At the Meadow with Priten Shah

Hello, and welcome to this, the first episode of Season 3 of At the Meadow. I'm your host, Joseph Carver. The Meadow is where we gather to discuss all things related to innovation, creativity, and independent schools. I'm excited in this inaugural Season 3 episode to welcome Prit and Shaw, the CEO of pedagogy.cloud, as well as the author of AI and the Future of Education, Teaching in the Age of Artificial Intelligence. Prit and I are old friends, so the conversation is familiar. It is welcoming to people that do not know a lot about artificial intelligence, and yet we talk about some of the most pressing issues facing schools today. I hope you enjoy this episode of At the Meadow. So welcome, Prit and it's great to see you. As we were just talking about on the sort of pre-call, you and I have had a number of spots where we've intersected, whether it was working on transitioning to distance learning at the start of the pandemic or speaking together at the NDCA conference. I'm always super interested in the areas that you choose to turn your attention to, because you're a valued thought leader in education. So I'm very grateful that you've made time to participate in this call. We want to talk mostly today about artificial intelligence and education, and specifically about your book, AI and the Future of Education, Teaching in the Age of Artificial Intelligence. I feel like, and I'm hoping you can talk me off the ledge a little bit, I feel like by the time we publish this podcast, most of what you and I discussed might be already field-dated at the rate of the sort of acceleration that's happening with AI. So let's start there. Is there an analog to the speed at which AI is changing education in any other part of educational history that you can think of, and is there reason to be concerned about specifically this rate of change and development? Yeah, that's a great place to start and thank you for having me, first of all. I think that, you know, this is the part that I think we try to make clear to everyone we're talking to, is that we haven't seen a change this fast before, and so I think there's lots of reasons why that's true, and we can talk a little bit about that, but I think historically we've had lots of major changes, right? I mean, the education system has been hit with lots of different transition points, and COVID-19 pandemic, which we brought us together for a little bit, was one of those moments. But I think what really sets this one apart, I think is the rate at which it's happening, the permanence with which it will stay, and I think that's what really gets me worried about, the rate at which we're adapting to it, because I think when the COVID-19 pandemic came around, we were all, it was a rapid change, it was overnight similarly, but we all knew there was an end to it, and so I think our, some of our adaptations over our measures to deal with it were temporary. There were things that we knew that we didn't have to keep forever, they could be banded solutions. One of the problems we're facing right now is that folks are trying banded solutions that bands across the country we're seeing are a great example of a banded solution, but we really need to be adapting with the pace of this. The other major difference between COVID-19 and AI is that their pace continues to accelerate, and so once, you know, after the school starts down overnight during the COVID-19 pandemic, we did hit a plateau in terms of what kinds of changes were happening and what kinds of things we needed to adapt to. If we just think about the last year that we've seen the rate of change has been really rapid, and every, you know, every time she's just like, okay, the students can't, won't cite sources, it makes things up, it doesn't have data up to a certain point, it can't read writing, although it can't do math problems, those kinds of hurdles get solved really quickly, and so the pace is pretty rapid. I'm hoping though that the kinds of things that we need to talk about today are going to be probably relevant for the near future. I think there's some high level things that will all need to be continued to have conversations about as educators in order to better serve our students in the new age of AI, and I think that the pace of development in the tech world, we're not, we're not staying up on par with it anyways, and so I think there's some meta-level conversations that we'll be having today, and probably that will stay relevant. Yeah, I mean, that's a great point that, that while the folks that are working in the tech sector are going to be moving at the speed of light, schools don't have to. We can be a little bit more circumspect about it. So let's take a step backwards and start with what I think a lot of people will listen to the podcast will at least be wondering if not allowed. What defines in the most simplest terms the difference between the kind of technology that we're talking about and that you're talking about in your book and something like Siri or, you know, whatever's built into the the Amazon product cycle, like I think that for some folks they sort of feel like, oh no, I've had this sort of built-in personal assistant on my phone for a long time. What's all the fuss about? Can you talk a little bit about the difference between, say, that and like what we can do by leveraging large language models? Yeah, that's a great pace to start. And so yeah, I think that you hinted at the key difference here, which is the fact that these are large language models, and I think the word generative, which is thrown at a lot, I think is really helpful for folks to understand the difference here. And so the way to think about it is the AI technologies that we've already seen in the past, whether it be the Netflix algorithms, whether it be Siri, they're really trained at doing a particular task, and they're really good at recognizing things, and they're really good at outputting more generic content, right? So for example, Siri does a really good job of being able to pick up some basic things that it often hears, and so something like, what is the weather? It has tons of training data on and knows how to recognize all the different ways that somebody might ask that and what different tones somebody might use, but it's not really good at if you started throwing a new question at it, the most it will do is run a Google