AI, Literacy, and Equity: How CELaRAI is Shaping the Future of Education at UB

In this episode of Achieving Joy and Mastery in Public Schools, we tackle the urgent issue of early literacy and how generative artificial intelligence could be the key to transforming young readers' experiences. Only 33% of U.S. fourth-graders were proficient in reading in 2022, and this crisis continues to grow. But what if AI could help change the story?

Join host Brian Graham as he sits down with the team behind the Center for Early Literacy and Responsible AI (CELaRAI) at the University at Buffalo to explore how innovative tools like the AI Reading Enhancer (AIRE) can support young learners in developing essential literacy skills.

Our special guests include:

• Dr. X. Christine Wang, Professor of Early Education & Learning Sciences at UB, Senior Associate Dean for Research, and Principal Investigator for CELaRAI. Dr. Wang shares her vision for how AI can augment teacher capacity and address the literacy crisis.
• Dr. Kristen Smigielski, Project Manager for CELaRAI, who specializes in inclusive education and social justice. Dr. Smigielski discusses how AIRE can empower teachers and ensure equitable classrooms for all students.
• Dr. Zhaohui Li, Post-Doctoral Research Fellow at the National AI Institute for Exceptional Education & CELaRAI. Dr. Li dives into the technical side of AIRE, explaining how it uses AI to generate personalized texts and provide real-time reading support.
• Zhuoyun (Joy) Cai, Ph.D. student at UB and AI policy expert for CELaRAI. Joy explores the ethical considerations of using AI in early education and how policy frameworks are shaping its responsible deployment in classrooms.

Together, they highlight the exciting possibilities of using AI to improve phonics, word recognition, fluency, and comprehension, especially for culturally and linguistically diverse learners. We’ll also discuss the importance of responsible AI practices and how CELaRAI is leading the way in developing ethical guidelines for AI in education.

If you’re looking to learn more about how AI can help bridge the gap in early literacy and support teachers in their mission, this episode is a must-listen!

https://earlyliteracyai.org/

Find us on Twitter @NiagaraErie

Or visit our website E-NSSA.org

Transcript

0:04

Hello everyone and welcome to the Erie Niagara School Superintendent's Associations podcast, achieving Joy and Mastery in public schools. I am your host, Brian Graham, superintendent of the Grand Island Central School District. We are really excited that you're listening today. This podcast is designed to celebrate all that is good in public education around Western New York. We'll be featuring programs and innovative ideas that inspire and influence our students, faculty, staff, and community in new and exciting ways. So everybody, let's get started. Today on the podcast, I am thrilled to be here at the University of Buffalo with four amazing scholars who will be sharing important information with us as it relates to the responsible use of artificial intelligence for the social. Good of early literacy skill development. According to their research, America's facing a literacy crisis, only 33% of US fourth graders were proficient in reading in 2022. This dropped to 31% in 2024. Once children reach third or fourth grade, and the focus shifts from learning to read to reading to learn, those who are behind often struggle to catch up. In this episode, we will talk with the team behind the new. Center for early literacy and responsible artificial intelligence, I guess being, uh, people are calling it celery, which is very interesting. So, uh, we're gonna be talking to this amazing team and exploring how generative artificial intelligence might help turn the tide for young readers across the United States and perhaps the world. Our guests today are Dr. X Christine Wang. She is a professor of Early Education and Learning Sciences at the University of Buffalo. She's also a senior Associate Dean for research. She's the director of Fisher-Price Early Childhood Research Center, and she is the principal investigator for celery. Dr. Kristin Salki is here. She is the project manager for Celery and just achieved her PhD in curriculum instruction and the Science of Learning. Also, joining us today is Dr. Jaho. He is the post-doctoral research fellow at the National Artificial Intelligent Institute for Exceptional Education. And of course he's doing amazing work here as well. Additionally, we have doctoral student Joys Si who is, uh, studying for her PhD, uh, with respect to curriculum instruction and the science of learning within that program here at ub. So we're very excited to have these. Experts with us today. I do wanna give you a little background information about their mission. The Center for Early Literacy and Responsible Artificial Intelligence is funded by the US Department of Education's Institute of Educational Sciences. It's my understanding, they received a$10 million award, which, uh, if there was an auditorium here filled with students, we would all be thunderously applauding that accomplishment.$10 million for the University of Buffalo and these amazing people that are leading the way with ai. Truly remarkable. So the center aims to integrate generative artificial intelligence into early literacy education. It is developing. Air, and that would be spelled A IRE, and that is artificial intelligence reading enhancer. We're gonna be talking a lot about this tool and how it is in a prototype right now. So this student facing tool will be designed to generate personalized text, conduct realtime reading analysis, and provide just in time literacy support. The center focuses on improving phonics, word recognition, fluency. Comprehension and particularly will be designed to appreciate those that come from culturally and linguistically diverse households. So this is really exciting for me. I love what's happening here at the University of Buffalo. And Christine, we're gonna get started with you. I wanna welcome you to the podcast. Thank you so much. Uh, it's really, truly a pleasure. I've been looking forward to staying down with you, um, to share with you about what we do at salary. Um, so I know some people joked about our equity for our center. Gosh, my gosh, I know Salary Ray though. Everybody says Cel Ray. The, that's the proper way to say it, correct? Yes, yes, yes. But think about salary is really healthy for you. So this is why we kind of connecting to that, but we don't want people think about us as a. The veg vegetable lot people don't like unless you put a little peanut butter on it. That works. Oh, oh my gosh. So I really appreciate that. And I love Right. Sometimes, uh, these, uh, institutes that are created have, you know, long names and to be able to remember Cel Ray is a great way to, uh, crystallize this, uh, notion in our minds. And quite honestly, I can't wait to really dig deep into this conversation. But as you know, uh, this is the podcast Achieving Joy in Mastery in Public Schools. What is bringing you joy right now in your role here at ub? I'm so glad you asked that question. Um, what truly brought me joy this year is really witness witnessing the birth of something new and necessary. So we launched a salary. Um, it has been a incredibly energizing experience, so we have built not just a research center, but a movement, uh, evolves a community of educators, AI researchers, learning scientists, and the literacy specialist all coming together to reimagine what early literacy with AI as a partner, not the replacement can do. So I would say like when we go to local schools, including your schools, the local school leaders, the teachers. Told us when we describe the tool, the tool we're developing, they, the response would be they feel that the tool like Air is exactly the kind of learning tool they needed. To support K 12, uh, k to two students learn how to read. So that is just priceless. Yeah, I love that. I love that so much. And I think you are absolutely right. And I would also say that, uh, educators, particularly, uh, educators that are helping children to read, uh, to be able to incorporate a tool like air, and I know we'll talk about that mm-hmm. Uh, in the, in, in later on in this podcast, uh, as a partner, as