00:01.47

Olivia Lara-Gresty

Hi everyone, I'm Olivia Lara-Gresty and I'm here with Dan Cogan-Drew, Co-founder and Chief Academic Officer at Newela, where he and his team are bridging gaps in K12 education in part by providing articles from reliable sources that are adjusted for different reading levels, which empowers teachers to engage students while also meeting diverse literacy needs. I know as a teacher I used this product all the time and so I'm really excited to be talking with you, Dan.

00:25.91

Dan Cogan-Drew

Thanks. It's great to be here Olivia

00:28.63

Olivia Lara-Gresty

Well, I would love to just kind of get started and, um you know, share a little bit more about yourself and and about Newsela. So I'm curious, kind of looking back, I know Newsela filled a really important content gap for K-12 schools. And so I wanted to just, you know, mention today we're seeing a huge explosion of those, of just edtech schools in general. So curious how your perspective on the role of technology in education may have evolved since you've actually co-founded Newsela.

00:55.88

Dan Cogan-Drew

Yeah, I mean, I think it's a really interesting question because I think in some ways it has, and in some ways, it's sort of some of my earliest notions that have been reinforced. So prior to Newsela, my job was basically to review products, hardware, software, and different learning models. It's sort of the early days of flipped classrooms and when Chromebooks were making their entry. And I think one of the things that was always, I was always mindful of when I was looking at the technology and its use in the classroom was um whose job is it to do what? You know when the technology enters the the learning space in whatever context, and I was looking at products in elementary, middle, and high, across subject areas, mostly in math and literacy, um you know what does the product think its job is? where does the what's the water's edge in terms of like where it thinks that's the teacher's job? um and naturally, the teacher comes into this sort of like teacher-student software or hardware technology with their own notion of that, and the student comes in with their own.

And usually, but for the most part, the student is the receiver of this. They're told, like, your job is to do X and when these things come in conflict, it's pretty jarring for the teacher when the software starts to do something that the teacher feels is their job, or conversely, doesn't do something that the teacher feels is the software's job. And so some like very typical examples of this are sort of playlists. I remember looking at two math products in elementary grades back in like 2013 - 2012 even and one of the challenges in working with the teachers who were, let's say third, second, third, fourth grade, working with their students is they would get on the sort of playlist of the math product and they would sort of disappear into their own world. And by intent, that was what the software wanted them to do and move at their own pace and so on, personalized learning. And then the teacher was trying to figure out like, okay, how does that intersect with what we're doing as a class, like the 20 of us in a more social learning environment that's not one-to-one.

And sometimes it would you know intersect in joyful ways. Like there would be a teacher would ask a student, the student would say something sitting on the rug and they would say something that the teacher hadn't introduced and the teacher around a concept they hadn't gotten to yet. And the teacher would say, well, how do you know this? Like that's the sort of like look of surprise.

And, oh, well, you know, in this case, Gigi told me the penguin and, and the teacher was like, oh, and they take that very high regard. Like, wow, there's really something powerful here when that transfer jumps from one-to-one computer-based to this setting because now other students are hearing the students say this and seeing the teacher's reaction and so on.

But in other cases, like an example from Newsela's, I resisted very strongly for a long time. As you say, we published authentic materials at multiple reading levels. And one of the things that I resisted is the idea that we would lock a student at a reading level the way other products that I had seen would do because it seemed to me that the product was operating beyond its remit and sort of saying, I know that Dan is a seventh-grade level reader and everything he sees is going to be at a seventh-grade level.

One, that that wasn't backed up by research. You're a different reader for many things that you read. And two, it was asserting understanding of a context in the classroom that it just could never have had. Like It was something that um the teacher could know and perhaps the student metacognitively could learn what level reader they were over time and take that and work with it. But there was like, to me, a sort of a jarring experience where if this the software is asserting this and the student is like, no, this is not, I'm not seventh grade on this.

I've been reading about baseball or gardening for, you know, five years. I'm obsessed with it. It sort of breaks the spell of what the software is trying to do. And so I guess this is a long answer, but to say, like I see that happening again. And that was one of my earliest impressions. and I see that happening. We can talk about ways that that's happening. But my perspective on the role of the technology in some ways is very similar to where it was like 10 or 15 years ago, though the technology has advanced incredibly since then.

