Loading summary
A
This session was recorded live at the 2026 ASU GSV summit in San Diego.
B
Our focus for today is around really understanding what is happening and how much good and how much harm might we be doing when it comes to enabling our teachers to bring their best to the work they do. And I'm in Australiania, I am the Chief Transformation Officer with tntp and I have the privilege of moderating the conversation with this very diverse experience group. My invitation to each of you is going to be to just introduce your name and where in the ecosystem of this work you sit so that we can really spend the next 40 minutes understanding your experiences, lessons learned and advice to the room today. Yeah. All right, so my first question, and I'm going to open this up to the panel, is that we've made a lot of promises, you guys, to teachers for decades, right? Smaller classrooms, more tools, more support. AI is just another promise, potentially. Why should anybody believe this time is different?
A
This time is not different for teachers. Teachers already have thousand tools that are coming at them and maybe they find five that are useful. But even though it's not necessarily different for teachers, AI is really going to change our world in a big way. But it's going to happen in five years, not this year or next year. So we have to kind of figure out this balance. And I think teachers are the key. My name is Jason Palmer, I'm a co founder of Socrate. So Crate listens while teachers teach and then does their busy work for them in the background. It's voice powered AI. This is going to. Teachers can be the leaders because they are the ones who actually care about young people and have been caring about young people despite all the chaos that's been thrown at them for decades now.
B
Maybe one more thought.
C
So I am Megan Mitchell and I happen to be an educational practitioner, AKA I'm a sixth grade teacher. And so this is a really difficult question and I say this because there have been a lot of promises made in education and with tools, with scaffolding, with professional learning, with subject matter experts who may or may not have actually been teachers themselves. And I think the issue with a, the issue not just with what's happening with AI now in the classrooms or how certain districts or organizations are pushing out AI. I think what has been missing for, let's say the implementation of Common Core has been putting teachers not as the product, but as the designers, a part of the decision making process, a part of understanding not just what the tool is, but having the understanding of how I'm going to actually implement this. And that goes whether it's for AI, that goes whether it's for like a series of specialized researchers providing professional learning, that goes if it's curriculum, I think it spans the educational spectrum where we're seeing a lot of things promised which are only adding more things to teachers plate and not alleviating what they said that they were going to do.
B
I'm actually glad you said that. So let's, let's build on that a little bit. Jim, Heather, you've both been part of work designing solutions for teachers. What's one or two things you've learned in the process? Maybe things that you know haven't gone right?
D
You want to start? Go ahead. Okay. I'm Jim O'. Neill. I'm the President of HMH's Core and Supplemental business. I think, I think building on what Megan said, designing systems ultimately has resulted in a bit of a patchwork system in our classrooms. And I think part of where I think AI comes in, back to your opening question, is there's this great hope that this is actually going to be the thing that can start to iron out the seams. And what is a pretty fragmented set of tools. Many teachers may navigate across 84 different ed tech solutions in the course of their school year. And that turns teachers into systems integrators, which is, I don't think that's what you signed up to do. Right. I don't think you wanted to teach kids. You didn't want to navigate across different platforms and different tools. And I think that's on us, on the industry, is to iron out those seams and solve for that patchwork. And AI has the capabilities to do that. We have not realized that potential. So it's early to prejudge where we are with I think AI adoption. I think the capacity for change is still huge. The realization of that, I think is still pretty early.
E
Yeah, I think there's a reason for that playing out the way that it does. I'm Heather Annakhini and I run the Chicago Public Education Fund. We do a lot of work with principals and teachers at the systems level in Chicago. I think one challenge for systems is we create guardrails that become blockers. So for teachers that are ready to implement, to use tools in ways that are meaningful in their classrooms, the guardrails we've put up to protect kids, to protect data, to protect the system, stands in the way of really talented teachers. That's one side of it. So then you try to counter that with what becomes, I think the Patchwork were just talking about, well, let's offer lots of different options to people. And in that process, we throw lots of one size, fit all tools at people without recognizing that anything designed to fit all will fit none. Zero. Teachers will find it helpful over time at a systems level, if that's the design of the tool or the program we're rolling out.
