Episode 30: A Beautiful Prompt: How Utah is Trailblazing AI in K-12 Education
Ryan Lufkin (00:00.984)
Hello and welcome to another episode of Educast 3000. I'm your co-host Ryan Lufkin.
Jody Sailor (00:06.518)
and I'm your guest co-host Jody Sailor. On today's episode, we are excited to have Matt Winters, who is the AI specialist for the Utah State Board of Education. Matt, welcome to the show.
Matthew Winters (00:18.779)
So excited to be here, thanks for having me.
Jody Sailor (00:21.416)
Absolutely, thank you. Before we jump into today's topic, Matt, would you mind sharing just a little bit about yourself so our audience can get to know you?
Matthew Winters (00:29.755)
Yeah, absolutely. So my background is I was a higher ed adjunct professor for a long time, teaching English language and literature. And then I switched over to working in K-12 about a decade ago, taught in the land of bad decisions, better known as middle schools. Shout out to Jessica Pack for giving me that phrase. Started building makerspaces, doing educational technology, getting really into artificial intelligence and 3D printing and all sorts of stuff.
Ryan Lufkin (00:46.702)
Ha
Matthew Winters (00:58.097)
And then started working at Utah Education Network just in 2021 and did training all over the state on all sorts of topics, including AI, Google, Canvas, all sorts of things. And then moved into this role about just over a year ago, which was fantastic.
Ryan Lufkin (01:16.118)
Awesome, good timing. There's a lot going on in that space. Yeah.
Matthew Winters (01:19.729)
Absolutely.
Jody Sailor (01:20.054)
Absolutely. And Matt, you also served on the Utah Coalition of Educational Technology.
Matthew Winters (01:25.017)
Yep, I was the first back-to-term president of UCET and we moved the conference to the Salt Palace during my presidency. I was going to say reign. I think that's the wrong phrase. I think some people may have thought about it. But yeah, no, during my presidency, we kind of got bigger, which was fantastic. Brought in some really great speakers and had a really good time doing that.
Ryan Lufkin (01:38.633)
You're right.
Ryan Lufkin (01:49.167)
Awesome. And for our listeners outside of Utah, the Salt Palace is the big convention center downtown here in Salt Lake City. There's no bigger space in the state to really have events. that's great. Now, one of the things we always like to ask our guests before we dive into the meat of the conversation really is we ask for your favorite teaching moment or one of your favorites. We all have kind of a lot. Every time I think we ask this question, I think of a new one. But that can be something where you were the teacher or it can be something where you were the learner.
So what do you think? Give us one of those.
Matthew Winters (02:22.255)
Yeah. so I've told this story many times when I've been doing keynotes and things like that, cause it's, it's a pivotal moment for me. it's a moment where I was being taught. when I was 15, 16 years old, I got really bored in school and I had teachers that, a teacher that I didn't really love. and she made decisions around what was happening in my classroom or the classroom I was in, not very fun. And so I started working on
paperwork to leave school. And so I wasn't going to graduate high school. I was just going to go off and work and figure things out. And I kept working, going to school. Like I was one of those kids that like, you know, I didn't want to drop out, but I didn't want to be there. And I got placed in an English class the next semester. And I was like, okay, I'll give it one more try. And this teacher ended up, she's one those teachers that puts, put up like on the board, like a prompt every day and like, go write.
Ryan Lufkin (03:05.805)
Yeah.
Matthew Winters (03:18.353)
Whatever you want to say, just write and I'm looking for a page. And so I kind of took it to heart, started writing. She introduced me to some really great authors like James Joyce and John Krocower, J.D. Salinger. And about two weeks into the course, she pulled me aside and she's like, I'm moving you dead center front row. And I was like, well, because I'm, those of you that have never seen me, I'm six foot eight and a very large man.
Ryan Lufkin (03:32.014)
Yeah.
Ryan Lufkin (03:45.293)
Yeah.
Matthew Winters (03:45.423)
So being in the front row is weird for me because I block everyone. And she, was like, I'm in trouble. And I said that to her and she was like, no, no, no, no. I just need you in the front row because everyone needs to hear what you're saying. And it changed my perception of what school could be. Made me want to be a teacher, made me want to be kind of into English language arts, all those sorts of things. And it just kind of shifted my mentality. And so I always look for...
as a teacher those opportunities where I can really highlight the kids that need an extra set of guidance but then as a now as a know state employee or when I was doing coaching at UEN looking for people who are doing amazing work and being the microphone for them handing them the mic so they can do really cool things.