05:36.27

Olivia Lara-Gresty

Yeah, that's interesting. So almost like, you know, we're reverting back in certain ways to, you know, there are so many products that I think, to your point, I loved how you mentioned, like, what does the product think its job is? And I think that's a really important perspective for teams out there to be considering. And something you mentioned about, you know, I like how you have not locked students into certain reading levels. I think about, you know, so much of, for students, a lot more than I think we previously understood has to do with how a student perceives themselves. So if they view themselves as a student who, you know, we talked about, we had students who were in the red in school, where they had their homework grades were in the red. And so then students start to perceive themselves as a student who's in the red. They don't really view themselves, or it's harder to, I think, break out of that. So I think to your point, you know, giving students that ability to view these levels while they are, there's scaffolding that's important. They are not necessarily locked in. And I love how, especially, you know, when you get that kid who doesn't want to be in school and then all of a sudden you find that one topic they can be an expert in, it completely changes their role in the classroom. They're no longer the kid who is you know failing every test, but they're the one who's teaching everyone about dinosaurs or about roller coasters or something really unique. So I love how you're kind of allowing students to tap into that subject matter expertise that they can bring

07:01.86

Dan Cogan-Drew

Yeah, I mean I think that's super important. It's a sense of belonging and like, you know, what do I bring? Who am I in school? What do I bring to this? And how does the school recognize me as an individual? How does my teacher recognize me? How do I become known? What's my reputation, so to speak, um academically or otherwise, just like who am I in this classroom? And just to go back to that example, like I also want to sort of land the point here that um we've learned, too, that we can't just, ah it's not enough.

We're not doing enough of the job if we say to a teacher, well, you know you give the article and Newsela will offer it at the level we think that's right for them, but they can change it. Teachers don't want that all the time. They're on an assignment-by-assignment basis, again, because they have the context that we don't.

They want the ability to lock a level because they are using that content for a particular purpose. And to go back to the very beginning of this, like that's, I think, fundamentally, like my understanding, as long as we're working in classrooms, my understanding is we have to be hyper-aware of what it means to work in a classroom. And we have to be mindful of the context in which we're being used um and not um you know sort of co-develop that division of responsibility with teacher, student, and us and software.

And for us, I think we see a progression over time. I mentioned metacognition, like the student becomes aware of how they think, where they're strong, where they're weak, where they need to work harder. And they own that. They own their identity as a learner over the years in school um so they can begin to make those choices for themselves. But it's a scaffolded process and the teacher is essential. You would misuse Newsela if you just gave it to your kids and said, go do Newsela. I'll be over there. You go do Newsela. independent reading is great, but that can't be the only way that you use it because that's not going to be effective.

09:12.22

Olivia Lara-Gresty

Yeah, I think that that's you know a really good point just to to consider that. And I think, you know, I want to think about kind of continue to dive into that student experience and that in the teacher's role specifically and how they're choosing products. And I'm curious, you know, how does this perspective um, and your kind of role in supporting teachers and creating and helping students meet their, form their identity as a learner, how does that inform the way that you approach the product development, either of new products, you know, within Newsela or the overall strategy of the whole kind of suite of tools, especially at a time when teachers are so overwhelmed by technology choices. I know for that, you know, there are so many tools out there, but there's only so many a teacher can learn and in um embrace. So how do you kind of think about that when you're considering product development?

10:07.03

Dan Cogan-Drew

Yeah. Yeah, I think it's it's maybe an extension of that sort of shared roles and the context question is that you've got you got to integrate and you've got to be thinking integration um all the time. How does this, how is what we are asking the teacher or student to do, or administrator integrate into what they're currently doing? We can't show up as though it's day one of school ever having existed. There's um habits and procedures, processes. Some are preferred and some are just required that we have to integrate with. That goes for technical integration and also workflow integration. So um one practical way that we think about this is, and I often talk about our pieces of content as sort of like free-floating atoms. They're just like out there. There's now, I don't know, 17,000 and counting pieces of content of all shapes and sizes and they coalesce and you can form them into sort of molecules and like have them be topically aligned and be a cluster that's like a text set. And that's all well and good, but really it's most powerful when it's actually aligned to the scope and sequence and the curriculum. Newsela, I'll often say, is not a curriculum. And we're not striving to be one, but we know that curricula are not sufficient unto themselves. There's usually a gap in the content between the relevance, accessibility, and which leads to sort of missed opportunities for engagement. And we think we play a meaningful role there, the way that our educators use us to take the scope and sequence, take the sort of spine of their curriculum, and we will help align our content to that spine. And so the integration is um tighter and not as happenstance and is strategic so that they know the topics where they really want to let their students go deep and they can find content from Newsela that complements that. But we're not there to tell them what's next. We're we're there to make the what's next easier, more engaging, and so on. And I think that's that's true at you know multiple elevations.