a tool for the teacher, but also as a partner for the student. Mm-hmm. I think that's really gonna open up, um, some interesting and important, uh, pathways for kids to continue to grow their literacy skills. So, uh, I think we're, I think the center is on the right track, uh, with its research and with the idea that every child. Can have access mm-hmm. To high quality, um, partnerships with, with air, uh, throughout, uh, their journey in, in our classrooms. So with that said, I guess this might be a good time for you, uh, to share the concept, right. Of, and I think University of Buffalo has been promoting this quite a bit. Mm-hmm. The social good, right? The social good of ai, the social good of artificial intelligence and helping children develop such skills. And I'm just wondering for our audience, maybe to hear from you the mission and the vision and anything else you think would help, uh, as an introduction to this concept. I think that the public good when we really believe, uh. Learning to read should be, uh, a write it's not luxury. Should be a opportunity to learn, to read, to master. Um, the skill to read is this foundational skills. Um, and, uh, unfortunately, like what you said in the introduction, um, the national, uh, achievement in this area is really lacking. So many children are left behind and, uh, when you in the early grades, that is the time to learn to read and you are not learning and then you just much harder to catch up. So, and then those kind of, um. Uh, achievement gap, especially glaring among students, come from different cultural linguistic backgrounds. So our center is really trying to ensure every child will have the right tool to, to be supported. And again, this is important to point out. We're not. Uh, replacing, uh, teachers, we're really trying to enhance and expand what teacher can achieve. For example, this tool is specifically designed student facing tool to support children's independent reading. Um, so independent reading, you know, in the K 12 classrooms, usually that have a time for kids to read. Sure. Independently, right? Teacher will work around providing kind of support scaffolding, but. Oftentimes one teacher, 2030 students, you just don't have the bandwidth to cover every child and every child's needs is also differs. Right? Um, so the air is exactly designed to address that kind of needs personalized and as well providing just in time, uh, support. Um, so do you want me to Yeah. Go into more detail? I, I would love it about the personalization. Yes, please do. And the just in time feature. Um, so personalizations really trying to gather the information about what the child, who they are, what they like, their personal interest, some family cultural practices. Um, so based on those information, I love what you just said. Super important, right? I think we all in education. Uh, subscribe to exactly what you said, but to have this a little bit, you know, structured and a little bit more formal as part of collecting mm-hmm. Maybe a baseline Right. Of information for each individual child, right? Mm-hmm. Somebody celebrates, you know, a holiday this way, somebody else celebrates a holiday that way or celebrate different holidays, right? Right, right. So. Super smart. Mm-hmm. And a great way for that heir mm-hmm. To begin understanding the child, but also developing text mm-hmm. That is personally connected with them. I, I know that's exactly what, what you're doing. Mm-hmm. I just wanted to emphasize that. Mm-hmm. I didn't, I didn't wanna. Uh, have you, you know, get off track, but I just love that those elements are built into that baseline understanding of who Brian Graham is, who Christine is, right, exactly. Yeah, exactly. So that make those text reading material a lot more meaningful. Um, so that, you know, meaningful text materials, you support children's engagement, keep their motivation high. Um, the personalization also goes beyond this personal interest, right? Because each child learns read in different pace. They struggle in different areas. Uh, even phonics, right? Some kids depends on the exposure they might have. They don't have issues with the particular sound, but they might have challenge with different, uh, phonics patterns, right? So this tool is supposed to be, we. When we complete it, um, this tool will be able to, um, based on children's pro progression and tailored the difficulty level leveling based on their, uh, progress. So that's another layer of personalization. This is a much harder way of personalization, right? Oftentimes when teacher does the classroom instructions, um. You usually based on the overall pattern of classroom learning, majority of students do not know. So we can target that particular content, but some kids more advanced might feel, might find it boring. Right. They are disengaged. Yes. The differentiate Exactly. To be able So differentiate the differentiation. Exactly. That's part, part of the personalization, uh, concepts. Yeah. Christine, this is so exciting to me. So help us, uh, understand, uh, where, um, in, in the journey of the research and, and the prototype. How you envision this prototyping used first by, uh, children and, and at what age would it be preschool. Would it be before preschool? Mm-hmm. Or would it be around kindergarten? So the target, uh, uh, grades is a beginning reader. So K to two, um, ages five to seven. Okay. Yeah, because that's the time really to learn to read of course stage. Yeah. And, uh, we specifically, um, target this, um, instructional time we call independent reading time. Make this tool more, can be more easily integrated into classroom because each school district, um, different regions, they have very different kind of literacy, uh, curriculum. Um, so when you. Design a brand new curriculums, much harder for the school district to integrate, but independent reading time, oftentimes there's not enough, um, material attention paid to. Um, so we feel we are targeting this relatively weaker, uh, link in the literacy instructions. Providing this high impact tool, uh, allows, uh, maximized learning opportunities for students. So given that age range and starting, um, and I know it's a prototype and I know that it's. It's gonna be further enhanced through research and mm-hmm. Practice. Do you envision this tool to be, say, web based and access through a mobile device that school districts have already purchased for children? Or it will be, uh, have its own hardware that might be available or a combination of both? I think it's some technical type of things. Yes. We will leave that to, uh, Dr. Lee. Okay. Perfect. Uh, but the general plan is, um. We would design this as an app? Yes. Uh, be that would be web based. Yes. So, or, uh, particular, uh, we can be easily integrated into the existing, existing school technical, uh, platforms, uh, depends on each school's, uh, usage. I don't know how that can be done, of course. Course. And Dr. Lee answer those questions. But the plan is trying to, um, make this, um, on the technical side, make the threshold of integrating into the existing school system much easier. And also because it's a federal funded project. Yeah. Um, this tool will be open source, uh, freely, uh, offered to any schools who would be interested integrating. That really is amazing. Yeah. So, um, so we talked a little bit about differentiation, right? Uh, each family's different. Each child is different. Uh, I have a child that. You know, who's now an attorney, but when she was a young person, she was reading, you know, at, you know, advanced for her age. Right. So would the, uh, heir tool be appropriate also for a child like my daughter, Maura, when she was five, as well as a child who may, may not even, uh, recognize the letters of the alphabet? So I guess the most kids starting kindergarten have some latter senses. Latter sense. So I would say, um, the targeted probably is mostly for appropriate for kids already start to recognize letters. Yep. Okay. Which is, which is good. Because when you start K, that's the basic expectation, I guess, at the very beginning. Um, right. Because if the phonics is part of focus, the letter can be built. Letter sense couldn't be built into it. But so far the focus is still mostly decodable text. Yeah. As, as a starting point. So they should be