Ryan Lufkin (04:30.626)
That's awesome. I love that story. And at 6'4", I'm very much used to being kind of pushed to the back in you know, in photos, I always go to the back in the, you know, in the classroom, tend to go to the back. don't want to, I don't want to block people. So I understand that. I love that, you're, you know, your teacher recognized that that was something that was actually keeping you from being heard or being near, you know, fullest potential.
Jody Sailor (04:32.566)
Great.
Ryan Lufkin (04:52.236)
So it's amazing. Well, so as, as Jody and I started talking about, you having you on the podcast, it's amazing because we both have students, that go to school in Utah, right? I've, have a, now a sophomore at the university of Utah, but you know, went to all grades here in Utah. And I have a son who's just left the, what did you call junior high? land of bad decisions. Yes. He's just left the of bad decisions and is moving off to high school. So, you know, and Jody.
Matthew Winters (05:13.423)
Land of bad decisions. Thanks.
Jody Sailor (05:13.43)
you
Jody Sailor (05:19.902)
And I have a daughter that is in that same exact boat and I'm really glad to hear that it's leaving that land. And a rising freshman in college, so very similar ages.
Ryan Lufkin (05:21.836)
Yep.
Ryan Lufkin (05:29.154)
Yeah, yeah. So it's interesting for us and I see the difference in how, you know, specifically AI is being addressed between those two environments. And so I'm excited to hear about, you know, what the Utah State Board of Education is doing in this initiative in general. So kind of what sparked this, obviously AI has been a hot topic for a couple of years now, but what sparked this from a kind of a state level with you all?
Matthew Winters (05:55.749)
Yeah, there's a couple things. Utah's had some really visionary leadership in education for the last couple decades, thanks to Superintendent Dixon, some great board members we have here as well, and then just the wonderful employees we have here at USBE, but then also our sister organization, UEN, Utah Education Network, has some really great leadership over there as well. so when Chad GBT hit in November of 2022,
Ryan Lufkin (06:23.81)
November 30th is my birthday. like to remember and remind everybody of that. That's yeah, exactly.
Matthew Winters (06:25.327)
Yeah.
You got the world wildest birthday gift that year. You know, it hit and then, you know, that spring of 2023, everyone was trying to figure it out. I mean, I remember going to USET and that was the first year around the conference and we didn't have any AI sessions. So I had to like ask people like come and present, like figure this out. We brought in some companies that they were very small at the time, but around AI now they're absolutely massive. And so...
There was this visionary leadership that was behind everything that was kind of looking at this. Over on the UEN side where I was at at that point, we started building courses. We started looking at how we can support our technology leaders in a variety of capacities, giving them training, those sorts of things. 2023 alone, I did about 5,600 hours working with teachers across the state. Most of that artificial intelligence work, trying to get them in the know about what is Gen. AI and how to work those things through.
And then spring of 2024 hit and we had some really interesting things hit that were very unique in the national perspective. So the first thing was we had a collaborative group that's run by my friend Emma Moss over in Canyon School District. And she ran this group, it's still runs this group. It's a large scale initiative with a lot of our LEAs across the state. So our districts and charter schools and basically the technology directors in those
those schools said, we need cheap, safe, effective AI tools. We just need these for the next couple, next upcoming school year and we need them quickly. So we worked on an RFP, a request for proposals for AI tools. We ended up being the first state to have consortium pricing around those. But also on top of that, the state legislature buys tools like Canvas and Adobe and Nearpod that all have some sort of AI component as well. And so Utah was, is setting up on
Matthew Winters (08:25.145)
that digital kind of equity, closing the digital divide question very, very quickly. And I am very, very proud of that. And that was over on the UEN side. And that was a little bit more public as buying contracts. Over here on the USP side, unbeknownst to me at the time, Superintendent Dixon, Deputy Superintendent Norman, who sadly has retired since then. Sad for us, not sad for her. She's very happy.
Ryan Lufkin (08:48.967)
She's enjoying her life.
Matthew Winters (08:51.131)
but they were working on with an inter departmental team, an AI framework. So a framework, our framework around AI is meant to be a living document. We're working on draft two right now, but they were working on this in the behind the scenes with a large group of people. And in that process, they went, we need somebody to kind of watch over this document and watch over this conversation and see where it's going because it's getting too big, too fast.
And they opened up a call for the first AI education specialist in the entire country. I happened to apply for that and they pulled that in. So Utah has been really leading the way in a variety of capacities, both digital equity side, but then also providing frameworks and opportunities. Right now, one of the biggest projects we're working on is a professional development program that will...
Right now, Zoray trained 3,000 teachers across the state of Utah with artificial intelligence. We're working on possibly between 5,000 to 7,000 teachers by the end of the summer. So there's a lot of really unique things that are happening here in Utah, and it all comes back to we have a great community and visionary leadership.