search for you and then tell you to here's a link to it and click it. Generative AI models that really became popular last year are popular because they're able to generate new content that mimics human-like content, and that's really where there's all the potential for fear, but also all the potential for the benefits that we can see in education. And so the way we want to think about this is these models are, so AI in general is a field that works to try to mimic human-like cognition, and these large language models have figured out how to mimic human language abilities, whether it be speaking or whether it be writing, in all sorts of domains across all sorts of expertise areas. And so that's really where we start to see uses for education, where in the past maybe asking Siri to summarize a current event, who would get you three links to the New York Times, CNN, and Fox News, perhaps now ChatGPT or another language model could perhaps pull those links up, take that content, and then create something original of its own for you to consume. So, and that's a big distinction because we are talking about replicating cognition as opposed to trying to do your best to determine what the request was and what sort of lines up with that request and other available resources. So some of it I'm sure has to do with the size of also these models and what they can do, which is an interesting question in terms of where that data is coming from, who owned that data? I think those are interesting questions. But one of the things that I feel like for our listeners would be helpful is there's a lot of conversation about AI and a lot of that we're seeing people use ChatGPT almost as in the way that people use Kleenex to describe tissue. ChatGPT is a specific part of AI. Can you talk a little bit about what ChatGPT or what you refer to as the other models, what some of those look like or who they are and what differentiates what seems like ChatGPT seems to be kind of the bully on the block right now and why that is? Yeah, I mean I think there's a first commercial advantage that ChatGPT has and so this is where Google became synonymous with searching something on the internet despite the fact that there are lots of other search players and in fact there were search players before Google but because they became the first popular largest one, we now often will be like, I'll Google that later on. I think ChatGPT had the same benefit here. They were the first ones to make something that was public access that didn't involve any coding skills in order to access and so the public could access immediately, you need to be a developer to make use of a language model despite the fact that these language models, even open AI's own language models existed predated last fall. They were just only accessible to a small group of people who could use them in their programs that they were coding themselves. So ChatGPT became the first player because they were able to make it accessible to someone who didn't know any coding and then their model is also so far one of the best. The only reason I say the best today and not one of the best and not the best today is because Google just came out with their new Gemini model which now finally there is some competition for GPT-4's model but prior to today the best model we had in terms of all the benchmarks that were put out was open AI GPT-4. But there are other models and so we have Claude is a major competitor for them, has a model that's similarly competitive, it's created by X open AI engineers, Google obviously has their suite of models that they're putting out and now we're going to keep seeing all these things pop up in other places. So Amazon has one, Twitter slash X has one and they're all basically trading on different purposes of data, corporate data in order to learn how we think and talk. And so we'll continue to see more of those but ChatGPT continues to be the most accessible and the best model that we have seen. It's funny there's these days that I'm sure when I look back on my academic career I'm going to remember with real clarity obviously one of those days was when I was brought in to a meeting and told we're sending kids home in the next two hours we don't know when we're going to have them back what can you do to make sure that they have devices to leave with all of that. That was a memorable day. Another one was clearly the day that ChatGPT became available to the public and I remember very clearly there was a real breadth in responses among leadership in school and it was great because it was like its own little ecosystem. There were the folks whose immediate reaction was we're just not using computers in class anymore at all. There were the folks who were immediately thought of what this meant they could leverage like oh I could see how this could be used for this this and this you know but in the middle there were just a lot of people who just sort of were thinking I'm not sure what this means. The funny part is those people that were out on both ends of that spectrum probably have all moved to that middle now where everyone is sort of in this still trying to figure out what that means. So for folks that are not spending a lot of time researching or reading about it can you talk a little bit about some of the most basic ways the schools that you've dealt with are integrating AI into the work that they're doing now and that can be academic it could be you know something not related to the classroom experience but where do you see it making the most immediate impact without the heaviest lift? Yeah so I think we we've been splitting all of our you know whatever our PDE that we do and conversations that we have into two buckets in terms of utility and so there's the teacher-facing content and there's the student-facing content and I think that the easiest way that schools have started to integrate this is in the teacher-facing realm and so it's much easier for us to start having conversations about how a teacher might use it to plan a lesson, to create a worksheet, to draft an essay prompt, maybe even to grade a essay at first draft for the grade please and all of those kinds of things we're seeing much more use of across the country across many different school districts. I think the the barrier is much lower for someone to be comfortable with that because the student doesn't necessarily engage with AI