B
Jodi, I'm curious, what's on your mind right now? You come from a different perspective, and I think it's worth naming that. And many of your graduates are walking into AI enabled classrooms. What reflections are you having from this?
F
Sure. My name is Jodi Faikema. I'm the Chief Academic Officer and Provost at American College of Education. And we were created by teachers to support teachers. So we have a number of degree programs, but a large percentage is teacher licensure programs. So we are working to prepare students to join the classroom and adding onto what they're saying. Our students are aware of AI. There's no shortage of tools and products. And the biggest gap that we're seeing and where we're really focused is how do we educate our educators to effectively use those tools and not lose sight of the critical thinking components that are necessary to be effective. And I think that is a gap at the moment. Again, there's tools all over, and AI is very popular and everyone wants to use them, but we cannot lose sight of the critical thinking component. So one way, of course, to address that is how do we integrate and leverage AI throughout curriculum. Encourage them to use it. Encourage them to then take something that is AI produced and use critical thinking. Create a lesson plan and then critique it and go through it and take it one level deeper than simply asking a product a question and getting an answer back is taking it that one step further.
B
Thank you. So there's a lot of nuances, right? And those of you who have kind of the engineering background, like, we can break these down into Lego pieces of what design would look like in use. So I want to spend a minute actually talking about the work that each of you are actually doing right now. So, Jason, I'm curious, like, if you can share a little bit about voice AI. It's intimate in a way, for kids that, like dashboards, cannot be. And with that comes risks and rewards. What are you learning about how this tool is being used and the value proposition it brings.
A
Sure. So. So Create was started by a teacher for teachers. My co founder, Maria Anderson, has been a teacher for 20 years. She keeps starting a new company every five years, and then she Goes back to teaching and then she starts a new company. And so one of our core values is teachers first. And this creates some very interesting conversations. So we have 900 teachers using the product now in 45 states. As an impact investor, I've invested in 35 education companies. I know in the long run we're going to get almost all of our revenue from district purchasers. However, when we talk to districts and people at the higher levels, teachers first, it's kind of a jarring conversation. We tell them the teachers are all part of building the product with us. They are the voice to us is exactly what you were talking about. They say, what about administrators? We do let them be part of the process, but they're the second tier. And that's not usual in edtech, usually ed tech. It's more about satisfying the buyer, satisfying the district need, creating the patchwork that you're talking about, Jim. And that's going to make it harder for our company. But it's a core value of Maria and I, and she's done multiple companies that we will always be teachers first. Here's an example. So when a teacher teaches and all this data is collected from his or her classroom, it knows about praises, warnings, classroom management, what has been covered in class. Principal cannot see that. Principal sees aggregate data of all the behavior and things that are going on for a teacher across the whole semester or for a student across multiple classes. But the principal cannot double click into that one class. Teacher could share that class to a coach or to a principal if he or she chooses, but it's not automatically available. Also, there's no recording. How is that possible? You're listening, there should be a recording. Nope, no MP3 file ever gets created. No transcript is available to anyone either. We listen to what the teacher says and then we turn it into structured data and action. So attendance tardies class participation. The teacher will say, the teacher will call on a student. That student will speak. We know that student participated, but it doesn't actually collect even student voice. We've stayed away from student voice altogether too, because that's a whole privacy conundrum. That's a deeper level. And that's because teachers said, you don't need to collect my student voices. I know what's going on. And actually what's interesting is they're now learning. They 75% know what's going on, but there's things they're seeing in the data that they didn't know. And I think in three years, maybe we will collect teacher and student voice. But right now, just teacher voice, just solving teacher needs. Teachers first. If you remember only one thing about so great teachers first,
F
what comes to my mind as you were just sharing that is teachers first. And something else that we see on the higher ed side is they're on a spectrum. So some are incredibly proficient and interested. And I think there's a connection there between both interest and proficiency. So with those 900 that you referenced, they may all be at a different part on that spectrum. And so that's one thing that we're trying to solve for. And again, the previous question was about a gap. That's it is how do we ensure that everyone who leaves our programs and graduates and goes into the classroom has an appropriate level of proficiency? And then taking that even one step deeper, it's not just the individual teachers. I think it's also districts. You know, some are very AI forward and are investing heavily. And then I live in a small town and AI is not really a big topic of conversation in the K12 district yet. So I think all those factors come to play. Yes, put teachers first. But it is not, as the previous comment, it's not a one size fits all approach.