So there's in each classroom, then there is sort of in the abstract across teaching and learning in classrooms. And then I think there's also integration into district strategic planning um and how districts are thinking about what the outcomes are that they're striving for. And I think that's sort of a critical point of alignment and integration um that we have valued over the years and really turned a lot of our energy to be very deliberate in demonstrating like what are the modes of use in which you can expect to get the outcomes in literacy, for example, that you're striving for and you tell us your strategic plan and we'll make sure you understand how to use us, what good looks like, with what frequency and how would you know walking into a classroom and then what's the data going to show and how is that going to roll up year over year?

I think especially because they're overwhelmed by technology choices, not just teachers, but admins, especially with the expiration of ESSER funds. They have to make hard choices and the customers we're working with get that and we're very deliberate about laying out that integration for them.

13:37.06

Olivia Lara-Gresty

Yeah, I think it goes back to, you know, a product knowing what its job is, and think I love how you said that, you know you know, understanding that you're not the curriculum, but that you do fill an important gap. I know, you know, as a science teacher, when I was trying to prepare lessons for both a 12 to 1 to 1 classroom and an ICT classroom for the same day, there were some topics that, you know, some kids got that others didn't, or just, you know, a lesson I need to revisit where, while I may have had a robust curriculum, we needed a review day. And so Newsella was a really useful tool for me to just kind of, um you know, revisit a topic or just create more engagement or excitement around topics that I wanted to to dive into. And I know that a while back I was listening to something you mentioned that when you were first starting out, correct me if I'm wrong, but I believe that you were actually directly emailing teachers that you were working with who were signing up just to ask them their opinions. And I think that um from what I've seen, that really shows in how you're kind of fitting into a teacher's, the variety of teacher's schedules, but what their day-to-day looks like.

14:39.28

Dan Cogan-Drew

Yeah, yeah, for sure. We had this email that we would all get um that would give us a list of all the teachers who'd registered that day. And it would come out at midnight. And so I would be on my computer waiting for the email. And then I would send an email to any any any, if there were two teachers from the same school, two or more, I would send them an email. And I remember people writing and they would write back because, you know, I think it was a kinder, gentler time when people didn't get so many freaking emails, they would actually respond. They still do. But ah there was something really sweet about the interactions that I could I could form with teachers there because we would really just be writing about, you know, what tell me what you're up to and what you're doing. And I stayed in touch with a lot of those teachers over the years some of the earliest points of contact that I have um which is harder now as you mentioned there's just so much going on.

15:38.64

Olivia Lara-Gresty

Definitely. And I think it's really you know important that it sounds like as you're understanding the teacher's role more, you can then form the product's role to, I wouldn't say, I would say to to fit and complement that. And I wanted to, I think, bring our conversation a little bit, you know a lot of what we're thinking about and what I've spent my time thinking about, especially is really what the role is for a teacher in this age of AI and how products can really support that teacher in that space. And so I'm curious, um you know, as we're kind of thinking about like a future-forward, um how do you think that you ensure that AI is being used on the Newsela team to genuinely transform learning rather than just being this buzzword or a tool that schools might purchase, you know, thinking on all the things that they're considering with ESSER funds, but then might not be able to actually implement effectively.

16:32.48

Dan Cogan-Drew

Yeah, um I think there's definitely a tension between like the, I think the pressure everybody feels to understand and keep up with, keep their understanding current with what is AI, where is it headed, what can it do today that it couldn't do to yesterday. And so, um yeah, we have to be careful. And then I think this gets to some of like the first question is, how has my thinking changed and maybe stayed the same with regard to tech and in classrooms? I'm still wary, as I was years ago, of the sort of tech-forward or tech first um point of entry, like this is in the classroom because it can do a thing and not necessarily because we have a problem that we're trying to solve or we understand what the gap is.