Ryan Lufkin (10:02.19)
That's amazing.
Jody Sailor (10:02.548)
Matt, it's really incredible to hear the work that you've done. And as a Utah educator and parent myself, it's just so impressive to watch the way that the entire USP really has been coming together across the different departments there, but also with all of the leaders across the different LEAs, like you mentioned. I'm curious, you're talking about the professional development with the teachers. And we know that is where we need to make sure that teachers are comfortable.
and confident in the way that they can roll these new tools out to their students and to use in their own classrooms. And so I'm wondering if you could just share a bit more about really how you're working with the different schools and districts to help them to understand how to responsibly roll these tools out and thinking about, you know, the ethical and the secure ways of using these new tools that they have now available.
Matthew Winters (10:56.133)
Yeah, absolutely. So a couple of things that are pretty important in this conversation is first, PD should be ongoing and recursive. We can't just do one and done, especially around AI. It's such a fast moving conversation that if you go, we're done. We did an AI training. You're going to be on the cold after about six weeks. So there's ongoing training. And I think we need.
Ryan Lufkin (11:07.202)
Mm-hmm.
Ryan Lufkin (11:14.574)
Mission accomplished, yeah.
Matthew Winters (11:24.643)
And we've shifted this in Utah, but it's not everywhere. I always love what Ken Robinson says about PD, which is that it's an investment. And I think that we need to invest in teachers and our, not just our teachers, but our paraprofessionals, our administrators. Everyone needs to have invested professional development around not just AI, but a variety of things every single month, every single year. And it's important to do that because we got to keep up on what's new.
Around what we've been doing with artificial intelligence, this program that we've been working on thanks to a grant from Intermountain Healthcare basically creates a tiered system. So the first layer is a one hour PDE. We come out to anywhere in the state of Utah, any charter school, anything publicly funded, a publicly funded district as well. And we do a one hour training that basically looks at a couple of key components. What is AI? Why is it important in education?
which basically comes down to if you don't look at AI right now, you're going to lose jobs and you're going to lose, there's some issues with how AI is presenting itself in the marketplace and making sure that people are thinking through in education what their students are going to be graduating with and whether or not that's actually the acceptable skills that they need. And then we talk about data privacy and security because that's a
core central tenant here. It's a different technology than we've seen previously. It has some volatility with how we think about our data and how it's used and absorbed. And then we end up with looking at tools and we look at three big tool types, constitutional chatbots, rag databases, and multimodal AIs. We look at tool types because every district, every charter school is a little bit different on what they have selected. Local control is a great thing and a real thing.
And so we figure out what they need, but we focus on tool types so that they understand if they switch schools or something like that, they can work that through. This PD, again, we hit the 3,000 teacher mark last week since April. And we're hoping to do about 6,000 to 7,000 by the end of the summer, which is roughly about a quarter to a third of our teacher population in the state of Utah. The goal being is that we continue this.
Jody Sailor (13:38.667)
in month.
Ryan Lufkin (13:47.747)
Yeah.
Matthew Winters (13:50.873)
working that through. Now that first layer, then we move on to a Canvas course optionally, and at the end of it, they write a lesson plan in their content area and grade level that either uses artificial intelligence with students or teaches about artificial intelligence with students. when they do that, if they are accepted, we can pay them $200 for donors to choose. And so it's this paid model that thanks to the grant that is allowing us to build a database of lesson plans where teachers are using AI.
Ryan Lufkin (14:07.968)
Awesome.
Matthew Winters (14:18.469)
with our content area and grade level. And so far, we don't think there's a database like that around globally. And so we're building that out not just for here in Utah, but across the globe so other people can benefit from it as well.
Ryan Lufkin (14:23.906)
Yeah.
Jody Sailor (14:29.238)
incredible.
Ryan Lufkin (14:29.422)
So those lesson plans can be shared with other educators and they can actually model kind of what good looks like, right? So one of the things I think is so interesting is, those are pretty, I mean, that's pretty effective impact, you know, when you talk about somewhere between a quarter and a third of all educators in the state, how do we continue to drive that? And so we've given that consistent experience in, you know, every school across the state of Utah.
Matthew Winters (14:34.651)
Mm-hmm.
Matthew Winters (14:56.143)
Well, that's where recursive modeling comes in. looking for, we're looking for layer two funding to, to maybe do this again next summer and things like that. But then also working with our LEAs to figure out what they need and then figure out the avenues for that to happen. So things like, let's say a district wants to do more AI training. Maybe I'm not the provider on that, but we help them find that provider or we encourage them to figure out what
the best practices are for themselves. And so we're looking for multiple avenues. There's things like our, we have an upcoming document that's coming out that's called the Portrait of an AI-Infused Student, Portrait of an AI-Infused Learner that's connected to our Portrait of a Graduate. Again, first of its kind document. We were trying to have it out, but there were some complications with that over the last couple weeks. So we're working on it for early July. But the goal is to tie it to our Portrait of a Graduate here in the state of Utah.