A
Definitely not.
B
Thank you for that. This is making me think, especially when Jodi, you said proficiency. There's so many different components here to build that proficiency around. Jim, your work experience is very interesting. Your team is building systems that integrates instruction, assessment, coaching. You essentially have teachers in the entire ecosystem of your product and work. What are some critical design principles that you've learned or believe still need to be worked on?
D
Yeah, I think that's a key question and I think this is going to sound like something you shouldn't say out loud at this conference, which is bring it out. Yeah. The best design constraint is what happens when you can't see the technology. More importantly, you can't see the AI. So if you keep putting the technology in the forefront, it changes the attention of what you want your user to do. So let me paint a picture. I was in a fourth grade classroom three weeks ago and teacher was doing a lot of great teaching at a smart board. The kids were getting up, they're using a blending board, changing letters, building sounds, and great instructional practice, explicit systematic literacy instruction. On her lap was a spiral bound notebook of what she needed to teach that day. So what happens when you start flooding that teacher with nonstop, iterative, generative content that you expect to be adjusting on the fly so she can work with her 30 kids? That just doesn't. It's not practical. That doesn't work. So there's this delivery from what you want from a prep session. It may make a lot of sense sitting in front of your AI tool and prepping and doing it on your own. But then how does that make its way into daily instruction that's going to actually use evidence based practices to create really good outcomes? Then how do you measure those outcomes and then how do you build the capacity to do that? Well, what if the technology is something that's working silently in the background to make that happen? And then back to the point around getting back up to the district administrator, what if the district administrator actually has a visibility into that and then can start the school year? What happens when? So we're, I'm charged with, among other things, building high quality instruction materials that has to go through very rigorous, compliance driven content reviews at state levels, state boards, parents, communities, things like that. So once we get all that built, things are standards aligned. If you take that and you put it up against a district calendar, what's the first thing that's going to happen? You're going to probably find out that you're anywhere from three to six weeks over time and the teachers can't teach it. And you're expecting kids to engage with really rigorous grade level instruction. But what if 80% of the kids can't access that, but that's what they need to do that day in class? And so technology can make a lot of that stuff happen. But if you kind of put the, if you put the attention on the technology, you take it away from all the things that we actually know that work in education, which is combining HQIM with good professional learning, with good measurement and then just the simple practices that we know from Hattie, that works. Let's make sure there's good classroom discussion happening in the classroom. Let's have kids get feedback in the moment, not three weeks later after you grade a paper. So. But technology can make all that happen. You have to stitch it. That's the stitching it together to make it happen.
B
Okay, I kind of want to see you nodding. I want to know what's in your mind right now. What is landing as true between both what Jason and Jim are saying and what makes you skeptical too?
C
I'm very grateful that I have the protection of the union. So I can say this without anything blowing back on me.
B
Say a little bit more about.