able to. Have that correct base knowledge. Yes. Which is perfect. Yeah. But, but to that end, could somebody who is already putting words together and somebody who isn't mm-hmm. Still use the platform at the same age? I, I believe so. Okay. Good. There will be different levels Yeah. Of, uh, perfect scaffolding. Right. So when you talk about personalization on the material, what we just discussed so far is the tax generation basically generating personalized materials. Yeah. But there's a personalization through the second, third part is we, um, diagnosis based on children's engagement with the material, identify the type of needs needed for different children. Yeah. Right. Some children need more support on the spelling side. Some children need more support on the background. Yeah. My daughter was the, the spelling side for sure. Yeah. She's an attorney. But right now, you know. She was such a voracious reader. Mm-hmm. But putting, you know, uh, the spelling part was, was a gap in her period. Right. So those other functions, the tool when is complete will have this deeply personalized and also the just in time kind of feature. But of course we are still in the early stage. Yes. Where that's really just the first year. Correct. Right. Uh, first year, us, as you know, those are large, uh, federal funded grants. Takes time to get this whole thing started. Even the count set up took a few months to set it up. And also this is a large team, uh, UB is the lead institution. Yes. I was gonna ask you, we also have, um. Uh, co-PI, uh, co-principal investigators, we call co-PIs in other five, uh, institutions. Um, they include University of Illinois, Atlanta, Champaign, uh, UCLA, university of California, Los Angeles, and Stanford, um, Michigan State, and also, uh, east Carolina University. I love it. So 10 pi co-PIs across six institutions, uh, including ub. So UB is the lead. Yeah. And as well, it should be. Well, Christine, uh, we're gonna come back to you, but you really helped, uh, set a foundation for the conversation today. And I'm, I'm just really, uh, thrilled now to talk to Dr. Salki. Kristin is here. Uh, so Kristin, you know, we, you heard us begin the conversation about joy. What is bringing you joy in your new role? Awesome. Well, I just wanna say thank you so much for having me. Um, it's really meaningful to be here today. This work is so close to my heart, so I'm honored to be part of the conversation. Um, something that has been bringing me joy is a great question. Um, kind of reflecting on the end of my PhD journey to get, um, defend my dissertation and everything, but also remembering my years in education. My, the purpose of why I was doing what I was doing as a teacher and now, um, working in higher ed and working at this wonderful, um, on this wonderful project is that I have been so excited to have the chance to support and nurture future educators, um, in the classroom. It brought me so much joy working with students every day, and I just wanted to help more and be there more and support teachers. And so having the chance to support, um. These teachers as they grow into joyful, responsive, and inclusive, uh, educators. And also as we're doing with our tool at Cell Array, um, to be able to provide more opportunity to help the students and help them learn and grow. And so I've also been so filled with Joy, working with such a fantastic group of educators and researchers every day. I am so thankful, um, to work with the team here at UB across our institutions and also incredible district leaders like you. Um, it has just been wonderful and I'm truly very honored. Yeah. Well, that, that is great. And it's just a great way to celebrate the work of, uh, all that encompasses, uh. The potential exciting growth of early literacy, skill development, uh, for everybody in the world. So this is super exciting. So, Kristen, just a little bit about your journey. Um, what inspired or influenced you to pursue first, you know, a PhD in this area? Uh, how, how did you come here? What was your, but you were a teacher, right? Yes, yes. Yeah. Talk a little bit about that. Absolutely. I'd love to share. Um, so born and raised in the incredible Western New York area. Um, and I come from a family of educators, uh, especially my mom who was a educator in the Buffalo Public schools. That's great. My son is teaching at the Buffalo Academy for Visual and Performing Arts. Oh, amazing. He's in, he's starting year four, so. Wow. I I love that. I love that. Your mom, uh, was a teacher. What was her specialty? So she did, um, English as a second language. Wow. Wow. So she was working with. Student she had, um, yeah, probably sometimes. So many languages, right? Yes. And she was the first ESL teacher, um, at one of the schools that she worked with. Uh, and so she had students from all over and, um, so many different languages represented in one classroom. And so she just approached it with joy of how are we gonna do this? We're gonna figure it out, um, as we go and helping all these different students find their strengths, find their, what they're great at, um, and then filling in those gaps so that they can, um, reach their highest potential. She did so many wonderful things of, uh, speaking contests. And, uh, shows and all of these things. So really having that passion instilled from a young age of taking students where they are and growing, finding what they're great at, finding where their strengths are, and then nurturing that growth along the way. So that kind of very instilled in me. Yeah. Um, the passion that I have and the DNAI love it. Yes, it's there. Um, so I, um, am professionally certified in New York State for childhood, early childhood and music education. I also, oh, oh, music too. And music. Yes. Really? Wait, wait, wait. You're going too fast. Okay. So when you were entering college, where did you go? Mm-hmm. So I went to Buffalo State. Yeah. Yes, for music. For music, really I wanted to be music. What specialty, what an instrument? A singer. So voice. So voice, singing. Nice. I wanted to be, um, a music teacher. I originally started, I wanted to be a music teacher, um, at university. And then I kept going down to middle school and then interesting childhood. I found that interesting. Found that passion. Yeah. So music was my start. Just a creativity. That passion. Yeah. Yeah. Love it. I was really excited that I would have students of all ability levels in the same classroom and I was so ready to, to go forward and see how I can help everybody grow. That's fantastic. Yeah. So I, um, started my teaching career at Buffalo Public Schools and, um, especially teaching the middle school students. Yes. Wow. I wow. Witnessed that. Um, the literacy crisis of every, like the challenge of, of so many different abilities For sure. Um, and so I felt that strong pull towards intervention, which brought me down to the early, early years. Um, so I transitioned into general education, um, classrooms. Before I did that, that music teaching. So I went to Buffalo State and then I did my master's in music at ub. Wow. I went back to Buffalo State for my master's in childhood education, um, before I went back to university at Buffalo. So I went down to general education classrooms with my last two years being in kindergarten, which was my absolute favorite. Um, because kindergarten kindergarten's such a powerful place to be. Yeah. You are building that foundation for learning and literacy, but also the most important work is helping students fall in love with their learning. And was this still, was this still in Buffalo or did you transition? So I transitioned, I was in Amherst and then I was at Elmwood Franklin my last year. Was this Nice. Good for you. So building that foundation for learning and literacy was just the best part. Um, and my, it, I found so much joy in meeting students across their wide ranges of abilities and experiences and showing them, um, especially those who. Couldn't yet read or thought they couldn't read. Um, even those that didn't know the difference between numbers and letters. Um, taking them and showing them that they are capable and clever and worthy, um, of joy in their learning was just the most wonderful thing. Um, and watching that belief in themselves grow was the most rewarding. Yeah. Perfect. So, um, you can't