Ryan Lufkin (15:36.8)
Ooh.
Ryan Lufkin (15:44.462)
Yeah.
Matthew Winters (15:53.041)
outline what are the skills, knowledge, and dispositions that a student needs to kind of work through being a part of an AI infused society and doing those things appropriately and safely. then also on the teacher side, here are the skills that they need to thrive as a teacher in this kind of environment as well. And so we're building all of these out as documents that can help guide conversations in LEAs across the state or even in individual classrooms where they can look at this and go.
this is stuff that I can embed into my standards. can go into my course content that isn't a heavy extra lift. It's just something that I can add that gets the kids to the skill sets they need.
Ryan Lufkin (16:25.41)
Yeah.
Ryan Lufkin (16:33.644)
Yeah, that's one thing I love about education just globally is that it's so collaborative. It's so, you know, no one should feel like they're starting from scratch. Someone has started laying that foundation. So I love the fact that you're really focused on being able to share whether it's the framework and we'll link to the framework in our show notes as well. But I love that you're making this kind of publicly available so everyone can benefit from it. Amazing.
Matthew Winters (16:57.413)
Well,
Jody Sailor (16:57.628)
I love that it's extending that sphere of what your professional learning community really is, right? These communities of practice and it's helping to really help those teachers, like I mentioned, feel more confident. When you can see what I, you know, I'm maybe a teacher that's feeling a little bit nervous about using AI, but I can see what this portrait of a student and a teacher looks like. You're just making it so consumable and actionable for educators and students across.
Utah, but across really the nation. And it's so incredible.
Ryan Lufkin (17:28.11)
Well, yeah, and the challenge is even it, I see this in higher ed, but it's more, it's more, I think widespread in K-12, but this idea that we can't just move beyond the idea that AI is a cheating tool, right? A lot of educators are very hung up on the fact that it is, it can be used for cheating and because it can be used for cheating, they don't get very,
creative in its use cases or they don't start thinking about all available use cases. And so I love that you're modeling that behavior because I think they need to see what it's capable of so they can understand that it's more than just a cheating tool.
Matthew Winters (18:08.049)
We know Emma Moss from Canyons and Jonathan Stewart over there. They've done some amazing work and they're leading out, I'd say some national conversations around this, they've come up with a really interesting chart that you guys can link to if you'd like to that's around AI adoption. It's like a 10 layered thing. But one of the earliest conversations is getting stuck on plagiarism. I always think about it like this, is that when we work with teachers,
Ryan Lufkin (18:25.731)
Yeah.
Matthew Winters (18:37.169)
there's a continuum here around AI. And the first layer is productivity. They start to see AI, once they get past the fear-based conversations around plagiarism, around, and let's face it, with plagiarism, those conversations were happening long before AI. I remember being a teacher in 2010, and I had students buy papers off paper mills. It exists.
Ryan Lufkin (18:55.084)
Exactly. Yeah.
Ryan Lufkin (19:01.772)
Yep. There's a reason Chegg's stock has dropped aggressively since AI came out.
Jody Sailor (19:06.014)
That's right.
Matthew Winters (19:08.177)
And so we have to remember that this is an ongoing conversation, but if we are able to harness the conversations around AI in interesting ways in our classrooms, then that's actually going to cut down the amount of plagiarism. It's just a matter of understanding the underlying technology and really working it through. And so with teachers, one of the things that I point to a lot is that when we start looking at AI and you get past those fair based conversations, you move into productivity and productivity is great. Like writing a lesson pattern fly out faster, getting through the administrative stuff, like making sure that
your workload is cut down.
Ryan Lufkin (19:38.99)
Administrivia. That's a good one. I've been borrowing that. Administrivia.
Matthew Winters (19:43.537)
I got a, that one goes to Carl Hooker, one of my buddies. He taught me that one. And so when we, when we think about that, that productivity conversation, if we get couched in that though, if that's the only way that people see AI, generative AI, then that's a problem because then it only becomes a tool that helps me basically speed up my workload. AI though, when it really gets down to it is a tool that really thrives on creativity.
Ryan Lufkin (19:45.838)
Yeah.