C
But what was something that Jim said that really, really, really like touched me was the fact that so much of the design of these platforms and these new tools are designed for digital citizens, meaning they have grown up their entire lives with technology. Then you have those of us that are elder millennials who remember a time where I even in college, in undergrad, I remember when my university got wifi on the campus and that was my junior year. I've been able to navigate between the two. But then you still have my teachers that are teaching who never use technology. The erasure of the conversations that's happening at a lot of districts is based around not just compliance as far as student privacy, but compliance with all of the things we need to do. We need to Prepare for the CAS. I'm in California. We have to prepare for the CASP. We have to prepare for the CASP. Then we have LPAC testing. Then you have now the K through 2 reading difficulty screeners. And so when they're purchasing this technology, they're purchasing it to meet compliance. But then they're not one involving teachers in the conversation, in the implementation. They're sometimes not even really providing the training. And so what then is happening is that you have districts that are either mandating teachers, use the AI or use these tools, which then if I'm a teacher who started teaching in the 90s and I've been effectively able to have a successful implementation of my lesson delivery, I don't necessarily want to feel forced to use a tool that I have not been trained on and I don't understand. And so when we're having, yes, there are things that can help with teacher efficiency, with teacher workflow, but if the conversations aren't being had as to how this fits into implementation, how do we utilize this within the, within this framework of lesson delivery? Right, Because I'm not just going to do. We, I use hmh, we use savas, we, I teach at a public district, so we use everything. Right. I am not going to assign a struggling learner on Savas. There's a thing called math buddy or math coach. I'm not going to assign that to them if they're struggling, because all it's going to do is continue to be information overload. It's a lot of reading, there's a lot of steps. Then if they get the problem wrong again, they have to redo it. By the time I do that, I'm like, let me just pull you for a 10 minute reteach. But I want that to be just as valued as if I were to. I know we're paying a lot of money for this curriculum, but if that's not going to effectively support the need of my Learner. I want to be able to have the autonomy and the control to say, this is what I'll use, this is how I'll use it. Maybe I'll use it like this this year. But if I get a new set of learners with learner variability, right, Because a lot of these tools are one size fits all. I can't implement it every single year. Or I may not want to, or I may not even understand. It's cool. But how, how is this going to support my pedagogical delivery? How is this going to support universal design for learning to where I'm actually meeting the needs of my different learners, not just giving them a glitz and glam fun tool that's on their Chromebook?
B
You guys were even pushing my own thinking. And I appreciate that in the work that I do. If I had to attempt to sum up, what I'm hearing right now is that AI really needs to earn its place in the classroom. We cannot just default assume it exists, so it needs to be there. I appreciate the push all of you are bringing to that. We've talked about many constraints. In the spirit of that, Heather, I'm curious if you can talk a little bit about your experience in Chicago. You're not always working with the most well resourced environments. In fact, there are some real constraints that some of the schools you're working with and educators can you talk a little bit about what are you seeing among this kind of first wave of AI adopters, the educators? That is promising.
E
Yeah. I mean, one thing I can tell you all about educators in Chicago is that we're very scrappy. Our path to adequacy as a district and as a state, we're on the journey, but we are not anywhere near where we need to be. About 70% to adequacy. So we have, you know, 70 cents on every dollar we need in the classroom. What's been interesting to us is that we're, and I'm reflecting on this in the context of this conversation. We talked a little bit about proficiency a minute ago and the spectrum of proficiency. I'm hearing you talk as a very experienced, exceptional teacher. So we see that same kind of proficiency and tenure range with the principals that we work with in Chicago's public schools. And it isn't sort of, you know, always the really young leader that is taking on something like AI as a tool to advance the work for us. It's often our high performing, so very proficient, very experienced leaders in our lowest resource schools. They're saying we're Doing everything we can with everything we've got and we know we need to do more. This is a tool that might help us and we have the cognitive space because we're good at what we already do to think about how to advance that. So resources are a constraint, but creativity is sort of the product that comes out. And those leaders are using AI right now, not so much to coach a kid, but much more to take the administrative burden either off them or off their teachers. Now where that gets tricky in a low resourced district is that you often only have restricted kinds of tools you can use. Right. So again at this conference, not to call anyone by name, but you know, we use Google Suite and CPS and Gemini breaks sometimes. So you design a whole agent as a school leader, you roll it out and it doesn't work to save your teachers the time you want it to save. If you ran it on ChatGPT it would, but you can't, you know that because you ran it on your home computer, not with the data. Right, but you're pretty sure that would work. The restriction is there again coming against that creativity that we see among our leaders. So, you know, it's all kind of a moving target. But it would be a mistake to say we're going to see, I think that we're going to see the best use of this tool in highly resourced environments. I think you're going to see the best use of these tools in environments where people are scrappy and experienced and are doing everything they know how to do to help get their kids what they need.
