More Than Personalization - How to Build AI Experiences That Empower Students

Connecting UDL Guidelines to SEE Mindsets

AI and UDL Work Better Together

5 Ways AI and UDL Work Better Together

The weirdest people in the world?

Amanda Bickerstaff

Amanda is the Founder and CEO of AI for Education. A former high school science teacher and EdTech executive with over 20 years of experience in the education sector, she has a deep understanding of the challenges and opportunities that AI can offer. She is a frequent consultant, speaker, and writer on the topic of AI in education, leading workshops and professional learning across both K12 and Higher Ed. Amanda is committed to helping schools and teachers maximize their potential through the ethical and equitable adoption of AI.

Dr. Nneka McGee

Dr. McGee is an educator, researcher, and advocate dedicated to creating or promoting access and opportunities for students navigating a future driven by artificial intelligence. She has served as an educator for nearly two decades. She taught middle school mathematics and was recognized as a district finalist for teacher of the year. Nneka has also served as District Gifted and Talented Coordinator, Director of GT and Advanced Academics, Executive Director for Learning and Innovation, and as a Chief Academic Officer.

As part of her doctoral studies, Dr. McGee researched the experiences of K-12 teachers implementing or planning to implement artificial intelligence in classroom learning environments. She has shared her work as a keynote speaker at events such as the NYSCATE Annual Conference and the Technology Leadership Institute. She has also served as a panelist or presenter at events sponsored by Carnegie Mellon University School of Computer Science, Vanderbilt University, Teach for America, ISTE, Foundations, Inc., National Education Association, and the American Association of Publishers. She has contributed to publications released by the U.S. Department of Education, Digital Promise, American College of Education, and the International Literacy Association.

Most recently, Dr. McGee served as project lead and principal co-author of “AI + Learning Differences: Designing for a Future with No Boundaries,” published by Stanford Accelerator for Learning. Her latest work, “AI for Young Learners: Balancing Tech Use with Authentic, Off Screen Experiences,” was published as a Jump Start Guide by ISTE in January 2026. She also served as an AI practitioner advisory board member for the Engage AI Institute and was selected as an EdSAFE AI Alliance Fellow in its augural cohort.

‍ 00:00
Amanda Bickerstaff
Hi, everyone. We're super excited to have you here. As always, I am so lucky to have such an amazing group of colleagues to be able to have these really important conversations with. As always, we really want these webinar series to be the foundation for just really good thinking and provocation about what Genai can and should be for our young people. And so there's nothing more important, I think, than making sure that the Genai tools that we build are really meaningful and fit for purpose. And so today, our webinar is in partnership with cast, with these lovely ladies that you see on the screen with me with Lindsay and Michelle. And we're really going to be focusing on the idea of more than personalization, like, how can we really build AI experiences and empower students?

00:46
Amanda Bickerstaff
And if you're familiar with cast, UDL and Understanding is such an enormous, wonderful, thoughtful framework for student engagement and meaningful, empowered learning. And I think what was the as always, just as a reminder, like, this is a community practice, so you are a community practice. You can say hello in the chat. And so feel free to say hi where you're from. Like, if you have a good resource, please feel free to share that. Just remember, though, make sure that you set it to not host and panelists, but to everyone so that you can share all of your wonderful thinking with each other. Also, if you'd like to have a question specifically for myself, Michelle, or for Lindsay, put that into the Q and A because that way we can make sure that we see it.

01:31
Amanda Bickerstaff
But as you can see, we already got Spain and South Africa and Scott, a good friend out in Ohio. So we're all, we're going to be a beautiful and amazing group of international folks that care about this. And so I want to talk a little bit before introducing our wonderful panelists about how this came to be. And so Lindsay and I have known each other for about two years. We met at a kind of a think tank event back in Gosh in 2024. But we didn't really get a chance to kind of sit down together until as GSV in April. And we had a lunch together before the craziness started. And I was so struck by just how important the work that CAST is doing today is.

02:10
Amanda Bickerstaff
And I think that especially now with all of the pushback around ed tech and screen time in general, but also our knowledge, I think I will speak for Lindsey Michelle and I of, like, a little bit of, like, disappointment. There aren't better learning experiences being built today with generative AI or AI in general. And so I was really inspired to have this conversation. And so Lindsay and Michelle are going to introduce themselves and then we're going to do is a couple of things. Number one is we're going to have Michelle present about if you're not familiar with UDL and the opportunities of today, we're going to do that and then we're going to have a conversation about what we see as whether you're an ed tech developer, an educator, a researcher, kind of what you can think about doing.

02:51
Amanda Bickerstaff
But then the best part of today is that we've created a brand new co branded resource that really combines our mindsets of gen literacy with these key core components of udl. So I'm going to hand this over to Lindsay and Michelle, introduce themselves. I'm just really excited to have you all here today.