themselves and then there's a there's a buffer moment where the teacher can kind of say okay this was ridiculous this is a terrible lesson plan I don't want to put this in front of my students or here are the tweaks I'm going to make and so those are you know lesson planning worksheet creation assignment creation is definitely the highest rate of uptake we've seen across the country. There are now finally this fall late this fall in fact schools that are interested in figuring out ways that students interact with and I think this is where folks are concerned about AI literacy they're concerned about students being able to navigate this ethically responsibly and so we're starting to see responses from schools interested in AI products geared towards students we're starting to see responses about oh like what are what are ways that we could all do this together in the classroom maybe I could put a GPT or bar on my smart board and I have I can type the things into it but everybody can kind of witness it together so that we're starting to have important conversations about it and this is you know those those that's finally starting to happen that's still definitely in the early stages most folks are finding it the most easy way to approach the technology as an educator is to figure out how it fits into your own workflow figure out what tasks that you find very daunting or annoying and figure out a way to get one of these language models to help you out. I think that one of the immediate responses as chat GPT started to reach a level of saturation and especially in the academic consciousness was these AI detector tools which you know I was dubious early on about the efficacy of those products and I think that time has sort of borne that out can you tell me a little bit about why a tool like you know a chat GPT detector is not a strategy that is long-term workable or valuable for a teacher in the classroom. Yeah I mean I think that I think you've already highlighted the fact that there it's not even short term workable and so these these technologies don't work there maybe initially when they first came out there were some evidence so that they could detect with some high level of probability still wasn't perfect probability but as the technologies have gotten more complex as more different models have come out it's basically become very easy for a student to go in and make it make it whatever they've generated not detectable by these AI detectors and so whether that be using multiple AI detectors whether that be putting it through one one and then a different one that is tailored towards avoiding detection tools I mean the amount of two options students have to evade detection tools today are pretty wide and effective and there's a reason why major universities are moving away from any tool that does a detection in the most school districts are also starting to realize that that is not an approach that they won't suggest to their teachers long term this is this is exactly why I think we this was a short-sighted measure in the first place we will continue these technologies are being built so that they can mimic human writing and so detecting human writing is not really going to be very effective in terms of hearing out of students writing versus AI writing especially as this model starts to be customized to a particular tone that a student's already taken as students become better at prompting the tools to take on a certain grade level vocabulary feed even their own writing into these tools first in order to mimic that writing all those kinds of evasion mechanisms will continue to see and so we need to rethink what we're having students do and so I think this is where this is this is not an easy thing I hate having this conversation because people hate me at the end of this conversation but you know when every time someone's like what can I do about the plagiarism problem my answer has to be we need to rethink the kinds of things where a lot of like the students can plagiarize and there's lots of approaches to this and so we're seeing STEM schools move towards more in-class essays and that's you know again that's a very effective short-term measure in terms of getting students and that allows students to be able to access these technologies while they produce the output that you want them to produce but there also are longer term questions we need to ask what is the value of writing an essay what are the skills we're trying to teach by having them write an essay how can we motivate students to see the value in going through the process of writing an essay and what are the ways they can build those skills that don't require writing an essay and those are those are all questions that I think we all need to be having long-term I don't think these are questions that we can answer today and then have you know a massive overhaul of the education system overnight but I'm hoping that these conversations start this year so that by next year we can start seeing some more innovative approaches to figuring out what how we're going to be teaching our students and keeping them engaged in the actual learning process and not just an output focused process where they're worried about getting an essay out of 3am for the next day. I'm gonna I'm gonna get on my soapbox for just a second and I'm hoping you can either say yeah Joe I think you're on the right track or you can course correct me. Recently I was invited to participate in a course for statewide educators at AI and education and I think that the work that they were doing is important valuable I applaud them for for jumping into the for the deep end but the value proposition for taking the course was we need to be using AI in classrooms so that we can train students to become effective prompt engineers because that's going to be the job of the future. Now my immediate reaction to that is we already know that that the machine learning piece of the computer piece of this will learn more quickly than the humans will learn. We know a bit roughly the way the brain works we've got you know we know what our computing power can do. My guess is in a foot race for AI to figure out what we're trying to ask and us figuring out how to ask it well. AI is going to beat us by a pretty long stretch meaning that this pull prompt engineer thing is going to be irrelevant. They're going to figure out what it is we're asking