D
Heather, it's interesting. Anthropic published probably was the largest qualitative market research study a few weeks ago and one of the things that like 81,000 open ended responses went out across the world and one of the things that they mapped was fear and concern about AI adoption of AI against things like wealth and education. And there's an inverse relationship between. This was regional. So basically the wealthiest, most educated countries had the most skepticism and distrust of AI, yet the lower income, less educated had the highest belief in AI. Which I think has an interesting parallel to what you're describing because it's like here's a tool that's going to enable me to do because ultimately what AI is, is it's an enabling technology. Right? And we've kind of, you know, we've kind of given it human faces and stuff like that, but it's basically an enabling technology. And if it's allowing under resourced and then scrappy and really aggressive educators to get for their kiddos what they need. Then we're seeing that and then you see this, and you're seeing this playing out globally in other economies as well.
A
I think we are all kind of systems thinkers, which is probably why we're on this panel. But I do want to maybe remind us or provoke you that 30% of teachers have left the profession in the last three years. We have a burnout crisis of epic proportions. School district of Philadelphia hired 1,000 people into teaching roles, special ed roles, other educator roles that had no teaching experience whatsoever. But that's how desperate they were. And I don't know how it is in Chicago, but I bet all the large urbans have this challenge. So in a way, we're kind of, you know, we're able to talk about this because we're at asu gsv. We get to come to this event and talk about how AI could have a renaissance in teaching. But meanwhile, the profession is dying. It's terrible. My sister left teaching two years ago. I wrote a substack about this that got more views than ever before. Just saying we need to pay teachers $100,000 a year if they're proficient and tell them how proficient they are and be transparent about what proficiency means. But we have this bifurcated system where a lot of it is evidence based, where it's about HQIM and professional development and proficiency in the large urban districts, whereas in the rest of the country, they're adopting tools and kind of scrappily doing it on their own any way that they can. And it's really hard to talk about the system as a whole without just recognizing teachers are under siege. Like, we could easily come back three years from now and have 40% of the remaining teachers having left the profession. We need to respect our teachers. We need to pay our teachers. We need to let them have the loudest voice in the conversation about technology in the classroom. Teachers first.
C
I'm definitely going to piggyback on that. And I say this because I think it is less about teacher burnout and more so about the treadmill getting longer. And as the treadmill is getting longer, there are more weights that we're having to carry as we're running this journey. Because the thing is that a school school year is a cycle. So every year I know I'm starting off with fifth graders who are now in sixth grade. By trimester two, I have sixth graders that are a little less feral. By trimester three, you're basically seventh graders. And this repeats every single year. But the difference that I had experienced of that year one is very different than what it looks like year 15. And for me it's more so about the lack of autonomy, the lack of support. And I say lack of support because a tool is a tool. But if you hand a baby a hammer, what are they going to do with it? They're either going to destroy something or they're going to just, just let it lay there. And so much money is being in. This is from a teacher and. Right. There's so much money that's being invested into tech and in tools and in programs that it is, if you like you said, a union. I'm a union vice president for my locals. So I attend board meetings, I read the budgets, I read the meetings. And when you see that we're paying vendors or we're paying for tools that are more expensive than our starting teacher salaries and our starting teacher salaries with teachers who are maxed out on the pay scale, which means that they have 75 post baccalaureate units plus their masters, but they're making $68,000 a year. But we just signed a contract for $100,000 for a person or a vendor or a tool. Teachers are becoming more aware of this and as we are becoming more educated and as we are being forced to do more or more onus is being put on us. Originally I was a teacher. Now I'm a teacher, a counselor, a mental health professional, a nurse, a lawyer, an advocate. Like I'm doing all of these things, but the value of my work is not being recognized in compensation. And I say that because I don't think anybody's really looking into teaching because they're going to make a whole lot of money. But the desire to be in teaching is because we want to make the change, but we also need to be respected as these changes are being made. We don't need to be told what the changes are. We need to be a part of the decision making process. And I think that's why we're leaving is, is not because we fell out of love with kids or with teaching. But we are in a profession where there is no. I don't know where the goal, I don't know where the finish line ends. I don't know if this year is going to be the same as next year. Like we don't know. And so much has changed. 2020 teachers were heroes in 2024, 25, 26. You have states that are talking about replacing teachers with AI in their classroom or hiring people that have no experience, no credentialing no. Anything and putting them into classrooms. Why would I become a teacher?