But we have a new technology and it's and now it's here because it's everywhere. It also belongs in our classroom. And I think that can be just, it's not necessarily a problem straight up, but it creates all kinds of distractions at a time when you really need to focus on student learning and outcomes um where things aren't well defined. I think some of the the ways to think about how to use AI is, I would say first and foremost, like don't rush. right like um Identify specific use cases and see for yourself how well the AI can fulfill those. If we're talking about like AI and core instruction, if you wanted to do like AI as a club or AI as like a sort of supplement and in a dedicated time and space where you don't need precious minutes for direct instruction or for learning or for you know um core subject areas, then that's a great way to experiment.

But um I think you kind of need you need a place where you can incubate and then you need to be very judicious about how you put it in play within instruction. One study that I read a little while ago that I hope, though, is the first of what I hope will be more is a way of a series of interviews over the course of a year that a group from the University of Michigan did with, I think it was two dozen teachers, just comparing how the AI in the hands of these teachers who were were relatively new to its use, they contrasted the idea of seeking input versus seeking output from the AI. And ah they found that seeking input, that is asking the AI to generate different alternatives and alternative approaches to how could I teach something, how could I assess something versus seeking output like, can you make 50 of these if I give you the first couple? The input was more effective, more efficacious for the teachers, their own self-reported productivity. They felt they were um sort of more productive. They could think more effectively through some of the instructional challenges they were having I think that is an interesting idea to use the AI in that case where it's like a discussant an interlocutor or someone you can something you can have a dialogue with and ask questions and get it to tell you some of its thinking around a topic that you give it, um as opposed think more effectively through some of to like sort of using it as a sophisticated photocopier.

It can do that as a time saver, um but it may not be the most um impactful for you know the amount of time ah you spend on it. So um I think it's with most things, you have to kind of have a good sense of what your use case is. What is it that you're trying to solve? If you're going to use it in the classroom instructionally, like in your own time, there's all kinds of, you know, zany things you can do, which are great and playful and so on. But I think you just have to be careful that in, you know, in core instruction, those minutes are precious. Teachers know this, but um as someone and myself who's prone to distraction, maybe this is just me talking to myself, but like I got to differentiate like serious work from playtime and not let the two get too intermingled in a way that just like doesn't result in much.

21:09.92

Olivia Lara-Gresty

Yeah, I think that's great, I'm also prone to distractions. I think just even, um you know, providing that framework for how we we do things that it's not just, you know, there is, I think the best advice when it came to all these tools was kind of to play around with them. And while I still agree with that, that there's nothing better than just getting your hands dirty and trying it out, but having a framework and a mindset for how you're doing that so that it's a little bit more intentional sounds really, really useful. And I think that so some of that, you know, does come down to really being thinking critically and kind of doing, you know, the what you're trying to get out of these um these tools, asking the right questions. And I think that that's something that many schools are actually concerned about specifically is that you know, AI might replace critical thinking or creativity, especially when it comes to writing. And so I'm curious if you could speak to, you know, what you're, you think the responsibility of edtech companies is in this place to see that AI really augments, rather diminishes these skills and in our students.

22:15.24

Dan Cogan-Drew

Yeah, I think it's a super interesting question. I think um yeah creativity, it sort of depends on how you want to think about creativity. Let's just say, let's take something. We've been working on a writing product called Newsela Writing that uses AI. And I think one of the things that um where we're mindful of and aware of, and I'm a former high school English teacher, is that um students can struggle with, let's just say, argumentative writing, making an argument and supporting it with evidence.

In the next generation science standards, the claim evidence reasoning frame, um or in social studies, the document-based questions, um you need to find evidence for your argument and you need to explain how the evidence relates. And it just that basic um structure is enough for students to really have some challenges with and you want creativity, but you know it's like I can't teach you to um to improvise if you don't know scales. It may sound like the musician is playing all kinds of notes in random order, but that's not without a great deal of training, right? Jackson Pollock was um didn't just start splattering a canvas. He is is a very talented, and more traditional artist before he went much more experimental. And I think as a teacher, one of the things I'm probably guilty of is like overemphasizing. First, I would say having not been taught how to teach writing, but having been taught to write, but not how to teach it, I had to make up how I was going to teach writing. And I applied my own sensibilities as to the ones that had been applied to my writing in graduate school and as an undergrad and in high school.