03:06
Lindsay Jones
Yeah, thank you so much, Amanda. We're excited to be here with you and it's always great and thought provoking to chat with you, especially in wild places like asu, gsb. And we are, we're thinking a lot at CAST about how we can make sure that we get the best learning experiences for students. So I want to just for those of you aren't who don't know what CAST is, we're an education nonprofit based in the US but we work around the world. We focus on teaching and learning and how to optimize that experience so that student academic achievement grows. We were really started focused on disability and my own background is a lot in disability rights, justice and education here in the United States. What the CAST UDL framework really highlights is learning design is within our control.

04:00
Lindsay Jones
We have a great understanding every day more about neuroscience. We are all variable learners. Every brain is as different as a fingerprint. We help to work with educators around the world, with tool designers, with others, to make sure that we can have classrooms that address that variability. We'll talk more about that today, but I'm joined by my great colleague, Michelle Soriano. So, Michelle, do you want to say a few words about your background?

04:24
Michelle Soriano
Yeah. Hey, everybody. I am Michelle Soriano and I am a professional learning and implementation specialist at CAST. And you know, I've been in education for over 25 years. I started out in general education, went into special education, and really my foot has been into inclusive practices my entire way through. So I live in Texas, y', all, so that is friendly language for you all. But in Texas, you know, I was able to really support educators and really identify, you know, how can we design better environments and curriculum to make sure that we truly are Designing for all. And that then pushed me into, you know, cast where I. It's more expanded, right? It's broadly. It's not only thinking about universal design for learning. We're thinking about accessibility.

05:19
Michelle Soriano
We are thinking about development of tools and resources and environments, not only using universal design for learning, but ensuring first and foremost that it is accessible for all. So it is a passion of mine. I've definitely been able to work alongside of our. Some of our amazing vendors, right. Who are. Who are building the products that we are using in the classrooms, in our agencies. And so that collaboration really has been just such a spearheading, strong point for all of us because we all know that we're better together when we can design intentionally.

05:54
Amanda Bickerstaff
And I think what's really great is that we're all. All educators. And I think that it's just different. I don't know, man. I hate to say this, but, you know, I've always had this really strong feeling, especially when I moved into ed tech, is that education technology feels like educators and technologists are, like, far apart and like some teams really get it. But it's really important, I think, for the organization that you're doing to not just be putting, you know, you don't just put out guidance into the world, but that you're really trying to shift practice not only at the individual teacher level, but. But also to the tools that are coming in.

06:27
Amanda Bickerstaff
Because I know, Lindsey, not to go too quickly, but were talking about how people want to put UDL on their package, and then when you press them and what that looks like, it actually just feels like it's much more a band aid or nice to have and is really thinking deeply about what the. What UDL means.

06:44
Lindsay Jones
Yeah, I absolutely agree with that. And I think, you know, Amanda, you guys have done in the world that's moving at this speed, right? I feel like the information you're putting out is so helpful, and it has a consistent thread, which is critical thinking, right? Not just teaching critical thinking skills, but being critical about what's coming at us. And I do know that we work with a lot of school districts, and they are overwhelmed right now, and they ask us kind of is this. They say this is aligned to udl. And honestly. So we have a product certification, as you know, we look at that. But what I see most often, and I think Michelle and our team too, drowning in options. I could do it in video. I could do it in this.

07:28
Lindsay Jones
And then presenting the representation part in many ways almost, that is not udl, right? That is just, in fact, that's exhausting. Anyone with so many options, it's too hard to make that decision. So I think that critical thinking that AI in education looks at, and that's why the document, I think, is really interesting, combining the mindsets with the framework can help us get to better development of tools. And I do think it's critical in this time just to say one last thing, which is you don't need tech tools to implement UDL in a learning experience. They can add flexibility when they're intentionally used.

08:13
Amanda Bickerstaff
Yeah, no, and I think that's a really good point. And sometimes I, we just get caught in these. Like, UDL means I give my kids options, right? Like, oh, okay, so they can do the same exit ticket in three different ways. And that is universal design of learning. And it actually is quite a bit more pedagogically complex. So, Michelle, but not to scare people, but I think that it's important to understand that a really good pedagogical practice is something that is really intentional, but also something that you can keep doing over again. Like, once you start getting into this habit, you're going to see it everywhere. Right. Once you start doing really good UDL design, everything starts to kind of make sense to you in a different way.

08:54
Amanda Bickerstaff
So, Michelle, you want to scare your screen and talk us through a bit of what you all are doing and so that everyone, if you're not familiar but. Or if you are familiar, you can get a little bit up to speed on UDL itself.