B
Okay, this is bigger than AI. And we knew this. We knew this. Starting with Jody. What can we start doing to signal that teachers can trust us? Because even when we talk about ed tech tools or AI, decision making is happening somewhere else. Implementation is happening somewhere else. We should be called out. How do we signal and build that trust?
F
I'm so appreciative that you were so honest. Thank you for that. What we're attempting to do at American College of Education is engage with those teachers and that feedback. We're asking our students and district leaders to join us in the conversations, to join us and give us insights and perspective of what we can do to better serve them and meet their needs. And I think that's where it starts, is being able to have those types of candid conversations and coming together to create solutions. But the thing that really came to my mind too is I remember being a kid and if you were a teacher in the community, you were a hero. You were an like, you'd go to the grocery store and oh my goodness, you would see Mrs. Whomever. And they were the most respected profession in your community. And I wish there was a way that we could come back to that and then layering on AI. AI should be a tool, but it should never replace, in my perspective, the great work that humans are doing. Because just as we talked about the nuances of proficiency with, with educators, the same is true with students. And you talk about needing to fill every one of those roles. Well, that takes a human being to understand and be able to support and serve those students. So for us, we want to engage, we want to hear from. We have a student advisory group that is made up of teachers that is helping us inform how we approach our curriculum, what tools we leverage. We've created an AI literacy center, and much of the content in there is coming from teachers, superintendents and principals to make sure that we're doing all that we can to come alongside and be a support, but certainly not replace.
B
Jason, I'm curious, from your breadth of experience, there's a trade off, right? The last thing we want is the pendulum to keep swinging in opposite extremes. So on one end, there is a potential for AI to make the teaching profession more sustainable. On the other end, it could make it a lot more prescribed and actually lose the humanity and the judgment that we need teachers to maintain and execute on. How do we think about that trade off? And I'm not going to use the word balance because that's not the right word, but how do we continue to honor the human side but still create space for technology to be a value add? Simple question.
A
I definitely don't know the answer to that question, but I'll try to articulate the way we're thinking about it, which is like, I don't know if any of you saw Dan Meyer's post in the last 24 hours that Khanmigo is dead. So three years ago, Khan Academy. Big deal about Khanmigo. 60 Minute Special. Everybody's going to have their own AI tutor bot. It's an incredible demo. It's an incredible product. They basically have said it isn't working. Students don't want to use it. Students don't want to engage with a robotic tutor. Students want like a real engagement human. I mean, there's all this research about the number one. I don't know if it's number one. One of the top factors of success in life is did you feel like a teacher cared about you during school? And you know, I know who that was. That was Ms. James for me in second grade. And I ask people this all the time because they almost always know who that teacher was. And if they I didn't know who that teacher was, that really changed their life. I know they've had an extra tough life and they had someone else that was the inspiration to them. So long story short, the tools that enable teachers like our teachers say, oh my gosh, I don't have to keep everything in my head. I can just teach freely without having to remember every little detail about all 150 students I teach. My brain is free. Like that's the feeling we want. You want to be so in the moment with your students that you don't have to run back and forth to check in attendance or tardy or a behavior issue. You want the AI to be invisible in the background helping you. That's for sure the way this should go. I'm very worried it may not go that way, but we're going to be the force for teachers and AI working for them in the background. And by the way, we don't even say AI on our website. We are a voice powered class companion and we scraped the word AI all off everywhere because we're in a backlash against AI right now that barely is mentioned in this conference right now. But next year is going to be the trough of disillusionment because AI has been promising great things and this, you know, and it's not going to deliver in the next year in the headlines. But it will be delivering underneath to tools that focus on really being helpful to people, especially teachers.