And a lot of it came down to things that were very hard to define um and they weren't well-structured. And when I tried to communicate these to ah my students, I was talking about things like how a sentence sounded and what is the voice. And for a student who's looking for, just give me the, get me let me get my bearings, AI and a kind of a set of rules that I can enforce fairly routinely, it's in a sustainable, scalable way can do a lot of great work. It's not going to average writers into artists in writing. But you know in our way of thinking, like what if every student could get a B? They could write B-level papers and the teacher could take them from a B to an A in a way that the AI can't and I'm not talking about creative writing, just an argumentative literary analysis, writing from a source.

That's kind of our approach there. So we don't want to replace it, but we know that students need a lot of at-bats, right? Like one of the things that I learned very early was that if the only writing that my students did was the writing that I read, there was no way they were going to write enough. So that put me in a bind where I couldn't read everything they were going to write, but they needed to write more. And I felt guilty that I couldn't, but that was the fact of it. And they needed timely, specific feedback. And at the time, there really wasn't a good mechanism for this. But I think that's the AI. The value of the AI is with some constraints, some rules, it can apply those rules more and more intelligently. And if you can get students to a point where they start to build confidence in themselves, when you assign writing, they feel like, okay, I know what this is about. I have a structure. I have a schema in my mind of how I'm going to attack this because I've had a lot of at-bats. There's a structure, a sort of a strike zone for me as a pitcher that I can picture, I can see it, even if it's not actually in front of me.

Then you've unlocked writing you know in a way that like was never never possible before. um And yeah, there's always room to grow and to stretch. And that's, I think, speaking going back to you know whose job is it, I think that's really the teacher's job more than it is the AI. And ultimately, it's got to be the students. right But even you know writers have editors, so we know it's not just a solo endeavor.

26:33.73

Olivia Lara-Gresty

Yeah, and I like how you um are speaking to the student's mindset, that that's something that does, we so focus so much on skills, and I think that that is incredibly important, but a student's mindset when they, you know sit down in front of a piece of paper or an assignment or whatever it is and what they go into is so impactful I think that that really does speak to just improving their kind of emotional attitudes through actual instruction or instructional support is really important. I think that, you know, speaking of ah places where students can feel stressed, I think that you know, assessments are definitely a place where I saw a lot of testing anxiety from some of my students, even who are strong performers. And we thinking about kind of the best ways to really understand where a student is in their performance levels. And so I think that where we see a lot of this when it comes to instruction is with personalized learning and thinking about, you know, many different products that are trying to leverage that.

But I think that there's also the argument that assessments haven't really caught up to this. And so I'm curious how you believe that assessments really need to evolve to actually reflect kind of personalized learning outcomes and maybe any obstacles you see holding that back.

27:53.74

Dan Cogan-Drew

Yeah, I mean, it's a great question. I think there's a lot at stake. I think assessments, increasingly, I think assessments are kind of where it's at. And they deserve a great deal of attention because you kind of get what you measure. And I think assessments are holding us back. So, like to date, what I've seen in terms of software and products, digital products, is that they've worked successfully in three areas um and for similar reasons.

So in early literacy, the so-called learning to read math and and language acquisition, like world languages. So like you know Duolingo or any of the the math products, let's just say Khan Academy um initially and like some of the early literacy products. Each of these, it's possible to sort of map the domain. Khan Academy for a long time, they did away with it, which broke my heart, but they had a Google map of math. And I'm not like a math expert. I wasn't a math instructor, but I loved being able to see a domain mapped in terms of a sort of spider web of dependencies to help the learner understand where they were relative to the domain.

You're here. And if you're having trouble with this concept, we're going to push you to go in this direction because we think building the background for where you are now comes from, you know, if you don't know multiplication, I'm making this up, you don't know multiplication, time for more addition practice. I have no idea if that's pedagogically sound, but just for example, like that was great for me to be able to think of like a domain, to hold that image of a domain in my head because background knowledge in non-math areas is such a, ah you know, it's like five-dimensional chess.

It's very hard to get a clear map of but when you when you can see in like things like language acquisition, products like Duolingo, you can see a progression. And if you if you are on a progression and it's mapped, you can code against that. You can write rules, if-then statements, and know what to show a learner when. And that comes down to assessment, right? You need to know where they are. So you can constantly just like assess them constantly from all different perspectives um in those three domains, early learning, math, and world language.