Matthew Winters (20:12.333)
It can ignite people. Curiosity is a big part of how it functions. And so if we're able to actually dig into how AI functions as a creativity enabler, you can really do some interesting things. so when we push on, one of my favorite conversations I have with teachers often is I'll hear a teacher say, can AI do, and then they have this like two to three minute, like really hyper personal.
Ryan Lufkin (20:40.174)
hyper-specific.
Matthew Winters (20:41.219)
Yeah, example of like, can this do this in my classroom? And I'm, I always say, I don't know. Ask the AI. It will tell you whether or not it can do it. And if it does do it, then it's just going to do it. And you're going to have that thing done and then you can return tool it if you want to. If it can't, it's just going to, you know, give it a couple of months. It might be able to do it then. And so I always encourage teachers to think that through. But then once we get through creativity, you have to start thinking about inclusivity.
Jody Sailor (20:41.418)
Yeah, yeah.
Ryan Lufkin (20:48.749)
Yeah.
Matthew Winters (21:09.345)
and how AI can act as a tool to meet the needs of students who are traditionally underserved and helping them to work with it. I think about my role in the classroom when I was in middle school, high school, and I was always the kid who was smart enough to get work done quick, and I would always have to tutor someone else. And that's not an effective use of my time. It's just, the literature shows us that. If I had AI that would extend my learning and like encourage me to do really interesting things,
It may have been a different scenario for me. And so we got to think about our gifted and talented, our multilingual learners, our students with disabilities. How can AI be leveraged to help support those communities in ways that currently we're not able to? So I really like walking that through with teachers and showing them that AI is not just this amazing tool that can write my lesson plan. It's actually can do some really heavy lifting with us to
build our classrooms to where we want them to be.
Ryan Lufkin (22:11.256)
Yeah, and that idea that that doesn't undermine the value of an educator or in any way to track it just, it just powers them in new and exciting ways that makes the learning more engaging, makes it more personalized, right? Like that is, I love that viewpoint.
Jody Sailor (22:25.044)
Yeah, I think of all the things that I wanted to do in my classroom, right? But I had limited time and limited resources that if I were to go back into a classroom today, I feel like it would be a totally different experience for myself as an educator as well as my students because I could do some of that like you were talking about, Matt, extension or providing the different levels of information and different things that I wanted to provide to my students. My husband once asked me.
are you going to be working all night? Should we come home or are we going to be staring at the top of your head and maybe stop and visit grandma on the way home? And I thought, what are you talking about? But it was because I was rewriting texts to the different levels for my students. And the fact that we have this incredible tool that can allow us to do that and expand the horizons for all the different students and the different levels in our classrooms is just absolutely incredible. So I'm wondering, Matt, with that in mind, is there a specific success story or case study of an educator that you have seen or maybe it's
You you shared a lot about what Canyons is already doing and they're doing some really incredible work. It's really, really impressive what they're doing. But is there one specific success story where a teacher has moved beyond that productivity or has moved to one of these higher levels that you can share with those listening?
Matthew Winters (23:38.417)
Yeah, absolutely. I want to make sure that I say this though, too, is this technology is two and a half years old. When we talk about generative AI, November 2022 is not that far away. And so we have to remember that the research cycles to validate generative AI and education have not fully been completed yet. And in fact, they're just barely starting to some degree. And so when we talk about
Jody Sailor (24:01.43)
That's right.
Matthew Winters (24:06.437)
best practices around generative AI in the classroom, or are we seeing learning shifts or learning goals? There's some of that's starting to happen, and somebody that I know from Stanford calls it best promising practices. I think that's something to really think about and use in the vernacular around those.
Ryan Lufkin (24:19.286)
Yeah, that's interesting enough.
Jody Sailor (24:19.734)
That's right.
Matthew Winters (24:23.665)
In terms of my own stories though, we did a wonderful series of trainings that were funded by the STEM Action Center here in Utah. There were two day trainings we brought out. We did three of them, two on the Wasatch Front, one in Southern Utah.
And we brought out teachers from all over, groups of like 30 to 40 teachers every time. And it was intensive AI training. So two days we really dug into what it is, how it works, and then gave them time to be coached through developing a lesson plan and process that work. In that whole system, we got some amazing lesson plans, things that...
They then shared at UCET and some of them are sharing at the upcoming Utah Rural Schools Association Conference. perhaps there's a couple of favorites in there. One, there was a teacher who I sat with and I wish I would have taken pictures of her presentation, but I was just so thrilled with her project. She's a political science teacher and she had constitutionalized an AI to work with her students on their own political bias.
Ryan Lufkin (25:29.4)
interesting.