see on the podcast, but I'm smiling as I talk about this. This brings me No, I think we can all tell, brings so much joy. So, so Kristen, so you know, you mentioned a few minutes ago mm-hmm. About being at, uh, buff State, maybe for music and thinking you might. Be teaching. Mm-hmm. At the university level someday. Yes. And here you are. Yes. Yeah. So that, that's a big jump, right? Mm-hmm. From kindergarten, you know, you all of these levels. Mm-hmm. Now, right down to kindergarten doing all this wonderful work and then having that, remembering that desire to get that PhD and be here at ub. So did you come to UB for your PhD? Yes. So, um, while I was still in the classroom, my last couple of years in the classroom, I started my PhD in curriculum. Wow. That's great. Instruction. Signs of learning. I just wanted to, there's so many wonderful opportunities in the classroom, but there are also some challenges and I wanted to, um. During my first couple years of teaching, but also my last couple years, I wanted to make sure I was equipped with like the most current research and best practice. Yeah, makes sense. I wanted, I wanted to learn from as many people as possible, um, not just from my own classroom, but also a broader impact because our community, um, is incredible and big, but also it's across the us, across the world, like how are we changing education, making it better every day. So after five years of teaching, I transitioned into higher ed. Um, and I've done work over the past year coaching resident teachers, um, doing instruction for pre-service educators as well. And thanks to Cel Ray, I've been able to expand that impact even further. Um, so working with. Not just one classroom, but like I said, uh, multiple teachers pre-service, like working with all of these wonderful, incredible teachers and resident teachers that are going into the schools and get to do this incredible work. Um, having that ability to work with them and see their passion and see their joy and cultivate that for them before they go into the classrooms, but also work on tools like air that's going, that's making a huge difference. Yeah. Um, going to make a huge difference for students has been amazing. Yeah. And actually that helps, uh, segue right into that concept. Um, you know, you are the project manager mm-hmm. Right? For cell array. Uh, you have a great background in education, inclusive education, you know, how do you see, uh, air supporting teachers? Teachers want every student in their classroom to learn and thrive. Um, from that basis and understanding, we're also often limited by resources, people, and especially time. Uh, one of the biggest challenges I've seen is during independent reading time. So while teachers are rotating through their small groups leading strategy instruction, doing incredible work with small groups of students, it can be really challenging to ensure the rest of the class is truly engaged with the meaningful, appropriately leveled and supported text. Um, and making sure that students are getting the personalized support they need to learn and grow. So a. When they're doing their independent work teacher, there are only so many teachers in the rooms, and I see every single classroom I go into, teachers are giving 110%. They're giving it their all, they are showing up, pouring everything they have into their students, uh, but often wishing they still had one more set of hands for certain times during the day, or one more tool to help reach every learner. So that's where I see air coming in because it's not about replacing teachers ever. We're never gonna do that, but it's about augmenting and elevating what they can do, um, and that elevating that incredible work they're already doing in the classrooms. Um, so for example, kindergarten story, um, I had a student that started kindergarten again, not necessarily knowing, um, all of the letters or letter sounds and recognition. And so it was a journey and a fun journey, um, with all of the games and literacy learning tools. But when it came time where, um, this student was reading ready to read. Simple sentences, short books, decodable books. Um, he was super interested in space and spaceships and planets and stars and. It was so challenging to find as a teacher texts and books that were of like, this is a great example. I love this. Yes. And it took so much time to going on the internet and Googling and using, trying to use other platforms that were out there to really create at this specific level that he's at, at with these abilities and these le this letter knowledge and we're working on this skill. It was so challenging to find that because I was limited by the books that we had in the school and what I could find online. Um, and so it took a really long time to, to create a, a personalized text for him. But once he got it and he read it, he was the most excitement I've seen in the classroom. It still makes me smile to this day, but he was so excited that he could read it and was reading it to everybody in the class and took it with him down the hallway to read it to other people that he could, he was reading. That's such a great story, but it, yeah. And did this, this story occur. In your role right now at ub? This happened, um, when I was a kindergarten teacher. Got it. In 2024. Right. So you're already thinking about these ideas, right? Yes. How to personalize. Mm-hmm. And the amount of time, effort to differentiate. Mm-hmm. I think is a really, really important concept. I. I have the joy, mm-hmm. Of interviewing every teacher in the final interview in Grand Island. And I will guarantee you that I always ask how they would create a unit of instruction mm-hmm. That's differentiated for their students, and I'm finding. That there are gaps in their thinking about this, and it's just so great that here you just told this great story and, but you're also illustrating the amount of work. Mm-hmm. Right. It's, it's a, it's a, a lot of work and to have a tool like Air will lift a little bit of that. Work burden off so that the teacher can focus directly on instruction. Right? Yes. And I think that is a great message for our listeners. I think that's one of the powers of AI and, and the social good. It'll provide not only for children growing, but for teachers. Mm-hmm. Making their lives just a little bit easier as they do exactly what you said, how you started off this conversation, you know, wanting to. Every child to reach their fullest potential. So I love that story. And so then you saw just by all that hard work mm-hmm. And you doing all that research, this child gets so excited about being able to read something that they are truly interested in. Right. That's such a personalized approach. I love it. So, uh, and then now you take that idea and concept and you're obviously working on this great team and trying to mm-hmm. Duplicate that right. Through artificial intelligence. So Yes. That's super exciting. Exactly. Um, and I, I think it's incredible'cause every student is so different. Interest home languages, cultures ability and strengths, and having a tool that responds to that diversity, um, and meeting each student where they are and generating that personalized, um, culturally responsive text and offering that just in time liter literacy support is to me the most incredible thing. So if I could do what I did with that one kid to everybody in a classroom at the same time, that was to me seemed like impossible just a year or two ago. Um, and so being able to work on something that is helping all students access, um, rich, joyful literacy experiences and giving teachers the support that they deserve to do the work they love, to me is just the best thing. Yeah. Yeah. So well said. So, um. Before we, before I transition to Dr. Lee, I'm, I'm very curious about, uh, the work you and your team are doing when, um, engaging with, uh, classroom practitioners and perhaps sensing or getting some feedback on, um, uh, maybe, uh, the challenge or obstacles in their own lives for adopting ai. Mm-hmm. Is that something you're, you're focused on in helping bridge that gap and maybe address some of those challenges that teachers may have about ai? Yes, absolutely. Um, that's something that we've, um, done this past. Um, this past year we've been in schools listening