B
And so Jim, I'm curious of what you're thinking right now. You've been doing this curriculum work for some time, even before AI tools were around. And as you're hearing about this trade off between sustainability and autonomy for teachers, what's coming to mind?
D
I think, Megan, you're going to be in my head for probably, I don't know, for a while. I think when you listen to what Megan was describing, and I think if I had to paraphrase it, it's like she's making. And teachers every day are making thousands of decisions in their classroom. Right. And that decision support can be supported by AI and by edtech tools, or it can continue to create a longer treadmill that you basically just get. It's like you're overwhelmed with more information that just leads to not being able to make decisions. And so I think when you think about what you know, and again, we know what works from the research and the evidence and if we can take those practices, merge it with the things that can help provide good decision support. And that's like a lot of things that happen in the background around data analysis. So you talk about the student. Should I put my student into this tool or not? Well, that could take up a lot of time. And you may not be able to make that decision in the moment, or you may not be able to, or you may want to plan it out over a course of a week. That's where tooling can do that to provide you with the type of clarity and confidence that you need to teach your lesson. And what's one of the things that's going to lead to a really good lesson, to really good student outcomes, is teachers confidence in the front of the classroom. We're going to take away teacher confidence. If they don't know if their technologies work, they don't believe what might be a hallucination coming out of their chatbot. So we have a responsibility as an industry to make sure that trust and credibility is at the forefront. But ultimately it's leading to these micro decisions that are happening every single day in classrooms and let you get back to doing what you're doing. And I think the. I like to, I observe a lot of classrooms and one I was in about a year ago was in Brooklyn in New York City and the principal gave me a challenge going through my day of observing classrooms because I want you at the end of the day to tell me which of the teachers in this Classroom in this school had the highest gains on their assessment that they were using for like an interim assessment. And I came back and I came back with kind of like a mid tenured teacher. He's like, nope, you're absolutely wrong. It was the first year teacher who was just coming into the classroom and she was following the guidance and the support that came from the combination of hqim, that came from good professional learning. But ultimately, when I asked her like, so why were you successful? She goes, well, I was always confident in every decision I had to make because, because it was laid out for me clearly. I had the data coming in to tell me what to do. And then I always knew that in front of my kids I had to put on a good game face and I had the confidence and clarity to do it. And I think that's where autonomy. She had the autonomy to make decisions in the moment, but she had the clarity that came from some set of structures that enabled her to do what she needed to do.
B
Thank you. All right, we just have a couple minutes, so I'm going to do a rapid fire question here. I want you to be very concrete. 20 seconds. Some of you fund schools, some of you train teachers, some of you in the classroom, some build products. So thinking from the perspective of one of those, what is something concrete someone wearing one of those hats can do not a year from now, but next
C
week,
A
Include teachers in your product development process.
C
A thousand percent. I would second that. I would say include teachers in your product development because one of my favorite tools that I use is snorkel, but I use it because it gives me what I need to do to be a more efficient teacher. I'm able to gain custom insights. I'm able to extract and say, okay, my reteach isn't going to be 30 minutes. With 10 kids, you're for five. I understand what's going on with you and I'm still, still, I'm still delivering the instruction. It's helping me with the data. My last point is teachers need to be at conferences like this and no shade, because I very much so appreciate to be here. But I am here on a scholarship because I could not afford the $5,000 that it takes my district. I'm hoping my district will pay for my sub for me to come. Right. But when you think about it, it's hard to have an educational technology conference when you don't have the practitioners and the users that are on the panels that are leading the conversations, that are having the conversations with developers so that they know what we need.
B
Thank you, Megan.