09:04
Michelle Soriano
So just testing, making sure everybody can see the screen. Okay. All right. So, yeah, this is really great and this is really exciting because I think we've learned, all of us collectively have learned a lot as we're collaborating together. And so when we think about insight to impact, right. What do we learn and how does that really impact our learners in that really inclusive design takeaways? And before we dig deeper, I just want to ground us in the mission that drives us work at cast.

09:35
Amanda Bickerstaff
Right.

09:36
Michelle Soriano
At its core, CAST is committed to transforming education so that all learners have meaningful access and opportunity. And that means designing environments that remove those barriers before they ever impact a learner, and ensuring that variability is not an afterthought, but a starting point. So through our work with education systems, organizations, agencies across the countries, we've seen a clear pattern. Truly inclusive and universal design doesn't happen by chance. It requires intention, collaboration, and a shared understanding of both accessibility and learner variability. The fact that everybody learns and processes differently, and we've had the Opportunity to partner with educators, leaders, product teams to rethink how environments are designed. And of course, that includes that development of AI from policies and systems tools and daily practices. And through that work, a few key takeaways continue to surface.

10:42
Michelle Soriano
Ideas that are critical if we want to move from good intentions to meaningful impact. And so as we move forward, we want to highlight just a few of those key takeaways that can help guide this shift towards designing environments that are truly inclusive and universally accessible. And so that takes us to our key takeaways, which are first, the fact that more options, what you just said, more option does not mean more engagement. It's all about purposefulness, intentionality, right? Intentional options do. And then we also think about learner agency. And we're going to be thinking just a little bit in talking more like what do we mean when we're saying learner agency here? But we know it's essential. We know that our learners need to be able to drive their own boat, if you will, pick their own pathway to learning.

11:42
Michelle Soriano
And so that is another thing, especially in the area of AI, that keeps popping up. And then of course, edtech, y', all, Ed Tech is not automatically accessible, right? And, and we'll talk about that misconception there. And then of course, accessibility and usability, it must be built in from the start, otherwise we're going to continuously be spinning our wheels and retrofitting. And so if were to look at that moving forward and we think about, you know, how more options does not mean UDL does not mean more engagement. Universal design for learning. It's an approach to improve and optimize teaching and learning for all by setting one clear, rigorous goals, anticipating barriers, and proactively designing to minimize those barriers. Right? We're thinking about the barriers in mind. We're predicting possible barriers so that we can build intentional options to help reduce those.

12:42
Michelle Soriano
And then we think about accessibility, right? Accessibility is when everyone, including those with and without disabilities, can perceive the same information. They can engage in the same interactions and enjoy the same opportunities in a timely manner. That means at the same time, my neighbor's getting their stuff, I'm getting my stuff, right? I shouldn't have to wait. And so accessibility ensures learners can access the content UDL helps us design with learner variability in mind. From the start. Engagement doesn't come from more choices, it comes from the right choices. Options should be meaningful, goal aligned and responsive to our learners preferences and needs. So instead of asking how many choices can I turn this one activity into what we should be doing is what are the possible barriers in this environment and how can I design options to help remove those barriers?

13:40
Michelle Soriano
So I know exactly which option is there.

13:42
Amanda Bickerstaff
So Michelle, just. I know this is. I'm going a little off book, but there's some good conversation in the chat. So when you think about that, we know that the idea of more options is like the safe place to do udl, right? So can you think of an example in which you've seen with like an AI tool where someone has picked not 100 options, but like the top three and how do they go about doing that?

14:05
Michelle Soriano
I think that is a very interesting question. Right, Because I will say that a lot of the AI tools, they have lots of options. There's lots of opportunities and options to embed. But this is where I think you keep the teacher in the loop, right? You keep that educator in the loop to where as an educator, I'm saying, okay, here are three options that you can use within this AI tool. Which one do you feel is going to help guide you to your goal? Right. So I like the educator because, you know, every state, every agency, every district is allowing AI to be used in a different manner. So depending on how that access is granted through that, I am the consistent, the educator that the human in the loop is that consistent. That is helping bridge that.

14:52
Michelle Soriano
So we might have to take it and narrow it down based off of our learners who are in that environment, based off their needs, and say, here are three options that you have access.

15:03
Amanda Bickerstaff
To and Lindsay, I saw you come off mute. Do you have some thoughts too?

15:05
Lindsay Jones
No, I just. I mean, I think that the. I want to emphasize that two things. First, the teacher is really important here. Right. I think we want to think about tools that lift some of the. Of the burden off the teacher. Right. That help them get better data faster, understand it. But they are setting rigorous learning goals. None of it works without that. Right. And then learner agency, as you see here, this is really just. We want to graduate students with great understanding of how they learn. So what Michelle said, a point she just made, was really critical. I haven't seen any tools without a teacher, and I'm not looking. You've seen many more than me. We need reflection built into those tools. Yes, purposeful reflection.