Matthew Winters (25:29.937)
And like track that, like where are you seeing shifts in your own bias systems? And she was like, I've done this lesson for a decade and change. Every time I do it, I don't know where to take it the next day. I don't know how to track their bias. don't know. I don't have real actionable data. And she was like, with this, not only was I able to see it in real time, but I knew exactly where I could take the lesson plan the next day. And so we talk a lot about, we pay a lot of service in education to data.
being data-driven instructors. AI really opens that door for us to do that data-driven instruction that is important and engaging in a way that is real-time instead of it being two days or two weeks after the initial assignment. She was just so thrilled. And she was like, I could actually see how my students were shifting their perspectives in real time through the AI. she was just blown away by it. It was amazing.
But perhaps my favorite was in that same group we had an educator from a charter school. He's a special education teacher. He has small classrooms between four and six people, students per class period, and there are variety of disabilities, from nonverbal to students on the autism spectrum to variety. We were having a one-on-one post the seminar and doing some coaching, and he was like, I really wish I could create an AI.
that would meet all of my students' disabilities and that they could have a conversation where it's keyed into their individual needs in the classroom as well. like, if I have a student with autism, I can key the AI into working with that. But then also, there's a social studies student and they're struggling on that. Key that in as well. And I was like, you can. And he went, well, I can only make one of these per thing. And I was like, no, you can make as many chatbots as you want.
Ryan Lufkin (27:23.086)
Yeah.
Matthew Winters (27:25.783)
And so he started making a chat bot for every single one of the students keyed into their individual disabilities and their academic goals. And he was like, it was incredible to see that shift. Like the nonverbal students, particularly, he was like, I can actually see them learning and talking in the real time now. And they may not talk to me verbally, but I can actually talk to them in other ways. And I've never been able to do that or see that. And the best part about this particular teacher is he told me all of this and he was like, do you think that's any good?
Ryan Lufkin (27:30.595)
Wow.
Matthew Winters (27:55.619)
And I looked at them and I was like, yes buddy, you're doing amazing work here. And that's the kind of stuff that we love to see is teachers really keying into the academic needs, if you're working with a student with a disability, their personal needs, those sorts of things, and digging in ways that allow you to help the students in ways that have never been able to be done before.
Ryan Lufkin (27:56.504)
Cheers!
Jody Sailor (28:00.256)
Wow.
Ryan Lufkin (28:20.654)
Yeah, that's exactly what we talking about earlier, that curiosity, that exploration of what the tools are really capable of. And you've got to acknowledge that it's more than a writing tool, it's more than a chat bot. You can actually start doing some really amazing things. RMIT down in Australia did a really interesting...
their chat bot's called Lev, but they created individual personas. So if you were doing research, here's your persona. If you're exploring campus, here's your persona. If you're in need of mental health interaction, here's your persona, right? So that ability, so, you know, it's pretty good, pretty good, you know, Utah, Utah K-12 institution on the same page as RMIT is, you know, not a terrible, not a terrible thing to see, it's amazing. How do you, how do you?
Matthew Winters (28:58.897)
you
Ryan Lufkin (29:04.566)
You know, because these tools are only as powerful as the data they're leveraging. How do we balance that access to data with student data privacy, especially in the K-12 space? I know that that has been a bit of a hang up for some institutions on moving forward.
Matthew Winters (29:19.877)
Yeah, Utah's, so I'm actually on the student data privacy team at USP. So this is part of my day in, day out. I'm not a student data privacy officer. That's John Lyman and Nicole Sanchez. But I do a lot of work with them.
Utah is one of the few states across the country that has extra student data privacy concerns on top of FERPA and COPPA. One of the big things that we do here in the state of Utah that's pretty unique is we have a licensing system for data privacy agreements between companies and districts and charter schools. So let's say that a big district wants to work with, or even a charter school wants to work with a specific company.
company can sign the DPA with that district or that church school. But then as soon as that gets done.
and signed any other district or charter school across the state can license that same agreement and start using that product. So it saves some time on every single person's, every single company's kind of perspectives and their workflows, but it makes it really easy for everyone to see what's available and get it done really quickly. And so we're, one of the stances that I took before I was even on the student data privacy team was if I'm going to show off a tool or talk about a tool, it needs to have data privacy agreements.
Ryan Lufkin (30:10.548)
interesting. Yeah.
Matthew Winters (30:35.441)
here in the state of Utah. And I think that's incredibly important for any system, even outside of the state of Utah, is to really think through, in the age of AI, is there a data privacy agreement there? Do the teachers understand what that means? Are they... I tell this story a lot to teachers. I had a wonderful day in my class back in about 2018. I had come across a tool called Flipgrid, and I was like, this is awesome. I want to use it.