And software in other areas has not been as successful. And I think in part, that's because the assessment challenge has been the burden. just You couldn't fulfill it because you had a limited way. You'd only assess effectively in a limited number of domains and a limited number of methods. And obviously for Duolingo, not being able to fully, like you know languages are meant to be spoken. And if you can't assess spoken language, um then you're missing a whole dimension. And I know they're working on that. I think this is sort of like the bleeding edge here. If we want to fulfill some of the promise of personalized learning is to to embrace a more comprehensive set of tools to enable um the the same level of sort of comprehensiveness that teachers have when they're trying to fully assess a student.

If they really want to get the bottom to the bottom of what Dan knows, they're going to do much more than multiple choice to find ah find the answer to that. um Teachers do all these you know oral presentations, discussions, writing, of course, and other things, but they're not sustainable to at scale. You know you can't do it as often as you want for every student as much as you would want. And so there's some power in technology where I think the evolution in that direction with um being able to automate trustworthy measures of multiple forms of expression, that's pretty exciting.

One of the stats that comes to mind in this context is a company called TeachFX, who in agreement with educators, listen to voices in classrooms. And they published this stat that has stuck with me ever since I read it, is that after listening to however many hundreds of thousands of hours of classroom instruction, they could tell that for the average student, in the average hour of teacher-led instruction, the student-only heard their own voice seven seconds per hour. And if we think that students speaking is a form of you know synthesizing and expressing knowledge that we want to hear, seven seconds, like not to mention the students' disengagement with learning, but like there's just not after the how are you going to do that?

So writing is great, right? Because everyone can write at the same time. But the technologies that could help us rethink assessment I think are just arriving

32:47.94

Olivia Lara-Gresty

Yeah, I mean, yeah that's a great point. I think, um you know, going back to a couple of things you mentioned, ah you know, this is not a math test. So it's speaking of assessments, so well we can glaze over any kind of mathematical dependencies. But I do think that I love that you're kind of thought here about um helping students kind of understand where they are within their their learning experience. I know that you know, even when I was when I was teaching science specifically, students always questioned me when they had a lot to read. And they always said this is science class, you know, why are we reading? And so we talked a lot about that reading is this, ah you know, a skill that kind of goes across all subjects.

And so just giving them that why was so important to their instruction. And I think that when it comes to assessments, students also sometimes have those questions, how is this helping me? And I think that teachers as well, I don't think that, you know, we don't want to spend lots of time assessing students unless there is going to be a benefit to that. And so I think that um to your point, you know, assessments are a bit broken at this point. And there is a lot that can be done to really ensure that students, that they're more effective. And so I'm curious kind of how you and your team are actually rethinking the concept of an assessment, if AI has as a role there, and and how we can kind of go from there.

34:12.22

Dan Cogan-Drew

Yeah, I mean, I think so. The most valuable assessments are formative assessments, and any assessment could be used formatively or summatively. But if you know take the measure of someone's understanding of something and then you just sort of take it and walk away, you're not really giving anything back to them for the time and effort that they spent in trying to demonstrate what they know and can do. You give them a grade, maybe they get some comments, um but it doesn't fold into what they do next. And obviously life is, you know, is made up of summative events.

So, you know, you finish a grade level and maybe things don't carry over to the next grade. But I think we miss a lot of opportunities in in the way that classroom learning is structured today, where the assessments, the summative assessments don't roll over. They don't move. There isn't like an accumulation or an interplay and a carrying forward of, well, what did we learn from the last, what you know where what's missing or what are we going to build on from the last unit or from the last assessment that we did?

So part of the rethinking, if you want to call it that, is just to drive home the incredible value of formative assessment. And there are some great examples of how districts that we're working with are really leveraging the power of formative assessments through basically a very sort of like old fashioned backward mapping is basically, you know, planning backward from what proof would we expect or what would we accept as proof that a student knows and can do what we're asking or what the standard is asking them to to do. And then um what are we going to do? How are we going to get that student to be able to demonstrate this? And then once we've asked them, and once we've assessed them, like, what are we going to do with the students who need more help and the students who are ready for acceleration? And just that taking your end-of-course exam or your end-of-course assessment and just backward mapping it to periodic common formative assessments and common is really important here because you want to have a discussion. Like this is the work of teachers to be very clear, like not the software's work, but the work of teachers is to knowing what they know about their kids and seeing what the data is telling them about their understanding at that point in time in a timely way, like compare notes and compare plans that they might make for individuals or groups of students. The charter school network I used to work at, we had so-called data days, like every six weeks, the schools would literally close for a day. I think it may have wreaked havoc on home life, but they got a day where they were basically like, okay, now what? So on Friday, there was an assessment. Monday is a day-to-day. On Tuesday, that information is being put into use, and we're reorienting our approach for each of our kids and groups of kids.