15:54
Lindsay Jones
Because the whole point is, when I have nothing and I need to learn something, how am I going to get there? So I just wanted to add this.

16:02
Amanda Bickerstaff
Yeah, it's really interesting. I mean, let's. Let's connect this a little bit to Gen AI and we can Move to the learner agency component. But one of the things I think about a lot is that everyone, not everyone, I'm going to broad brush it, but a lot of organizations have thought about, like recreating tutors, and they're Socratic tutors because somehow we decided that all kids want to be asked 100 questions by a bot that's not pedagogical. Right. And maybe there are some kids that really want that. Right. But where is it for the kids that need something different, that they need to ask the bot questions, a teachable agent. Where are the places in which, you know, there are opportunities for that reflection to happen? Like, you know, what do I, like, here's the bot for this thing.

16:45
Amanda Bickerstaff
Do I don't even want to use it. I'm going to say no. And I think that is an option too. But I think what ends up happening again is that, like, especially right now. Yeah. So we see like flashcards or quizzes, like take the same thing and put it into a short video versus not, I think a lot about like notebook alarm, that there's some really interesting things with notebook alarm and those kind of features. But they're not all created equal. And they're not all created equal based on what you give it for context to like the audio overview podcast versus the video with the flashcards versus, like, but it kind of feels like it is because it's all equally weighted. Right. It's just all there all the time. And no matter.

17:23
Amanda Bickerstaff
You'll never get a lot of tools, will never say, no, you don't have enough information or no, this is not the right approach. Or have you thought about the impact on students? And I think that's where I think that, like, it gets like really muddy, especially around that options and learner age, because it feels kind of like the opposite.

17:41
Lindsay Jones
Yeah. And one other just quick thing here, which I think we're really concerned about and I'm growing more concerned about, which when you mentioned the tutors and the chatbots, there are obviously social, emotional issues with. Right. But two things like large link, large language models are not designed to be teaching. Right. They are designed to give information, but they are not designed to help with productive struggle. Right. Everything we're seeing, and that is learner agency, it is pushing through, doing the hard work. The evidence we have the very little evidence that's out there. Right. But the evidence we have definitely indicates for some students, giving them a hint hurts their productive struggle. And we know they want the hint. And it's very hard to not want the hint. Right? Yes.

18:34
Amanda Bickerstaff
Yeah. Especially if you know it's there.

18:36
Lindsay Jones
Yeah.

18:37
Amanda Bickerstaff
It's like the teacher you had or the homework helper you had that would like, you knew if you were like, oh, please, I need help. And they'd be like, here's the answer. Like gaming that system. We get very good as young people.

18:48
Lindsay Jones
And I, you know, what really worries me are two things there. And that is what we say at CAST all the time is education has a tendency to blame the student. Right. The design is what we control. But we often go, oh, the student's lazy. And I see this a lot in disability.

19:04
Michelle Soriano
Right.

19:04
Lindsay Jones
With non apparent disabilities. Oh, the student just doesn't want to work hard enough. The student just doesn't. At cast, we just take that out of the equation. We're not going to blame the student. We're going to change the design and figure out how we can make this work. Right. Because we control the design. And so I get very concerned because I feel like there are some incredible things for assistive technologies. Right. Technologies that are designed for people with disabilities. There are some amazing advances for AI in that arena. What I do not want is for there, and I don't think anyone wants, is to just start feeding all the answers to students that are harder to teach because they're harder to teach. That's a real risk. I don't know how we address it.

19:54
Lindsay Jones
We're going to work hard at CAST to try to figure that out.

19:56
Amanda Bickerstaff
Oh, we can wait. Can I share something? This is. So I was with a wonderful, you know, in southwest Ohio on Monday and Ohio has machine learning graded tests and they're teaching the kids what words are picked up or not. Like what are. It's almost like an ats, like an applicant tracking system where, you know, if like it's not formatted correctly or it doesn't have the right words or some of the words that will be flagged. So they're learning to game the system. And I understand why. Right. We want your kids, like especially it's become so disconnected from like actual assessment of learning. But I mean, that is real, right. Both from the sense of feeding questions or like just starting to gain the system. If generally I starts to grade knowing that certain words look really like, you know, are. Look really smart.

20:41
Amanda Bickerstaff
Right. Or are expected. But I know we're a little off track and Michelle's so patient. But I mean, I think Michelle, as you. As we think about like the learning agency component, I think sometimes we kind of get into this mindset that like, if kids have choice at all, then it's agency. But you're actually framing. You frame this way different than that, right? That it's like authentic decision making, that they feel comfortable advocating for themselves, that there's a safe space for risk right. Taking. Like, if things not working, like, I don't know how I learned yet. We're going to try things out. So, like, how do you think about that with AI systems specifically?