to teachers, their stories, their experiences, observing classrooms, um, doing focus groups and interviews one-on-one to really understand. The specific need. So we know this can be incredible and wonderful, but if we're not building it to the specific needs of classrooms, then we're not doing the right work. Um, and so teacher's needs, they're incredibly varied and there's no one size fits all approach. So. What we're doing is taking the journey teachers take every day, which many go above and beyond, um, despite the limited time, the training or the tools and creating something where teachers can feel empowered and not overwhelmed. Um, and so really listening to their stories of they wanna do this work, but they don't have the time. They're learning so many new things. Um, so we're trying to make tools that are intuitive. Um, oh, I love that you used the word intuitive. I think that really is, uh, so important to break down any barrier mm-hmm. Right. With technology that it's just easy to use. Yes. And you just touch it, it doesn't break, and off you go. Yeah. Like, I, it's such a simple concept, but I think it's so important. And I was actually going to ask, uh, uh, later about that intuitive, uh, connection. So I'm glad you brought that up. That's great. Yes. Thank you. Yeah, it's, um, we really wanna make sure that it's an easy. Thing to transition into a classroom. Teachers are already doing so much wonderful work that it's very student facing this tool. It's exciting, it's fun, and teachers can input the data that they need and then trust and feel safe that it's gonna continue this amazing work. So teachers have so many different opinions, and ideally they all wanna be able to trust the technology and know that it's safe. So including that, um, transparency and reliability and the clear alignment with student growth is at the core of what we do as well. Um, but also working with teachers, um, they, they need to see how AI can help. Not just, it's a great thing, but specifically what is it? How does it. Relate to your classroom, to what you're doing, knowing that it's never gonna replace them. Um, but to continue assisting and augmenting their relational, inclusive work that they already do, I think is really important. Yeah, that is so well said. Now, as I, before we transition, I, I'm just thinking about any advice you might have right now for educational systems like Grand Island and others. Mm-hmm. Are there things that we should be thinking about as we prepare to integrate more AI into the classroom? Are there systemic changes, systemic, uh, thoughts you think we should be thinking about? I think that's a great question. Um, thinking that AI is coming, AI is here and is very Yeah, pretty much it's already here. My first time using what I knew was AI was chat GBT in spring of 2023 was the first time I tried it. But then learning that we have been using AI and machine learning from playing tic-tac toe against a computer to Netflix recommendations, um, kind of understanding that it is just another new thing that we're getting comfortable with and using it in a different space. So being open to ideas, um, being open and, and, and hearing when people are concerned about it, um, but also being excited about and learning and knowing what it can do. Um, so it can't do everything and. If we are prioritizing that relational aspect of teaching all of the, the good things that teachers are doing, how can it elevate what they're doing? How can it stretch what they're doing a little bit further? Um, and kind of providing those moments, uh, for teachers where they can, they know that great things are happening. So I think just being, um, open to ideas, listening and learning, keeping up with it. Um, hearing all of the different sides, the excitement, the concerns, everything, but prioritizing student learning throughout is, is the best. And you do such a great job in your district. I've heard your Oh, thank you. Your stories and everything. So, um, yeah, I think that's good. So from the school district perspective though. Mm-hmm. Um, I, I know that you've probably heard, uh, over the last year at law 2D right. I'm sure this is baked into all the work you're doing. Mm-hmm. That, that the data that's being collected by air right is, is going to be protected. Yes. There have, that is one thing that I think is. Sets education different from other industries and things like that is we have student data and student privacy, and so we have been working diligently, um, and also partnering with a couple other districts to really make sure that we have the safest options possible. So making sure that we, while we have this tool that can be deployed nationally, um, all student data is, um, locally stored, locally housed, um, for each school district. Um, things like that they're working on. Dr. Lee can expand. Yeah. Oh yeah, for sure. Everything. Um, but making sure that it is safe, super critical. Yes. Yeah. Mm-hmm. And, and of course, uh, districts may require a signature, you know, that indicates right. That mm-hmm. That that data is gonna be protected. Yes, absolutely. Which is huge for everybody. Uh, so Kristen, that was so well said. Uh, thank you so much for giving us that overview and emphasizing, right, the importance of systems and how systems, uh, can adapt to the needs of their children through technology, of course, but also ensuring the, the privacy that, uh, uh, that is so valued when we're working with, uh, all of our children and staff. So, Christine, that made me start to think about timeline. Uh, maybe just share, uh, with our listeners, uh, you know, you applied for this grant. You're in year one. I'm assuming it's a five year grant, but I don't know, so maybe you can share a little bit about that. Yes. Uh, it is a five year, um, grant. Um, so the. Research and development of the air tool is one of the central mission of course. But beyond that, there's other things we need to do as well because, um, this one of the four IES funded national centers, um, all related to AI in different content area. So our, ours focus on early literacy. So, um, the year one, the main focus is really the needs assessment. Kristen touched on that we're Yeah, it makes sense really trying to tailor, uh, the design prototype and the meet. The needs in the classroom, the needs of, uh, teachers, the needs of school, understand it side of things. That's why we're also interview school leaders. Uh, you are included. Um, and we also have a broader. Team, uh, uh, advisors. Um, yeah, we have researchers on our advisory board. We researchers including AI researchers, literacy researchers, and uh, uh, learning statistics for the research design side of things. We also have a practitioners, um, we have IT leader, for example, Juliana, uh, from Buffalo Public School is on. I did see that. Yeah. That's great. On our advisory board. Love it. Um, we have, uh, Mary Howard. Mary Howard, yes. From Grand Island. A teacher who is early AI adopter. Yeah. Very early. Exactly, um, uh, on our board. Um, so we covered this wide range of issues. So year ones really the needs assessment, the prototype, typing the tools, and the year two is mostly focused on the development, the user studies of the tool. And the year three, uh, is we are going to conduct a efficacy study. Oh yeah. Makes sense. Uh, deploy for school year, um, to see effectiveness of the tool and also really identify the conditions, the context and what works, why it works. In what kind of context to whom? Incredible research. Right. So, and then year four and five is really, um, playing a leader, uh, a leadership role in the nationals, uh, level, um, to support schools integration. Of the tool. Yeah, yeah. Right, of course. Um, so think about the questions you asked about what school need to be ready. Yeah. For this kind of tool to be deployed, we feel the timing, you see this current administration really set ai. Education, AI in education as one of the strategic priorities. Yeah, I saw that. Yeah. For K 12 schools. So I think the timing is perfect. Um, so hopefully right now, actually everybody is really trying to address this, responding to this needs. So there's a lot of professional development is starting to support teachers to gaining understanding about AI tools and how AI tools can be