before we figure out all the necessary tools meaning that the value proposition for all of these teachers is a role that probably will be irrelevant by the time they finish the next academic year. So what first am I on the right track do you think that prompt engineering is really something that is temporal and then secondly if it's not prompt engineering what's the value proposition what why am I as a teacher am I making space to teach these tools in my classroom. Yeah both great points so I think I'm right on there on this whole box with you the prompt engineering for the masses approach I think is a short-term one I think these language models are being built to be able to discern what we're asking and you're right that they'll be they will much more quickly figure out how to do that then we will be able to train the entirety of our you know student population let alone teacher and student population or the population at large how to write a prompt as is currently needed by a system to get the exact output you need. These language models were built especially like chat to PT, bar, cloud they're all built so that they are better able to take into account your follow-up questions your feedback and I find that that is the easiest way for anybody to approach this is have an actual conversation it's not a one off I put the first prompt has to be exact right thing so that I get the exact right up the right away go ahead and tell it what it's wrong what you didn't like and that will get you very close to the final output you need and most people will intuitively then include that in their next first prompt once they realize that oh if I specify that I want this in a table format as the second message I should now include that in my first message and that those kinds of things will happen intuitively and that's also how we approach human conversations you know when we give students instructions in the classroom you know one year we might give instructions to students have a bunch of questions about it and then next year we'll include that as part of our initial instructions to ensure that they don't have the same question and I think that approaching this similarly with chat gpt and other language models is more than effective in the short term in it in fact 100% the only approach in the long run and if we don't want to waste a bunch of time learning things that are only applicable for on the day or two and we thought this we've already seen this in the past year last year folks were spending a lot of time learning how to write a mid-journey prompt in order to create images and those were complex there's a little bit of context like a domain expertise a particular sort of language and syntax if you just learn it or to do that properly and this year folks are generating similar quality images at least which gpt that now writes the writes the image prompts for you and we'll continue to see some more but then I think you're right the question I think that when we're talking to teachers about this we we try to make clear that the goal here isn't for them to learn how to use a particular tool the goal isn't for students to know okay this is exactly how I'll use gpt the goal is for them to start thinking about how they might augment their own capabilities using these AI technologies how might they be responsible consumers of the application technologies create can they sit down and look at something an added tool I generated and be able to see through you know what what they need to fact check are they developing into versions to know okay this might not be true this is what I should double check here's how I can even effectively double check and do my fact checking properly what sources can I rely on those kinds of digital literacy skills that were very important in the social media era in the last two election cycles are even more important now and those are the basic skills I think we need to keep teaching our students what is responsible usage of it in terms of what information you share with these tools are we what should you put into your chat gpt bot what information should you not share with these AI models those kinds of conversations I think are really important to be having in our classrooms but I think if we're focusing on how to exactly craft was perfect prompt we're really missing the vote on what we need to do to prepare students this is more my personal curiosity about what you've observed and then I want to talk a little bit about your book but I've been a bit surprised that schools haven't embraced strategies for using AI on the operational side as quickly as I would have thought in other words I see the other day there was a discussion about scheduling you know in the amount of time it takes for us to complete a schedule for say an upper school division I'm not suggesting that you know chat gpt can generate our schedule for us perfectly yet but there are countless areas in the operations of a school where you could increase your overall efficiency with a pretty simple you know built out you know you can now build your own gpt and chat gpt where you know give it sort of an identity and say this is what I want from you I mean obviously you know this but for those that are listening if they haven't done it and I've built a couple of interesting I built one that was a handbook gpt and it was basically just a field questions about our employee handbook so how many vacation days do I have what are the requirements for my dress code you know just to see how efficient it was so has that been a surprise to you that schools haven't looked more broadly at like how to use it on maybe the operational side yeah I think that this is part of I think the transition in the narrative this year has been very like universal across the school districts and I find really interesting and so I think there was like this wave of let's talk about plagiarism let's talk about AI detection and I think now we're like in the transition point between the early fall which was let's talk about it in terms of like curriculum development and like what can we do with our students but I think that there's there I've really seen pockets of folks interested in thinking about how they can use for email drafting how they can use it for you know streamlining some of their communication tasks permissions for generation those kinds of like very mundane things that maybe