I think that is, I don't know if it's so much a rethinking, but it's um it can have transformative effects, and it can really enable ownership of assessment at the classroom level and even at the student level but I think right now at least when I was teaching assessment felt to me like something I had to do wasn't necessarily something that I owned and was sometimes in some ways perhaps being used to judge me so it was being mandated from on high. And I think assessments is actually a place where classrooms and district offices actually can meet really productively to the benefit of kids and and parents.

38:00.44

Olivia Lara-Gresty

Yeah, that's that's a great point. You know, again, always understanding the why, what is the purpose of those assessments. And it's interesting to hear about that adjusted school schedule. I have lots of thoughts on how school schedules could be changed or a bit creative, but all would end up being, you know, wreaking some havoc. But I think that that's a really good point to consider. And I think ah supporting teachers in having that ownership over, you know, the assessments that they're giving and why they're they're being put out there. And just kind of circling back a little bit there, you’ve touched on it a bit, but I'm always thinking about all of these tools are great, there's many edtech tools, but how do we choose? And I think in the same vein, there's also just so much data. We're in this information age. And so it's not always actionable. And so just um if you could elaborate a little bit on that as to how we can really support teachers with just the massive amounts of data, any ways that tools can really support teachers with that.

39:00.94

Dan Cogan-Drew

Yeah, I mean, I guess they're it's a big question, but let's say just to take it down to like a micro level, you know, I'm in a classroom, I'm delivering a lesson. I'm trying to, let's say, I have an IWIU structure and there's like something I want to communicate. I want to land a point um and make sure, I mean, this is the key is that make sure everybody understands this. Because like, if you don't understand what a thesis is, I'm about to ask you to write a thesis. And I want to make sure that we're you're ready to do that. Because if I tell you, okay, you know, take a look at this article and tell me, you know, what's your what's your argument? Do you agree with blank? And then I release you to go do that.

And we're not aligned on what I'm asking you to do, then it's going to be chaos and you know disengagement and so on. So what in that moment, like that's formative assessment and that you can do this as with most things, well, some things, let's say AI aside, you can do these things on paper, right? Like there are ways you could do this manually. And a lot of great products have actually been born of um processes that were unscalable, unsustainable, but a human could do.

It was a human that originally was thinking like, I'm going to do this. So like just finding a way to sort of effectively, like what would I take as proof that my students are ready to go write a thesis? It's probably not raise your hand if you know what a thesis is, right?

Like that's much too shallow. Like it's very easy for a student to just raise their hand and be like, I think, you know, everyone else is raising their hand. So that's not enough. um Write me a thesis and I will read it and I will tell you if it's a thesis. That's what I'm about to ask you to do. So I'm not really ready to ask you to do that. I want you to go do that, but I need to know if you're ready. So like, I need to find something in the middle that will give me a sense of like, uh, your readiness for the next thing. And this isn't at the level of like gobs of data. This is just at the level of like in the next three minutes, should I proceed? Or should I reteach? Or should I like, mix it up and have students work in small groups? try to bury this like, there's a, ah I think it was a physics professor, Eric Mazur at Harvard, who did a lot of really innovative work around just how to make a lecture more interactive.

And a lot of it was just like pairing students and having them argue out their answers um so that he could get a sense of where they were in the lecture and their following along in what he was teaching. And he was doing this you know with hundreds of students and eventually, you know created clickers and so on. But it's not a fundamentally technical process or digital process. That's just enabling a pedagogical need.

Olivia Lara-Gresty

Yeah, I've never really questioned why the turn and talk stopped in college, but they're definitely effective and enjoyable, even as an adult when I'm sitting in a meeting or something of that sort. And I definitely think that, um you know, to each of these places, there's so much that that can be done and to kind of support teachers and in um really enabling them to help their students just be engaged and and feel in that sense that they're they're ready. I think some of the best advice I got as a teacher was when, you know, the teacher that I was, one of the teachers I was training under really, really ingrained in me that when I was preparing a lesson, I needed to focus on what mastery sounded like from a student voice.