integrated to improve the instructional innovation that also school-wide statewide, uh, effort to, uh, put place, um, policies. How do we think about AI can be safely, securely deployed? Yes. Uh, while protecting children's data, uh, security data safety and protect their privacy issues. So that's a, a huge need on the policy side. So that's one of the, um. Goal for our center is developing this kind of guideline. Yeah. Related to like Ejo might be doing that. Exactly. Responsible AI use. So that is part of our national leadership role, um, to address that. So I feel, uh, for school we designed this to try to meet where school are, but school also have all those things happening. National level, state level, school level is gonna be perfect timing when this tour will be ready to be Depo fully deployed and widely adopted in year four. We think the timing will be great. Yeah. Oh my gosh, I thought that, thank you so much for sharing that. That helps us, uh, really, uh, begin to see that in a few short years. It kind of sounds like it's a long way away, but not that we'll be deploying and utilizing great tools like this all across the United States. Alright, so Dr. Lee, it's such a pleasure to meet you. I understand you. Just came back to Buffalo from Italy. Yep. Oh my gosh. How was Italy? Yeah, this is great. Um, I love this country. I love visiting Roman and Sicily. So the conference in Sicily, um, and I love the, the, the CEO there. Yeah. Yeah. So did you were a presenter at a conference? Yes. Fantastic. I'm a presenter with, I present, uh, the Air Prote prototype there. No way. And I got a lot of useful feedbacks and, uh, we come up a lot of ideas that's kind of related, the recent joy you asked for. Yeah. And I, I definitely want you to focus on that. What is bringing you joy and maybe just coming back from Italy is all you have to say. Yeah. But go ahead. Tell us what's bringing you joy. Yeah. So the most recent joy I have is, uh, going to the conference. Yeah. So, I mean, our beautiful country and talk to people Yeah. And travel around there. And I really like it and especially when talk to people from different discipline in, from different, uh, communities like learning science. I'm a computer science people, so I allowed to, um, collect ideas from the people from other educational areas. So, so that made me very excited. And, uh, yeah. A few years ago when the AI Institute opened for Exceptional Ed, I was invited to a retreat and it was like three days. And you had, UB had all these wonderful presenters from all over, all over the United States. So I definitely would find joy in exactly what you were saying, being amongst other computer scientists from different countries with different ideas that has to inform your work in such a positive way. Yeah. And different culture background. You were able to present this prototype. So just let's, let's get into it. Um, in this presentation, I mean, are you allowed to share a little bit about what this prototype is able to do at this, at this point? Uh, yes, because, so I show the demo in the conference and people interested in how, oh, how you do that. How can you generate such a, a story and how you generate stories, how you generate text, how you generate figures and narration voices. That's, uh, yeah, yeah. Oh, wait, so I didn't even know this. So besides generating text, you said it, it, it can narrate as well. It, it can speak to the child. Yeah, that's great. Did I spoiled our research? Um, I mean, yeah. No, I love it. That's great. And, and can it create images to make the text more enhancing and exciting and engaging for kids? Yes. That's, uh, part of our goal. So our research, my research, I'm, uh. Uh, I'm, my research is about large learning model using generative AI to create, um, text and images for this pro, uh, this project. Um, and now as we all use, since we other people are using chat GPT and we can use chat GTP to G PT to create the lesson plans and figures and everything for you. Yeah. Yeah. So, and it, isn't it amazing how just those tools that are more common for educators, uh, have grown exponentially in its power, in its ability to respond to the appropriate, uh, prompt. Right. I, I, I dabble with Chache Tea, Claude Perplexity, Gemini Notebook, lm. I'm like, I'm in all of those spaces. And so I've seen this incredible growth over Yeah. You know, the last year. And so now. Yeah. You and your team, you're building your own, right? You're really, you're not partnering with open ai, you're building your own ai. Yes. Um, so I'm currently, our team is very interdisciplinary team. So we have people from computer science to build backend algorithm, like you said, train a large, a model for generative for, for generating the text. And also have people from HCI, human computer action to do research how we are gonna interact with the students. That's, uh, also one interesting thing. And also we also have the amazing learning science team, um, for us to, uh, to guide us. They have rubrics, they have pedagogies, learning pedagogic goals and how we can make AI to generate the thing that they aligns to, aligns to Pedagog goal. Yeah, and that's very crucial point. That's the differentiate between the air and uh, the, like the general AI tools. Select chat GPT. Yeah. Yeah, for sure. Alright, so I know, and you've just illustrated it, it generates text that's decodable and tailored to the child's interest. Does it also anana, does it do a analysis? Does it look at the child's ability to read aloud to the app and then have some analysis that's usable for the teacher? Yes, our algorithm can do that, but. Currently and their development. Yes, of course, of course. We need, that's the goal. Yeah. That's the goal. Yeah. We need the rubrics from the, um, learning science team because, um, we want, we don't wanna the, we want a teacher and the, the children, uh, the, the student trust our tools. It's not about, I'll generate some random text and that makes sense. Uh, see if them is, no, we wanna. That's assessment to be, um, can be trust. It's like we don't wanna have some ethic issues or some other things that we want we do on our, um, tool is reliable. So yeah, it makes so much sense. And because you're right here at the University of Buffalo and there's so much research, you're gonna have, uh, plenty to inform that analytical aspect of air. Uh, one of the other powerful messages today is that it's gonna be linguistically and culturally, uh, sound. Can you talk a little bit about that and, and how is that being baked into air? Um, yeah. So the current, I think for example, um, I'm from a Chinese background and I, I'm, I'm, I was born in China and if we wanna generate a story about dumplings, so. Maybe the people from Italy, they, they have different dumplings. Yes, of course. So there are two different dumplings. Yeah. There's, uh, the cultural difference, right? So how, how AI do that is we have a thing called prompt. Yeah. So we input those cultural, the, the students profiles with their cultural background, their age, their um, their ages, their um, cultural backgrounds, and also the, their assessment scores. So the ai, we put that in the prompt. We tell AI here, oh, here's the student's information and here's our learning goals. We wanna generate our, like, like that, uh, for this case. And the AI will, will process the prompt. Yeah. Love it. And I put the, the, yeah, the right following result, like the. GBT, um, and Gini. Yeah. The way that works, for sure. Yeah. So clearly the prompt is the key. It's always the key in with ai. Have you have, are you considering having an agent that interacts with the child so that it better refines the, the input so that it can, uh, provide the proper outputs, uh, that are really, truly personalized and differentiated for children? Yeah. I, I think this is a great idea. We are already doing that now. So now in our, um. Background algorithms. We have a chat bot that can interact with the students. They, the chat bot is listening to the students', um, questions. They can answer students', uh, questions. And some also can do assessment as well. Yeah. So super smart. We can, yeah, we can have, in the future we have, so the child will just interact with this chat bot. The chat bot will be. You know, available can talk back and forth with