21:15
Michelle Soriano
Yeah, that is such an interesting question. And the reason I say that is because. Okay, story time, right. I'm a mom of four. One of my own kiddos, you know, went through the special education system and I started to get really worried, right. By the time my kiddo was making it the highest goal, I was like, holy moly. What's my kid gonna do when he doesn't have his amazing special education teacher there to tell him, you know, pick up this or do this or write this down or tab this in. That's learner agency. If they don't know how to learn or what they need to learn, I was worried, right? How am I gonna make sure my child is a productive member of. We had to start allowing him to fail forward, right? Just like we. That's how we learn. We fail forward.

21:59
Michelle Soriano
That productive struggle, learner agency is knowing by trial and error sometimes what it works for you in this context and what does not, what might support you better. You know, I make a to do list every single day because I know that is what I need. I learned that, right? That's something that I automatically do so in AI. So I've had a really great opportunity to work with some organizations in thinking through how AI is being developed. And one of the piece that were looking at, and several of them is that learner agency piece. How are we making sure that when we are developing AI, that there's embedded intentional questions that they're making the students reflect on? What is your goal in asking this? You know, what is the support that you might need?

22:45
Amanda Bickerstaff
Right.

22:46
Michelle Soriano
And what I've learned because, I mean, I don't know a lot about AI. I'm lear as we go through it as well. What I learned is that it can be built that way. Right? We just have to. Yes, it can be. We have to ask for it. We have to mandate that this is what we want. This is what is needed for our learners to be able to take. I mean, the 21st century technology that we're using, but in a meaningful way, so that it's not just asking them to do it for Them it is truly just a support when needed. So there's ways to do it. I'm learning through the coding and the design of it. But what happens is we can't do that solely.

23:24
Amanda Bickerstaff
Right?

23:24
Michelle Soriano
It can't just be developers, it can't just be edgy. We have to come together and have that collaboration piece to where we're building something even better.

23:34
Amanda Bickerstaff
Yeah, no, definitely. And some really good comments in the chat. Krista and Hi, Krista and Christopher. You guys need to put it for everybody because only us can see it back in the backstage. But I think, let's be honest because you know, I love a provocation. It is so much harder to build a edtech tool that does what you just said than to build a version of a chatbot that's on top of ChatGPT or Claude or Gemini. It's just a lot harder to develop those things. Right. To actually create spaces in which there is like an on ramp, that there is scaffolding. There are multiple entry points and exit points to a lot of our point. Right. But I think that right now this almost like rush to build even more.

24:21
Amanda Bickerstaff
But I don't think, I think with the screen time pushback that we're seeing nationwide, and not just nationwide, but internationally as well, is that there's been this influx for a long time of ed tech tools that have not really been intentionally designed. And it doesn't mean there doesn't exist intentional tools, but it is a lot harder, especially when you're trying to get people to pay you're trying to get kids to stay on the app, that it needs to be easy. It's really hard to get the different systems to talk to each other. But I think that like, it becomes something that it's a lot harder to do what we're saying. Right. And I think we have to recognize that.

24:57
Amanda Bickerstaff
But I think it's more important than ever because of that same friction, like the developers themselves should have, that same friction that we're talking about kids happening, having, and that it shouldn't feel so easy to create something without it really being situated into things like UDL or evidence of learning, like types of approaches.

25:15
Michelle Soriano
Yeah, yeah.

25:15
Amanda Bickerstaff
And I also.

25:16
Michelle Soriano
Oh, go ahead. Sorry, Lindsey.

25:17
Lindsay Jones
No, you go ahead.

25:19
Michelle Soriano
I was going to say yes. And I think we have to keep reiterating and keeping at the forefront. Everybody is different. That variability of. Because I'm seeing in chat a lot of great comments about, you know, students should have the choice to turn it off as well. We might need it in one contextual day to support us with something and we might not want it the next day. Right. Everybody's different. Context matters. And that is part of that agency piece, knowing when I need it, when I don't need it. And so yeah, these are great conversation pieces.

25:53
Lindsay Jones
Add really quickly to what Michelle said, which I was. I totally agree. And I think one thing Amanda, is I can't tell you how many developers I have talked to who really want to build with accessibility, want to have UDL included and get so much pressure from venture capitalists to move fast. They don't care. Let's deal with. They look at the market in this really weird way. Like, we'll deal with that later. Fix it for the normal kids first and then we'll help all the special. And it's like that is just not. First of all, it's not great for many reasons, but it's like not a great way to build for this edge for any education market. It doesn't address neuroscience. But you know that.

26:38
Amanda Bickerstaff
I know. That's why I don't. This is why everyone, like, if you don't know me very well, I was an EdTech CEO. I built technology. I decided I didn't want to do that again when I started AI for education for a lot of those perverse incentives like, it's really hard to like do this well. And it's because we have a really complex system and it's really hard to develop something sticky for that to this level that works every time. Not, not a sense of works every time for students, but just works every time in general. But Previn has a thing about the idea that like one of the things I think a lot about we've been talking about for the whole entirety of the time is that like gender AI can be a system to everybody.

