trend, feel incredibly privileged that we get to try to harness this, this transformative technology and ensure that actually it puts every child on a much more equal playing field. It has some real consequences for us as a company and design choices that then get forced out about because of that. And we do. Just to quickly mention we are trying to ensure that there is some free version of the product in. There isn't with the current product. The current product we have a very heavily discounted at 10% of the price. If you're on any kind of government assistance in the US we're trying to ensure there's some free version of the product in the US as well to make sure everybody can access what we're building. But yeah, it leads to really deliberate design choices. So we have to build for low powered Android phones as an example, which is another really hard problem that we're solving. But that means that for example, everybody could have access. It's not reliant on having the latest MacBook Pro or whatever you might need and then building the product in a way that accepts that difference in technical literacy and ensures that we are bringing the child through the learning journey and not just kind of letting them drop off at the first instance. Luckily, like that's what the technology we're building is for. And so that's very easy to then apply and to ensure that. So a lot to figure out some really hard problems we have to solve. But the fact that we are funded to be able to do this, to ensure that we are giving this opportunity to the kids that need it most, it's one of the reasons why everybody works so hard at Ello and turns up every day. It's like it's such an ins. It's such a privilege to be able to get to do that and feel very fortunate.
- [73:00]
A
That is truly incredible. And I had not heard that before, but that is unbelievable. And you know, I mean, we hope everybody in ed tech, and I think everybody in social impact around the world has this dream that AI could be a transformative technology in education, in health, healthcare that could be an even, you know, an evening technology that it could actually decrease inequality gaps. It could support continents with huge numbers of young people like Africa. But yes, you also see this funny evidence where AI, you know, the benefits sometimes accrue to the people who are paying a lot of additional money for AI tools or sending kids to camps where they're going to learn AI or learn through AI and the ability to lower the cost to free or 10% and then provided on devices. We saw, you know, Khan Academy famously do a version of Khan that was able to be used on very, very low bandwidth phones. It's incredible that you're in that world because I think it's going to be, it could be transformative.
- [73:57]
C
So.
- [73:57]
A
So let's look towards the future and I'm going to. This is going to be a little bit of a tricky question and I'm curious how you answer it, which is you put on your, your futurist hat, you look at three to five years from now. I'm curious what the sort of dream idealistic pie in the sky vision of what AI tutors could be. And I know you're heading towards that world like what is the sort of beautiful future in three to five years and then what do you actually expect to happen? How close do you think we'll get to that within that time period given all the realities of the world?
- [74:27]
D
Yeah, I think being a futurist in today's day and age is one of the hardest jobs you can have because everything is, everything is changing so quickly and anybody who claims they know what's going to happen I think is deceiving themselves.
- [74:40]
A
We're recording this for posterity or anything.
- [74:43]
C
Don't worry about it.
- [74:44]
D
I mean I think something that I do believe fairly confidently is that every child will have an AI teacher or tutor in five years time. I don't know if it's two years time, I don't know if it's four years time. But certainly by five years time I think the technology is too powerful and too important for learning that that won't be the case.
- [75:03]
C
Yeah.
- [75:03]
D
As we said our job is to make sure that it's not the iPhone version of the AI tutor for the wealthy and the, I don't even know the brands that we're. We're play testing on to make sure it works on the lowest powered devices. Like it needs to be high quality AI tutors for, for every child. So I think that's just going to be the norm. But parents will be picking the AI tutor. There will now be like right now it's the category creation moment. But in five years time it's like do I want the safest one? Do I want the one that promises the quickest results? Like there's going to be, there's going to be multiple AI tutors and you're going to be able to pick which is good. The one that best suits exactly how you want to think about your child's education. A couple of capabilities maybe of AI tutors that I think are really exciting. The first is that is memory is being able to really understand a child. Like my philosophical learning background is all around learner centered education and how do we truly center that specific child in the learning Experience and the ability for technology to really understand the child because it understands every interaction. Like a great teacher can do a reading assessment twice a year, given time constraints. Even with our current version of ello, we're doing formative assessment moment by moment. Understanding how that child is reading and the ability to connect that across subjects, across different modalities, is really exciting. Risk. As soon as you start building a picture of a child, you've got to really be careful around bias that's creeping into that picture and privacy and all those kind of things, but also really powerful. The other is, I think, much more multimodal, where we're talking about screens, backlash, and Ello is on a screen. But I think as we're going to see it, like, I barely type anymore. I just talk to my computer all the time. I think the multimodal aspect of learning, which is so powerful in the classroom, will come to an AI tutor as well. And then I think that the risk is kind of what we've been saying. We don't want to replace the human in this system. And so it concentrates the importance of the role of the teacher. My idealistic view is I want every teacher to have. They've got a class of 30 kids. They have 30 of the most capable teachers, assistants working for them. That feels amazing. And then the teacher is there to do what only the teacher can do. You can still have a 1 to 30 classroom, but you've got just so much more. We ask way too much of our teachers in terms of the pressure, the time pressure we put on them just gives them so much more capability. And so I think, how do we keep parents central to learning with these AI tutors being about in the home environment? How do we keep teachers central to the child's experience in the classroom, but then having these really powerful teachers, assistants, tutors that can support the human adults
- [77:51]
A
in their critical job, Fantastic. So part of my interpretation, and it's an amazing answer, and part of what I'm sort of hearing you say is you feel like in three to five years, AI tutoring will be widespread. It may be ubiquitous. Every child will have an AI tutor. The question is just whether we as a field can make sure that the tutors that students have are equitable, are incredibly powerful, are safe, are private, are age appropriate, are thoughtful, and play their role in relationship to teachers and don't seek to replace them. That's sort of the cliff notes version that I'm hearing. And it's a very powerful, I think, vision for where we're all going. It's exciting. And you're right. Category defining moment. We're all in this crazy moment and we've talked to the Khanmigo folks, you've talked to, you know, synthesis, lots and lots of people who are trying to figure this out. And it's really exciting to hear that Ello is expanding beyond reading into multi subjects and thinking so critically and with such an evidence base about how to put the pieces together. Tom Sayer is the CEO and founder of Ello AI Tutor, giving young children the personalized one on one learning usually reserved for the lucky few. My oldest son is turning four in two weeks and I know that I know one of the birthday gifts we're going to get for him as well. So thank you so much for being here with us on Tech Insiders. Thanks Alex. This season of EdTech Insiders is brought to you by Cooley LLP. Cooley is the go to law firm for education and ed tech 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.