require a little bit too much news letter generation we're seeing some of that but I think that there are larger ways that I think the examples you provide are much more robust so using it to generate some piece of writing is one thing but these models can do a lot more and I don't think folks fully know that yet and so you could potentially put in a spreadsheet of last year schedules some spreadsheets of your current student rosters your class rosters and have a kind of a pretty good first draft of your upper school class schedule if you wanted it to and it will write the necessary code for it and it will then like generate the final output for you so I don't think folks fully know that those things are possible you can upload you know last few years of grade spreadsheets and look for trends they can even tell you what things that you can notice I think there's lots of great ways to think about this in terms of the data size because it can do some of that data analysis for you and then figuring out bulk ways to streamline some of the communication tasks also I think there's a potential there but I think folks are maybe only seeing that right now starting to dabble with it but really I think like this this GPT example for your handbook is a great example the amount of questions I'm sure admins across the school across the country deal with about even HR related questions from internally with their teachers you could cut down on the time the handbook is the perfect example of why this kind of technology was super interesting to folks when I came out with you weeks ago I think at this point because I think that's exactly who's case is can we get it to talk about something really specific that's really particularly for our school and take away a bunch of you know human time that's being spent right now on something that really doesn't require like high-level human cognition and looking through a handbook and answering the question is definitely one of those tasks yeah so I was really excited I don't know how long ago it was but I was really excited you know it feels like almost a year ago when I got the email that was a request for me to read your book AI in the future of education and provide a potential cover quote for it first of all thanks thanks for that thinking of me I was it was an honor to be a part of the process I'm an exciting to see it come to fruition and to see a copy you know arrive at home what was the impetus for you I mean obviously you spent a lot of time at the intersection of technology and education but you haven't chosen to write a book about any of those other intersections yet so what what was it what what tripped that and what were you hoping to fill in terms of space with the book what questions were you hoping to answer and who were you hoping to answer them for yeah um yeah and you were one of the first people to come to mind because I think you're one of the other people who I see at the intersection of education technology and so I was like I need somebody else to speak both sides of this well understand what I'm trying to do here so I appreciate you taking that task and I know we give you a really short term I know that as well but but I think that the the key impetus here was that I was really excited about this technology we you know I there was a college project I had worked on with language learning that involves large language models we were thinking about it in the reasoning space for one of our clients in very hypothetical terms at that point because these technologies were nowhere ready for it and when it became that easy for accessible sorry I would like to see what we could do in terms of positive uses and so you know we built out our so-called AI tool to showcasing the kinds of productive uses of AI there might be in student-facing contexts but folks were not ready for it in fact most folks probably still aren't ready for it and I think that that's when I realized that the the narrative has largely been about fear and concern and plagiarism and I really wanted to take a step back kind of give a large overview of what this technology could do and mean for education in the in the positive ways but also really start asking some of the important questions I think we'll need to be asking in the next few years and so you know there's there's a lot in there about really practical uses that a teacher might have tomorrow there's a lot of prompts in there for teachers to start playing around with in terms of using it tomorrow but there are also lots of questions in there about what our education system might need to look like in a year or two years five years in order to better adapt to all these technologies and I think while the book plants a lot of those seeds I don't pretend to have any of the answers yet I don't think any of us have any interest because we don't know where we're going but there's there's I think hopefully the book asks the right questions to get us all thinking about those things and so that was one of the other major goals and putting it out and then there was also this middle line stance that I wanted to take here because I think there are folks who are just like okay AI can revolutionize things it can replace our teachers we don't need a public school system anymore everybody can just learn on their own at home and then there are folks who are proposing the bans and saying okay we need to teach students a way completely from the technology and I hope the book has taken that middle stance of tell us that we should be really cautious there's really important reasons why we still need humans in the loop why we still need educators interact with students directly why educators will continue to permanently play an important role in the development of our students lives but also here are really interesting ways that we can use it to streamline some of the things that we we want to achieve in our education system and I've been struggling to do so and that's that's the other whole care I had a headmaster that I worked for at Carrollton and I think you actually may have even spoken to him once or twice Olen Kalkus who was the head at Carrollton School of the Sacred Heart and years ago year and when I say years I mean seven years ago he said that he believed that there was a possibility that in the future only the truly privileged privileged would have access to in-person education that