the child, but it's also gathering important information that's based on their interest. Absolutely. Yeah. Maybe their cultural background. I love that. Yeah. And, uh, we are considering have different agents for different tasks. Yes. Like one agent to talk with the student, to have them to try to make them focus on the, the reading. Yeah. And one could be like doing assessment in the background and uh, really Oh, that's even more clever. And that makes sense too when you think about how AI works, right. Uh, to be able to tap into different agents behind the scenes to work and gather and, you know, do the analytics while the one agent is guiding the, uh, and facilitating the, the process and not having to worry about the analytics. That's really, really Well, I'm telling you. Yeah. You're the po, you're the. You're, you're in the postdoc, which in June, so you already know how smart that is. I just think that's great and I think if we have a lot of fans listening who are AI enthusiasts, they'll, that'll resonate really well with them. Yes, absolutely. And also the current AI is still artificial. Artificial, yeah. So it's not generative. Right. Is that where you're heading a GI? Yes. A GI. So we are currently very far from a GI now, I have to say not we at UB or we in the world. We in the world, yeah. Yeah. So, but people are saying in the next five years, 2029, uh, Ray K Wheel is saying 2029. Some people say next five years. Yeah. What do you But you think we're far, far away. I feel that's hard because uh, currently AI is most of engineering and data collection. Yeah. So we collect data. Yeah. We make the mix some from engineering and we generate the result they want to see. Yeah. But uh, for our test, for salary test, yes. This is huge. Yes. This is difficult because we are like the a r auto speech recognition is hard. We are dealing with the little kids. They have different accents. Sometimes, sometimes they say things way. Right, right, right. Yeah. And that's the importance though of having different agents. Right? Yeah. Because one agent can't do it all. We need to tailor the, the large term model to different agents and to handle different tasks. Yeah. It's super hard. It's, uh, it's include a lot of manual works. Um. That's, but that's all right. Yeah. Um, I think we can do that. Oh, I believe definitely. We have a really great interdisciplinary team. We have computer science scientist, we have, uh, learning science people that Yeah. I believe we can do that. I love it. Yeah. So Dr. Lee, so we were just like having a little bit of a tug of war with the thinking that there are so many, uh, people out there that says that are saying a GI is coming within the next four years or five years. Uh, I hear that from different, uh, leaders. Maybe it's just a marketing ploy for, uh, anthropic or for open AI to get you to get super excited and get more investment. Right. So talk about, uh, for our listeners, uh, the what should we realistically be expecting and how is it different from, you know, we're not really, this isn't magic, right? Uh, you know, no. Talk a little bit about that. Yeah. AI is not magic. AI is not, um, I can't help you to do everything. Think about it. If you, it's like using ai, it's like, um, you're shopping in the shopping mall, so if you wanna do wanna shoes, you go to the shoes AI shop. Yeah, that's right. If you wanna I like that. If you wanna Yeah. Eat, you can even food can just go to the cafe person wearing Adidas kicks today. Right. So she Oh, you guys are awesome. Yeah. So you're right, you're right. You're in this big shopping mall and you're going to these other agents to get help. Yeah. That's, this is like, um, AI have different functionalities. So if you're going to, um, do education, teachers can use AI to generate the things they want. But, uh, other than that, it's, I mean, my point is it is a designed for a specific task. Yeah. So if without task we can't say AI can do everything. This is, uh. AI is not like, uh, everything. Yeah. Not a wizard. Yeah, not a wizard, not magic. Yeah. I, I know a little, I, I'm a closeup magic, uh, amateur personally, so I totally understand what you're saying. I think this is a good thing.'cause, uh,'cause when you using AI education, we do want teacher to involve, we don't, don't want to, uh, we want to that you student focus on the instruction and, uh, yes. Don't be, um, distracted to the other things. So yeah, it's well said. Um, yeah, we, we had to limit it AI for. For the, a test, specific test for performance and for, um, security and uh, uh, and their learning goals in our task. Yeah. Yeah. Yeah. Wow. I'm so impressed, uh, by you and your team. Uh, thank you so much for, uh, spending some time with me today. We're gonna be transitioning to Joy in just one second. So Lee, thank you again, uh, for first being here after traveling from Italy, you're probably, uh, you probably need to take a nap soon and just get adjusted to the time zone again. So I definitely appreciate you being here and now I'm super excited to talk to Joy. Joy, uh, welcome to the podcast. Thank you for having me. You are absolutely welcome. And it was, uh, really nice that you waited 57 minutes to get on the podcast. I really appreciate that. It's always deserved too, hearing those aspects. I definitely enjoyed, uh, you, you were holding a focus group, uh, was it? This week flies. Yeah, we just finished it. Was you? Yeah, it was with me and my, uh, elementary principal. And it's just, uh, it's just great. You're gathering a lot of data. You're, um, pursuing your PhD. Uh, but before we get into all that, just what's bringing you joy, uh, in your role here at ub? It's really fun. Like ask the joy for the joy. Yes. Yeah. So for me personally, I really like swimming, swim. So swimming bring me joy. Yeah. Because I think it's really connected with my, a lot of work, like including my research work. So the feeling like you delving into the, the. Swimming pool, just like your, you know, the pursuit, your goals. Just like what I did for the salary right program, I lead a policy review, and so if I use the metaphor to describe this, I would say like swimming, the process is like the leading the, uh, policy review and the swimming, the watering, the swim pool is just like this kind of technology enriched environment. So it's really. Super excited. And you find your goal and value during this process. Yeah. That's great. It's, my wife was a lifeguard. Oh. And when we were in college, uh, she, we were at Damon and she would go across the street to the Y and swim. And one day I said, oh, I'm gonna come swimming with you. And I was not a good swimmer. She just crushed me in the pool. Uh, but swimming, it can bring joy. I totally understand what you're saying. Mm-hmm. And so give us, uh, you know, a little bit of background into your role with Cellar Eye. Okay, so I'm the research assistant working for our salary program, and we have three different kind of group and I'm, I'm in one of the group called AI Ethics Team and also collaborate with, uh, Stanford team. So we try and Oh, fun. Yeah. That's, that's super cool experience working with Stan. Yeah, yeah. Working across the country with Stanford. Right. They have such a great reputation. Yeah. And it's, your focus is ethics and policy. Yes, yes. So we are trying to like, to figure out what is our, so there is like a concern when we are using that AI tools, right? We were like considering about the positive side, but we are also consider about the negative side. So we're trying to like, to build a safety guard, like to prevent those negative influences for the even, uh, for the young learners. So we're doing the policy review and try to figure out what does the AI ethics means in existing policies and the way we should do, like, incorporate those connotation and values into our ai. Uh, creating process. Yeah. Yeah. So I love that the title of this institute, uh, includes responsible ai. So maybe from your perspective, you know, what, what should our listeners understand, uh, regarding, you know, why that was so important to include responsible artificial intelligence in the title of, uh, of the center? Mm-hmm. You know, what, what would you say it means about responsible ai. Yeah, I think, uh, we all have the experience like use the chat