F
I echo what they're saying, but one other piece of advice. I would say as an administrator, it's hard to create a strategy with how rapidly things are changing. And I think a piece of it is to really be mindful and aware of that. But the biggest part is to have regular conversations, include everyone you can be transparent and communicate with all of that. I think people expect senior leaders to have every answer and a very clear strategy. It's evolving so rapidly that that is not the reality. But to lean into that and to include everyone you can in any decision and conversation that you're having about AI. I learned the most from others who are have found a new tool and put it all out on the table so that everyone's aware of what's out there.
E
So I think we expect teachers to walk into classrooms every day and to recognize the individual needs of their students, to meet the needs of the whole group and to meet the needs of the small group and to meet the needs of the individual. But we don't expect districts or systems or states or the country to recognize those same differences in the teaching population. We talk about things like 30% turnover rate. We don't recognize that that means 30% of our teachers are brand new every year. And we don't design for that kind of differentiation. We just ignore it. We talk about involving teachers as a general concept, not as a specific need. And we talk about designing around teachers as a more often than not one size fits all, not a let's design for the different types of teachers that we have.
C
Thank you.
B
Jim, close us off.
D
I'll be fast. For every one thing we as a product development company, industry do, let's take five more things away from what we're asking a teacher to do. If we're going to ask them to do one more thing, we have to promise we're going to take five more things that don't create value for them.
B
This was actionable. I hope you'll all take it back with you. Thank you again to this brilliant panel and thank you for joining us.
Title: Will AI Give Teaching the Renaissance It Deserves?
Podcast: ASU+GSV Summit Sessions
Date: May 6, 2026
Panelists:
Main Theme:
This live, candid panel explores the challenges and hopes surrounding AI in education—especially its impact on teachers. The session cuts through tech hype to assess whether current AI advances are genuinely poised to transform teaching or risk repeating cycles of false promises. The panel highlights teachers’ voices, practical implementation realities, systemic hurdles, and concrete next steps.
[00:13–03:41]
[03:41–06:25]
[06:38–07:56]
[08:30–12:21]
[12:21–15:43]
[15:43–19:43]
[19:43–24:25]
[23:19–24:25]
[24:25–29:57]
[29:57–33:10]
[32:11–35:10]
[35:26–38:06]
| Timestamp | Quote | |-----------|-------| | 01:15 | Jason Palmer: “Teachers can be the leaders because they are the ones who actually care about young people and have been caring about young people despite all the chaos that's been thrown at them for decades now.” | | 02:11 | Megan Mitchell: “What has been missing…[is] putting teachers not as the product, but as the designers, a part of the decision making process.” | | 03:58 | Jim O’Neill: “The industry needs to iron out those seams and solve for that patchwork. And AI has the capabilities to do that.” | | 05:16 | Heather Annakhini: “Anything designed to fit all will fit none. Zero.” | | 08:30 | Jason Palmer: “Teachers are all part of building the product with us...Teachers first. If you remember only one thing about Socrate: teachers first.” | | 16:03 | Megan Mitchell: “I want that...re-teach, to be just as valued as...a glitz and glam fun tool…on a Chromebook.” | | 20:26 | Heather Annakhini: “You’re going to see the best use of these tools in environments where people are scrappy and experienced.” | | 24:25 | Jason Palmer: “We need to pay teachers $100,000 a year if they’re proficient and tell them how proficient they are...the profession is dying. It’s terrible.” | | 26:10 | Megan Mitchell: “There’s so much money that’s being invested into tech and in tools and in programs that it is...more expensive than our starting teacher salaries.” | | 32:47 | Jason Palmer: “You want the AI to be invisible in the background helping you. That’s for sure the way this should go.” | | 35:26 | Jim O’Neill: “Teachers every day are making thousands of decisions in their classroom...decision support can be supported by AI...or it can create a longer treadmill.” |
[38:29–41:57]
This panel makes it clear that AI is not a panacea and cannot “save” teaching without deep, sustained involvement of teachers themselves. The real Renaissance for the profession will require putting teachers at the center: designing with—not for—them, reducing burdens not just adding tools, and restoring respect, agency, and support to those at the heart of learning. Whatever the promise of AI, it must be kept. And the work starts now, with teachers leading the way.