that the impact of the evolution of technology would be that the masses of young people being educated would be relegated to technology teaching them and only those who could really afford it would be have the benefit of in-person education that's first of all I think that that was pretty interesting foreshadowing and so the shout out to Olen Kalkus but but it's also a really interesting question because at the center of that is the answer to all of the fear that you're talking about right at the center of that is nothing replaces in-person educational opportunities we saw that writ large during the pandemic right so I feel like there's a real possibility for people to just take a breath and say there will always be space for human to human knowledge transfer it will always be preferable even in a world of singularity or in a world where we're in a battle with you know with technology for for some sort of dominance the interaction of two humans or a human you know 15 humans whatever will always have intrinsic value and if they start with that at the core it feels like the rest of this becomes how do we do that more efficiently or and this was my next question for those who we know that's not going to be available to how do we give them the closest fact similarly to that and so my question is what do you think the overarching impact of this technology will be to non-traditional educational pursuits meaning people like Khan Academy, home schools, micro schools all of those that seem to me to stand to maybe gain the most from some of this I'd be interested to know what your thoughts are about that. Yeah I you know I want to echo the the importance of the human human connection first because I think that that's a really important point I think that does help quell some of the fear and I think that my go-to example is that we still go watch humans run watching races is something that we find entertaining we watch the Olympics and even though we can build these hypersonic jets that can go across the country run and there's some value in the in the human feet here and so I think that there will always be value to that teacher who can crash down next to a student you know tell them that okay it's okay to issue with this math problem here's here's a tip even though that same exact hint or tip can be given by an AI bot so I think it's we we need to be very cognizant that any proposal that you know pushes for mass replacing of teachers is probably in order to make sure that only the privileged access that is the in-person education so let's let's try to make sure we stay far from that but I think in terms of um thinking about um we might as well cut this part out well the second question you had I'm so sorry no no it's fine now I'm not going to cut it out because this is what human interaction looks like my question was do you do you agree that that the the potential impact for non-traditional education could be the opposite of what was being suggested before which is to give access to education to those who might preview you know might otherwise not have it those folks that are potentially you know in a homeschool, microschool, non-traditional educational environments where resources are dictated by you know cash flow in the family or what you can find like what it seems to me like the ramp up for them is much quicker than it would be in in traditional and educational environments yeah no that's that that's true so I think that there's I mean there's concerns here and then there's there's some truth there so I think that we will continue I think that they will have the most use of it in the short term I think Khan Academy has seen that I think that their target audience seems to be a parent at home who wants to like help the student out homeschool audiences are definitely a great audience for the economy go-bot where it helps students navigate the content you know in a personalized conversational manner which that now could be the parent or the homeschooling and person can like serve as a guide and not as the not to be the expert on the content area should be the topic area teacher and I think that that will hopefully see more of it I think we'll see smaller schools make use of that where they don't have the teachings they have to teach a bunch of electives they might be able to find ways to integrate AI into providing different elective options for students so that they can pursue whatever intellectual interests the students have and so I do think that there's lots of uses for this in the small-scale setting my worry is that we start to see a massive gap between who uses it and who doesn't use it and then the literacy gap will get worse and so that that's this is where we have to be really careful about you know we we might see a bunch of smaller schools say you know what we already have this figured out we know what our student population needs and wants and so we're able to provide all those needs and so we're not going to currently think about how to integrate AI and then at the same time we might see a large public school district embrace the use of AI technology and that you know again that those can be at parallels in any sort of dichotomy there but the worry is that we start preparing some students for this AI future who begin to have those skills and are ready for whatever the world throws at them once these you know technologies come even more due to this and kind of start to change career paths and then we have students who have no exposure to the technology and so while I think that there's there'll be different needs that the technology can fail at different levels of schooling I just hope that there's at least universal adaptation of it adoption of it in some way so that we can all start to become more adept at using the technologies but you're right the problems that will solve for different school systems and different educational contexts are going to be vastly different but I think there's there's some role in playing everywhere I guess is my cautionary note yeah I agree with you I would I think that my concern is probably a subset of what you've expressed which is that what we see in typical educational technology models is the you know the move from freeware to subscription models and even at you know chat gpt you would monthly expense where you can put tokens in to get more to get a little bit more horsepower out of