GBT perplexity or all of those kind of tools, right? So we can find that the output that the AI generated to us not always correct. They definitely like make the life more easier and make the learning outcome like more beneficial for all of the users. But we also need to think about those like the, this negative part, like the misinformation part and always our, like the privacy concerning those parts. So those are very, uh, like the big concerns for our users and also for the young learners. Yeah, yeah. Yeah. So I'm, I think it's really smart, honestly. Mm-hmm. And I, I know I keep saying that, but this is a university, so you're already, everybody here is very smart, but I just, I think that it helps. People make that transition, uh, the practitioners in public schools and parents, right, to make the transition that this center is taken that so seriously. Mm-hmm. And so your role, uh, with respect to ethics and policy is, is critical to its future success. Uh, with that said, talk a little bit about, you know, um, the things you're thinking about with ethics and fairness and transparency as this pro project moves forward. Mm-hmm. Yeah. So we are thinking that, um, because we're trying to generate the AI tool for the young learners, so we're thinking that the young learners, they're not like passive users. They have their, you know, their positive and also the initiative to do the thing together. Just like what you have mentioned that they even maybe. Very soon, maybe recently, over the closing futures, uh, they can just create AI by themselves. Yes. To customize the AI tools by themselves. So what kind of the rules that we accept for gener just to interact with technology? What other, you know, the humanity and those social norms that we want them to use the while they interact with the tools, right. Yeah. So those are very important part, and except for that, the privacy and transparency and those part, we are thinking about it because we generated the tools for their early literacy learning. So when they reading things, we hope that the reading context that we generated are responsible for them. Just like a human. When we are talking about things we always, like, sometimes we have that buyers when we're talking about things. So the AI as well, because they have gained the data from provided by the human. So we really like you, uh, we can just set up the boundary and give the safeguard when we give those. Output to the children. Yeah. Yeah. Super, super important. We only have a few more minutes left, but I'm very curious about this work, your partnership with Stanford, uh, building, you know, ethical, responsible ai. Mm-hmm. Uh, how is this work going to inform the design? Of this air product. Yeah. That is the main tax that we do this policy reviews. And currently we're just aware still in the data analysis part, we already reviewed like 24 policies across different like platforms such as Oh, smart, yeah, yeah. National levels, uh, in, uh, including the White House and, uh, the us um, department of Education. And also we, uh, covered the policies from the international platform such as the us, uh, un, unesco, and UN and eu. And those policies really give us enrich understanding about what does the responsive AI means and what different kind of stakeholders can do to set up the, uh, safety guard, such as the parents and teachers and the, um, the leaders and what they need to do. So the, our next step is trying to generate a checklist for like, uh, even we were communicate with the acting, uh, lead by Dr. Lee, and also we were on. Also we were, uh, included voices from the parents and from different stakeholders. Yeah, the stakeholders, right? Yeah, yeah, yeah. Into our policy. And you already, I, I mean, I'm, I am a witness to that.'cause you had a session with myself and my principal. And, and are you also having sessions like that with other leaders, other educators, other parents? Yeah. Yeah. We are doing, uh, this is collaborative work. We also work with Kristen and the other amazing team, uh, the colleagues. And so we are, uh, interviewed, uh, like you. Excellent. The superintendent and the principals. And also we included the curriculum leaders and the teachers and also the tech leaders. Yeah. So. Across different kind of field. Yeah. To understand their authentic thoughts. Yeah. On that. Love it. There's a lot of work, but mm-hmm. That groundwork is gonna really set the stage for the design and the, uh, facilitation of this tool. That's fantastic. So Christine, uh, I met you a few years ago at the institute, uh, when it, when that got started. I was so impressed, right. With Jun and Rana and Venue and the whole team here. Uh, met so many wonderful people, met you and here we are now. Uh, it's, you know, you're running, uh, this own your own institute. You secured and you and your team secured$10 million for the next five years to keep working on something that really is, uh, going to elevate. Early literacy skills and literacy skills in general with children all across the United States. It must be so fulfilling. It must be so amazing. And I know there's a lot of stress and work right, involved with all of this, but what, what a wonderful, wonderful testament to you and your team and, and the region, right? To be a leader here at the University of Buffalo and be the leader for, for the United States with this particular topic is truly amazing. So as we wind things down, what else would you like to add, share for our listeners so that they better understand all this great work. Thank you for being so patient, um, and allows the whole team, um, to discuss all aspects about our salaries, uh, work. And I think you are absolutely right. Um. Salary as a part of, uh, playing a big role as a part of UBS lead leading role in the nation. Um, we have those two major national center. One is you mentioned the National AI Institute for Exceptional Education. Plus our so is really cement our leading role in the nation, in the education arena. Uh, in terms. AI innovation. Um, so I guess the final message I would like to send is, oh, before I get to my final message, I do wanna send a heartfelt thank you to teachers and families and the school leaders who have opened their door to allow us to do research, including your school district. Provide Yeah, absolutely. Um, uh, important support for our work. Um, so I guess the final message is, is if you are listening, if you are curious or you have ideas, uh, concerns, we want to hear from you. And we are also looking for more research site. I love it. More participants, so get in touch. Our website is early literacy ai.org. Yeah. So from there, um, you can sign up for our newsletter. Um, there's an email contact there. So we are looking for, for the opportunity, um, to work with you and, uh, um, work together to really support, inspire young children to read. Yeah. Wow. That is so well said. Uh, for our listeners, I hope you enjoyed this episode as much as I have. I think it is absolutely amazing to live in Western New York and to see this incredible work. Uh. You know, the origin of it being here, right in our backyard. Uh, with that said, I would say to our listeners, uh, feel free, uh, to email or contact, uh, the Center for Early Literacy and Responsible ai if you are interested in participating in future research or, uh, eventually, uh, helping with the prototype of air. And, uh, and I know we'll continue to share their good news and, uh, elevate and celebrate all this great work, uh, around Western New York, uh, with your support. Uh, so once again, if you are a follower or subscriber to achieving joy in mastery in public schools, don't forget, you can always listen on Apple, Spotify, Odyssey, Amazon, and so many more of those favorite podcast platforms. Until next time, thank you so much for tuning in. Thank you for joining us today as we explore the concept of achieving joy and mastery in public schools. We will be interviewing school districts in Erie and Niagara Counties on a regular basis as we look to shine a spotlight on all of the amazing programs and practices essential for achieving joy in mastery with our students, faculty, staff, and community. This podcast is sponsored by the Erie Niagara School Superintendents Association, and we hope you consider subscribing.