So I might know the main takeaway of the lesson, but that wasn't necessarily in student vocabulary or even how they might phrase it. And I think when you're mentioning, you know, understanding of a student's ready to write a thesis, it's about knowing what the mastery sounds like at that level, but it's not going to sound the same at the end of the day or the unit or the year. And so just keeping that in mind is really important. And there's so many different um ways that we can support teachers with that. You know, I really, I could talk, I feel like I could be talking to you all day. And so I will try not to do that, but I wanted to just continue looking for a future-forward and kind of think about as you go into thinking about K12 district leaders, many of whom may be hesitant about adopting new technologies, specifically AI technology, technology but just new products in general. I know that you've had quite a bit of experience in advising on some of those choices but um you know in your past, but how can ah K12 district leaders really make for sure that they're moving forward and really kind of start small with some of those technology choices but also aim large.

43:45.94

Dan Cogan-Drew

Yeah, I mean, it's it's a good it's a good question. I think it's sort of an exciting question. um, I think first and foremost, you need to pilot in and incubate in your own context. So um it's not enough just to go to conferences and to hear others and to sort of um to listen. You have to do that, but that's not enough. You have to have some, you have to be experimenting in a deliberate way in your own district in some form. Like New York City had the I-Zone years and years ago. There are ways that even the biggest, whatever you think of New York City DOE, biggest systems can do this. And it could be as simple as like finding a teacher who's in a stable place in their career, has the capacity. They're not experiencing other stuff. They can take on something. um and sitting with them or asking them to come to you with a proposal, like what are some of the biggest assumptions or the hypotheses that you have about the use usefulness of a technology in a given use case? I think this will be good for this purpose. And like, what's the cheapest experiment you could run um that would, if you plot on it like a two by two, like what's what's our level of confidence and what's the consequence if we're wrong? And how can we test the thing that has the highest consequence? Like we're going to release this to 50,000 kids, but we're really unsure about blank. And we really don't know the answer. Like we are, our confidence is low. So you got to test that thing that you think you may at one point release to 50,000 kids. You got to test it ah early in a very cheap way, like the smallest size you can think of. That's the kind of experimental thinking that you need to have happening somewhere in your district um so that you're incubating things and not just trying to bring in whole hog, like like ah transplant an entire ah process from another district that you've heard about at a conference because they've done a lot of work. Every single time they've done a lot of work in their own context. So you need to find a way to do that. And at the same time, you have to recognize that like the incremental steps you're you're taking are in the face of like a tectonic shift um in which AI is going to be smarter than humans like soonish.

So like there's something looming, so you got to get going and you can't be distracted by it, but you have to be just be aware that like the norm here is um a pace of change and a space that I don't think any of us really understand.

46:27.54

Olivia Lara-Gresty

And I think that that's, you know, know, that's the role that leaders can take is being that future forward, but then connecting and supporting the people that are on the ground. And I think that your ideas there really make the case for, um you know, providing additional leadership roles for experienced teachers. I think that that's on a greater level something that can be challenging. Not every teacher, not every amazing teacher wants to be a school administrator. They might really want to stay in the classroom. And so creating additional leadership opportunities for those types of teachers, I think um this is definitely a really good opportunity for that. So I love that you're um encouraging leaders to to lean into that.

47:08.44

Dan Cogan-Drew

Yeah, I think that's key. I think, that like, look to your look to the resources that you have, build on the strengths that you have, and, you know, don't rush to import a technology. it's very hard. I think the best the districts that are best at this have spent a lot of time building resilient, dynamic systems around the people that they have into which they can put technology. But it's often it's often because ah if something fails, it's not because the technology necessarily caused it to fail. It's just because the district is managing change is difficult and they haven't figured out the right methods to do that. And technology is a form of change. So I think this this generally probably applies not just to AI or technology, but to any initiative that you that you want to do, you know, rethink has room to grow to scale across your community.

48:08.52

Olivia Lara-Gresty

Definitely. No, I mean, yeah, technology is definitely um always brings in a lot of change. So I think that's a good place to be really mindful of that. Dan, thank you so much. I really enjoyed this conversation. I think that I've learned a lot and I hope that anyone out there listening also, I'm sure that they've learned a lot as well, but we really appreciate you joining us. Thanks so much.

48:30.13

Dan Cogan-Drew

Thanks, Olivia, really enjoyed it.