27:17
Amanda Bickerstaff
So Michelle, you just talked about how you learned a to do list. You let your son fail forward. I'm going to be really honest, everybody. I was like good at school. I was good at school until I wasn't good at school because I actually, genuinely struggle with writing, especially long form writing. My brain doesn't want to do it. Like I, I struggle with it, but I kind of like powered through college, I powered through grad school, but then I could not power through a PhD program. And all I can think about as I started to see generative AI, whether it's the idea of for me ideating, getting a draft, finalizing the draft would have meant that I finished my PhD. I wouldn't have said no because of.

27:57
Amanda Bickerstaff
I knew that wasn't something I was going to be very capable of doing at that time in my life. And I think that UDL can. Is for everyone, but also generative AI systems. Their ability to personalize, like, deeply personalized meaning. Like right now, infographics, you know, voice to text to voice. The ability to create images that feel uniquely suited to the different things. Like, that's really unique. But it doesn't feel like that's what people are thinking about. They're thinking about, how can we create this all in one system that does all the things instead of, like, where are the places in which gender bi does something really well today? I think that's my biggest ballet with today is that it feels like genuine, I should do and do everything. But in reality, the jagged edge means it only does some things really well.

28:44
Amanda Bickerstaff
But we tend to see, like, we're just going to slap it on there. But, like, the idea that, like, oh, my gosh, okay, I'm just going to use this one piece 100 times over. Like, some of my kids really need to have it spoken to them, and I can't do that. So we're going to set up a voice agent that all it does is it sounds human. We make sure they know that sounds human, but they're engaged. They can stop it, they can restart it, they can have agency, but it becomes this monolith of, like, all or nothing. And I think that's where we get ourselves into trouble.

29:12
Lindsay Jones
What you described about your PhD makes me think, that is such a good example, Amanda. That is such a good example. We all have examples like that in our lives. And many people will ask me, because they don't use assistive technologies. And I'm worried about ourselves. State bans. I'm worried about our bans on tech. Not because I want more tech, but because I don't want them to stigmatize kids with disabilities. I don't want them to take away rights from students who have the right to use it. But also, like, I think people don't even understand what that means. And so there is amazing promise, especially for these types of technologies. And what the only example I use with people who have never used it is usually people have watched tv. You, I know, have watched TV and seen a Scottish accent, right?

29:59
Lindsay Jones
And you're like, oh, my gosh. And you're listening so hard to hear just the words that you don't know why the guy shot the other guy or why did the car come in here, because you can't Pay attention to the plot. When you add the captions, all of a sudden you can think, oh, okay, I can think about the story. Why is it happening? I can engage with the characters.

30:19
Amanda Bickerstaff
Yeah, the cognitive. You lowered the cognitive load in one place so that creative critical thinking and engagement can go up. Because our brains are not, like, they are not this unlimited thing. Right. Like, we have. Cognitive load is real. Right. And so I think that is such a good example and one in which I think it's. It's kind of funny. I think, like Michelle and Lindsay, I've been thinking a lot about this, and I'm going to throw. This is. I love how you come on a webinar. We think we're going to do one thing, and I'm like, I'm going to blow it up. Michelle was like, I prepared slides. I'm like, so. But I appreciate your. Your.

30:52
Amanda Bickerstaff
Your kindness with me, but one of the things I've been thinking about is with gender bi, the ease to which you can get answer and the ease to which that could over time degrade a skill is something that I think is, you know, like, I just. I. Someone's talking about negotiate choice. So I'm not very good at writing, and then I just stopped writing because I'm only going to be, like, brain dumping and I end revising. Like, could that degrade my ability to write over time? So one of the things I was thinking about is we need some kind of, like, way to show kids how if they stop using one of the skills they've developed, that in a short amount of time, it will go away.

31:32
Amanda Bickerstaff
So the one I think about with adults is everybody in here most likely was, you were all good in Excel at one point, whether that was when you were in college or in high school in a class. And then we stopped using Excel, we all lost all of those formulas and macros and how to run a pivot table, and we have to restart every time. But I feel like it's almost so hard to capsulate where. If you're not using these muscles, if you're not building that cognitive work over and over again, especially as a young person, that you. That either you won't develop them or you'll lose them. But how do we get kids to recognize that? Because it happens in a way in which we are not. We don't recognize when we lose a skill like that.

32:11
Amanda Bickerstaff
But I think we have to get better at showing kids that, like, that bypassing can have a really huge effect and really quickly.