And I created a lesson plan where kids came in the classroom, scanned a QR code. It would give them instructions to write a poem and then go slam it down the hallway. And so they'd have to record themselves like doing the poem. And about 30 class periods into that, my principal came in and he said, it's time for your eval. Surprise. And I went, there's the lesson plan on the board. And I was like, do it with the students. And he was like, that's weird.
So he did the lesson plan with the students. slammed a poem. It was awesome. And he came to me afterwards and he was like, if that's what you're doing every day in your class, you've got a job forever. This is the kind of stuff you want people to be doing. It's authentic. And OK. I found out a few weeks later that there was an endowed privacy agreement in place with that company. Now, not a knock on Flipkart or anything like that, or Microsoft at this point. But that just wasn't in place. And I didn't understand what that meant. I thought my district was trying to quash my creativity.
Jody Sailor (31:43.414)
Thank
Ryan Lufkin (31:43.662)
You
Ryan Lufkin (31:52.738)
Bye.
Matthew Winters (32:03.312)
You can't use that tool like blah blah blah. Now I understand how important that is and I would encourage a lot of teachers who are going, well I just want to use this tool. My district and my charter school has not allowed me to. You need to just, you need to listen to your technology directors and technology directors need to be transparent with their teacher faculties of why tools are not allowed or are especially around data privacy.
Ryan Lufkin (32:04.77)
Yeah.
Matthew Winters (32:30.641)
We know with large language models, if you're using a model that trains on your data, all of that data goes straight into the large language model. And my worst fear is that a teacher doesn't understand that, and they plug in student privacy data into that. And they create a whole mess where that data can come back at some point, maybe permutated into someone else's response in the large language model. And so all of those are parts of the components of why data privacy is so important.
Ryan Lufkin (32:52.855)
Yeah.
Matthew Winters (32:58.373)
I'm just really happy to work with the team here in Utah who are leading the way and making really great decisions around how our data is used and making sure that our students and teachers' data is saved.
Ryan Lufkin (33:12.014)
Yeah, so incredibly important. We've already, you know, it's unfortunate, but in a time of such rapid change, we've kind of seen the first round of some of these AI companies that got funding and they've burned through that funding. And, you know, we, I, I personally got a call from a company saying, Hey, we're, closing our doors. Do you want to buy our IP? Do you want to buy our, our, uh, you know, beta partner list, things like that. And you're like, I hope all of that was covered by, by, you know, protection contracts. hope they had the right paperwork in place. And so, yeah, you can't, can't even.
Can't underscore enough how important that is.
Well, one of the that Jody and I started talking about guests, you know, we were talking about with Melissa, guests on the podcast. One of the reasons that you were on the radar is because you in the seat of Utah won an award at ASU GSV, right? Tell us more about that.
Jody Sailor (33:45.622)
Thank you.
Jody Sailor (33:58.87)
Yeah.
Matthew Winters (34:00.143)
Yep, yeah, big thank you to Sarah Young, our former chief of staff here at USP. If know Sarah, she is a force of nature, the best way possible. She put in for an award, it's called the Education, I'm looking at it right now, sorry, Education Innovation Showplace First Place Winner for Impact. It's a global award, we were up against three or four, well, more than that, about a dozen.
Jody Sailor (34:08.372)
incredible.
Ryan Lufkin (34:21.132)
Awesome.
Matthew Winters (34:29.489)
projects across the globe and great projects. There was one from the Bronx that was like our finalist against us that honestly they're doing amazing work and it's for impact. And so we're hit, we have the possibility with this program we're doing, but also the programming I've been doing across the state of hitting every single student in the state of Utah, which that's a hard impact to beat. And so very proud of it. But it's a cumulative effort across.
the entire agency and across honestly the entire state. When we won the award and people have been asking me about it, they're like, how are you guys doing these things in Utah? Well, I couldn't be moving as quick as I am around AI if I wasn't for the work that Sarah Young and Rick Gaysford did a decade ago around digital teaching and learning, that the work that the agency has done around personalized competency-based learning, around...
you know, working with UEN, the contracting around certain products and things like that. Utah has a really beautifully crafted collaborative environment. And so being able to pull on that and use that to my advantage has been incredibly useful over the last year and change in this job. so awards like this, I'm so happy to receive them. but I'm very realistic. couldn't have done this without, it's not an award for me. It's an award for the community.
Jody Sailor (35:58.998)
So with that in mind, Matt, I mean, I've been able to have the privilege of watching a lot of this work from various roles that I've been in. And I can't help but think of things that you and I have talked about in the past with like backwards design and how we're thinking about what's the future state and how do we make sure that we're tracking towards the direction we want to go. So with that in mind, looking ahead, what's next? Where do you see everything evolving over the next three to five years? And what is the role that you and others can play in making that a reality?