what you're doing my concern is that so much of this technology will end up privatized that the folks that really stand to benefit the most the ones who right now don't have access to the resources that say the students at you know the independent schools that I've worked at will not have access to it again because it's become one of those things that's driven by profit I love the fact that open AI has built its board or at least tried to build its board around its public purpose so that that tells me that there's at least in the room there's a voice that's saying we have to be thinking about you know not how much money we can make from this product alone but also what it means for the world and that's to me is reassuring I don't know that all of the tools are approached that way but I but I echo that I know I've had you for over 40 minutes and I want to let you go I just want to give you sort of one last sort of prompt something that I heard and I'd love to hear your feedback about it I was in the discussion about independent schools and artificial intelligence and a school leader who I will not I will not name although I think he deserves a lot of credit for what he said he was the voice of calm and this was over a year ago so you can imagine where all schools were everyone was sort of fluttering figuring out what this was and what he said was I work at a school that from a time from the time a child walks through the door our singular purpose is to help them find their voice if we've done our job this technology is nothing to be afraid of because our students will never give their voice back I thought that was so poignant and prescient brilliant and also predictive what do you think AI has in terms of the potential to help us help students find their voices as opposed as opposed excuse me to being a threat to that search yeah I think that is very well put and I think that that highlights I think the potential that I see also from technology I think the the technology has the potential to help us achieve our long-standing goals and education I really don't think that we need to dramatically revamp what we want for our students in fact I think the technology will help us better achieve what we want for our students and I think that one of the I think the best way that I think these tools can serve that purpose is by serving as the students personal thinking buddy I think that we all have top processes cognitive processes that are dramatically different from each other even after we have similar lived experiences folks have different living experiences have even asked me more different cognitive processes intellectual journeys interests and all right like the entire mind space is very different and so I think these technologies can allow those students to kind of tap into that and the example like you know it might be something as simple as helping a student go through a particular topic in a way that's most interesting to them and so that might look differently at very different levels but for a civics project at the fourth grade level might allow a student to learn about the US government by focusing on a different issue that's important to them and so you might have one student who's really interested in climate change and now gets to learn about all facets of the government and the civic process in the context of climate change advocacy but you might have another student who's really interested in thinking about economics and really interested in the the wealth and then their entire exposure to the field of civics might be through learning about how it's relevant to solving the wage gap and thinking about the wage gap and I think that you know that's a that's a minimal example of just thinking about how we might be able to you know achieve some of the same goals we want we want our students to know how to engage with specific process but in ways that are more engaging to our students in ways that feel more meaningful to them and help them kind of figure out okay maybe they thought they were interested in climate change and halfway through they realized okay after all this talk about it I kind of am much more interested in just thinking about you know environmental policies and not thinking about climate change advocacy and those kinds of journeys I think will be much more easy when we can have much more tailored content and tailored feedback to students own intellectual pursuits and doesn't have to be so systematic and rigid that necessarily has to be right now given the limited resources. Yeah I think that's I mean that's a great place for us to stop. I want to say thank you again first for those who are listening to this please seek out AI and the future of education teaching in the age of artificial intelligence you can find it on Amazon. It is a really valuable entry point into these conversations and I think that it's not just a great tool for educators folks in the classroom too often we we see the term teaching in the age of artificial intelligence and parents forget that probably the most influential influential teacher our children have are their parents and their siblings so it's a great purchase for anyone and it's an excellent entry point into this conversation. Prit and I thank you for giving us time I think I could have probably talked to you for another two three hours on this and I've got millions of questions that I didn't ask for fear of maybe boring some of the audience who haven't spent as much time digging into this as I have or certainly as you have but again I think it's a great resource it's an exciting conversation I look forward to seeing how it evolves and I hope that maybe down the road maybe season four of At The Meadow we can have you back and see if if we were on track at this point. Yeah for sure that'd be very exciting. All right thank you Prit. Thank you. All right that's the end of episode one season three of At The Meadow I want to thank Prit and Shaw for joining us in a very fascinating conversation for those of you who are interested you can find Prit and's book AI in the future of education on Amazon or in other locations that you buy books I'm certain that this will be a theme throughout this season of At The Meadow as artificial intelligence seems to be a part of so many discussions right now in our daily lives thanks for joining us and we look forward to seeing you back here at The Meadow.