32:18
Michelle Soriano
Yeah, that is great. And that is. And yes, and I think it ties really nicely to that question that's in chat about how do we make sure that they're not losing it? And that when does the teacher input kind of overweigh what the student preferences are? I think it all goes down to the goal. It has to streamline to the goal, right? What is the goal of this activity? If the goal is to focus on writing, practice writing, then obviously there's going to be certain choices that are not available because that's the goal. However, this is a part of that agency piece that can sometimes, speaking from experience, be hard for educators and adults even to kind of relinquish that control into the students. Right? Allowing them to take on that choice to say, okay, here's the goal.

33:05
Michelle Soriano
The goal is not to practice writing, but it's really to have that comprehension piece and that understanding and making connections. What is the choice that you want? Now? I do think there is an opportunity for us leaders in whatever, you know, room organization we're in to have those conversations. Hey, this is why it's really important to consider this option here. Because this option can help build up a skill that, man, I use every single day in my job. Right? So we're connecting that relevant piece as well. But in the end, I don't really know what a child needs, what a student, what a learner needs that day better than what they know.

33:44
Amanda Bickerstaff
They know.

33:45
Michelle Soriano
I have to allow them to think about it purposefully so that they can guide their own learning. I might have some weighted, you know, conversations with them. Hey man, you use this tool three days in a row and you've done outstanding. I bet you can do just as well if we try to switch it up. Let's see. You know, here are other options that you can use in the AIs that you're pulling from. So having those conversation pieces can help with that as well.

34:11
Amanda Bickerstaff
Oh my gosh, I love that. And I love just the idea of like, hey everybody, like we're, we, you know what, we're going to use the stuff that helps us the most and then we're going to do a day that's just about productive struggle. I actually, the goal of today is that you don't actually get a good answer, get a good grade. It is like, I want to see you struggle in a meaningful and non frustrating way about this thing. And for everyone it's going to be a little bit different. But you know what, at the end of the day, we're going to See how much we change and move, and we're going to keep track of that.

34:39
Amanda Bickerstaff
Like, I was never put in a position, and I don't know about most people, but, like, have you ever been asked to just take a step back and do something hard in a structured, safe way? It doesn't and doesn't really happen. We don't give kids the space, or adults for that matter. One of the things I will say is that, like, I've learned more about adult learning theory through this work because I've worked with so many. But like, sometimes that psychological safety of like, you know what, it's a bingo card. We just. We just did on Monday for the first time. And it's just kind of fun and a little bit like, silly, but also like pushing you in ways that you've never thought about before. But letting you do that feels very strongly something that. It wasn't about an outcome.

35:20
Amanda Bickerstaff
It was about the exploration, it was about the trying, it was about the openness, it was about the that. That kind of experimentation. And I just don't think that we do that enough.

35:28
Lindsay Jones
I'm a parent as well, and like, it's so hard. Right. But I. We just had this conversation yesterday at cast. As an organization, we cannot be afraid that we're going to fail. What that does to you is par. It just stops everything, right. It stops innovation. And in some ways, I think that has happened so many issues in our. In our current systems facing students with disabilities. But a lot of it is. Is this instinct to protect, this instinct to. And whether you're a child with a disability or not, growing confidence means you fail, you get up, you try again, and the world hasn't ended. You're able to kind of rethink through it. So that just seems like an incredibly critical skill. We hear it in our early workforce, right. Today, many people. You're never gonna get somewhere if you don't fail.

36:16
Lindsay Jones
But somehow maybe we've changed that. And you know, the other thing AI gives us as Michelle's point is a really good one. It's still all about the teacher. The teacher is making the decision, right. In terms of provide. It's about the student choosing. But they still have in the educator someone after the three days of using the tool, right. To say to them, hey, let's try something else. I bet you in that positive way that teachers. Right. I bet you can do this just as well.

36:47
Amanda Bickerstaff
Yeah. You learn so much. And if not like, let's talk about, like, you might be over reliant. Are you over relying on your aid or you're over relying on your parent or you're over relying on your friend. Like those are the types of things I think can be really great to just have. Okay, so what we're going to do is Michelle, I'm going to ask you if it's okay. I'm going to actually I'm going to share my screen again everybody because we have, we've had such a beautiful discussion and please keep everything coming in the chat because you know that we love that level of discourse because is that we have a brand new resource for you all. I will say this is a perfect example of being two organizations that are getting our butt kicked in terms of how busy we are.

37:29
Amanda Bickerstaff
So the version we have is not fully accessible yet. So what we're going to do is we're going to share it with you all as a link that you should then be able to later today and in the email tomorrow will be fully accessible. So apologies. But you know what, we always like to be transparent. You know were rushing at that. But what you have here is this resource and what we've done. And this is so cool because we had our first conversation about this literally when were about to publish the C framework in April and we had these five mindsets and then they were like but there's also these five UDL guidelines that weirdly they work. We didn't try to do this but we almost ended up doing a crosswalk.