Matthew Winters (36:28.817)
Not a light question there.
Ryan Lufkin (36:30.99)
Just pull out your crystal ball and...
Jody Sailor (36:31.134)
No. No more softballs here.
Matthew Winters (36:35.281)
Yeah, that's a hard one. AI again is like two and a half years old, generative AI. so we don't know, I don't think anybody could have predicted some of the stuff that's happened over the last two years. What I'm looking at, I use a framework from Stanford's design school, d.school, called the Futures Framework quite a bit. And basically the premise of the Futures Framework is two things. One, how do we be good?
ancestors to our future descendants, whether it's culturally or biologically. And so it's the long view, 10, it's not 10 years, it's 100 years, 200 years, a thousand years. And so like really thinking about the impact of decisions now down the line, the other side of it is not always being focused on the fire that's in front of you, but the smoke of the next fire that's coming three to five years.
And so I'm really thinking about those things quite frequently and trying to prepare both myself and the community. That's part of the reason why we wanted to build a portrait of an AI infused learner and teacher is so that we can edit, revise that over time, but then also have this guidance document that helps everyone to see these are the knowledge skills and dispositions around AI in the classroom. And so those kind of framework documents really help to guide.
the change-oriented work that's happening. And then the other side of this too, and this has been feedback slash, I think this is for anyone who's in a change-oriented system. John Cotter in his book, Leading Change, talks about how if you institutionalize change around a single person, it fails over time because you tie it to the person, not tie it to the organization or to the actual structural change. And so over the last year, I've been very cognizant of...
And I would say this for everyone who's in this kind of role. I've been very cognizant of let's pull other people and give them a microphone and give them a platform so that they can start being the next generations of people like me who are doing interesting and engaging work. Because if it's, again, if it's just tied to one individual, it'll fail over time. People leave, people, I'm not saying that I'm leaving, just, you know, people leave, people make changes.
Matthew Winters (38:58.319)
So you need to have a whole structure. so in terms of what's going on with AI, if you're in a change oriented mindset, find people that are around you in your community that can really help you navigate these complex changes. And I feel like I'm punting a little bit, Jodi, but really what it comes down to is we don't know three to five years what's going to happen. And so we just have to think about how we build structures in place now that will support us in the future, even if they're not.
maybe ever going to come to fruition.
Ryan Lufkin (39:30.424)
I love that. Yeah.
Jody Sailor (39:30.442)
Yeah, I absolutely love that, Matt. think the change management and being comfortable with change is such a critical piece in all of this, right? I think we've talked a couple of times about AI truly being such a different shift and trend than anything we've seen in our 20 plus careers, your education, right?
Matthew Winters (39:49.179)
and it's going.
Ryan Lufkin (39:49.347)
Well, and those resources like the portrait of a learner and things like that, that's so valuable for us as a technology vendor, to help keep us on track and make sure we're developing in the right direction to support schools now and tomorrow.
Jody Sailor (39:54.908)
valuable. Yes.
Matthew Winters (40:04.475)
Yeah, it's one of those things that you really have to...
apply very thoroughly and then think through. I always think about with AI and with a lot of things, we don't know where we're going. There's things we can predict. But anybody who says this is where it's going is a liar. Because the shifts that we've seen.
Ryan Lufkin (40:21.612)
Yeah. Or they have a real functioning crystal ball, which you can make a lot of money off of, right? Yeah.
Jody Sailor (40:26.582)
And you're right. We all wish.
Matthew Winters (40:26.673)
Yeah, which if you have one of those, I would look to buy one. Please let me know. I have cash available.
Ryan Lufkin (40:33.762)
Ha
Jody Sailor (40:34.166)
I'll pitch in.
Ryan Lufkin (40:36.878)
Honestly, good investment. No, Matt, this has been amazing. Thank you so much for joining us. We'll add some links in the show notes here for all of our listeners, but also is there a place that our listeners can go to kind of watch what you're doing, see the latest and greatest?
Matthew Winters (40:54.105)
Yeah, I'm easy to find on LinkedIn. I post about once a week, trying to highlight people who doing really cool work. This will probably come out after it, but I'll be at SD this year, SCA, SCD, which is always a fun conference. And then schools.utah.gov. There's lots of great stuff on there, and we have an AI page that I'm going to be adding some stuff to over the next one to two months.
Ryan Lufkin (41:18.016)
Awesome. Well, well thank you so much. That is it for this episode of the Educat 3000 podcast. Hopefully you'll join us next time. Thanks everybody.
Jody Sailor (41:19.637)
Awesome.
Jody Sailor (41:27.744)
Thanks so much.