38:11
Amanda Bickerstaff
So what I'm doing is I'm going to pull this up and we're going to talk about it. And again like I said, we apologize. This is not fully accessible yet. It hasn't been tagged. But what we did is we came together and were looking at the CAS Universal Design for Learning Guidelines and then RC framework and then what we did is this little like wonderful handshake that happened and Michelle and Tanner on our team were instrumental in this. But what we found is that if you looked at what we've talked about today, right even Michelle was talking about the intentionality that if you look at like the UDL guideline of strategy development that this it is critical for goal setting and self reception to be brought into any type of gen AI use.

38:53
Amanda Bickerstaff
Like we need learners you already identify in the chat, everybody, we should be able to say no to gen AI use. We should be able to say yes as well or I'm going to watch or modeling like the idea that AI literacy equals AI usage is wrong. It should be our questions intentionality should be, should I use this tool for this thing? And I would include that for not just learners, but also for the adults in the room too. It might feel easiest just to go and give a whole bunch of Gen AI to your students without being that intentional as well. I think also what we want to do is always have a meaningful goal before using a tool. Man, I have not seen this happen as much and I have some.

39:30
Amanda Bickerstaff
I know we have some amazing teachers in the room, so if you've done this before, well, please, please share it. But I think it's really interesting to be like, okay, your goal for today is it with this, with Notebook Alum. Let's go back to Notebook Alum because we talked about it is we're going to have you come up with a short kind of video script about this, you know, this piece of, you know, literature. But then you're going to compare it to the video that the AI did. And like, that's our goal for today. You're not going to just like, you're not going to accept it, but the idea you're going to use this. And then if you're not comfortable with the AI over the AI video, you could do the same thing with the AI overview.

40:07
Amanda Bickerstaff
But you can have that kind of choice based on what you do. And I think the last thing we said was that it's the idea of student reflection. Checkpoints should be built in at every point. And I think that it's not a monolith experience where you're on a chatbot, but like, how is it going? Do you need more help? Do you need a human? Do you need your own time? Like, what do you need to be most successful? So I don't know, did I miss anything, Michelle here that you would like to call out?

40:30
Michelle Soriano
Yeah, no, absolutely. You hit on every piece there. And that intentionality of reflection, I think is so important because even once they do receive their quote unquote answer from their AI, it's okay. Why did you feel this was a relevant answer for the question how did this meet your goal that you were searching for? And so that reflective piece, it's just like this process that's keeping our students and learners thinking.

40:54
Amanda Bickerstaff
Yeah. Yes, yes, yes. I mean, that is okay. The thing is that everyone's kind of pushing like there are a lot of these new studies about cognitive surrender, which NVI systems, but that is just that right now is primarily the chatbots that give you answers. They're not these deeply thoughtful Gen AI ed tech tooling experiences that like, stop you and give you space and then push you off the platform or back into the platform that gave you choice. And so I think that we can't just, we can't also broad brush this as it's always going to happen. It's just happening now because we're not being that intentional in our creation. Right. Okay, so the next one is building knowledge.

41:32
Amanda Bickerstaff
And so for us, we found that like the idea of building knowledge, generative AI is moving from a passive consumer to an active critiquer of gen AI abstracts. So the idea of building your knowledge is also about staying critical. There's not a gen AI tool in the entire world that does not make mistakes, that does not have inaccuracies or hallucinations, that does not have bias in some way, shape or form. Whether that is pedagogical, whether that is ethnic, race, role. You know, we talk about weird. Oh gosh, English speaking. I can't remember exactly. Now we'll put it. I'll get, I'll get Wendy to put weird in the channel. But the idea of saying critical is like every gen output is a draft. You must fact check it every time and it must be iterated and revised. You would never take.

42:17
Amanda Bickerstaff
Lindsay is a beautiful human. She would never take something out of gen system without reviewing it and making it our own and sharing it. Because that would be. First of all, she's not building her, like she's not having a voice, she's not building her own learning. But also she's not being critical consumers of the technology. And that's really important. It's always important to demonstrate how to question a gen response. These tools unfortunately are sycophantic. So even if you say you're wrong, they'll be like, Amanda, you're right and you have nice hair. It's a very silly thing. But learning how to question and then be like, you know what? This is not the right tool for me right now. Again, going back to making those decisions.

42:54
Amanda Bickerstaff
Because if you're saying like, I know that this bot is wrong, it keeps telling me that I'm right, then that's when you like that is like you have hit the jagged edge. It is not working and we need something different. And then also that last thing is what we just talked about, that the idea of is this actually undermining my learning. Man, if we get kids in any way, shape or form to be able to ask that question of themselves, what a gift would that be even without generative AI? But because generative AI is So much more easy to use. It is more important than ever before. Like it is that easy button. It's just so ever present and it is a siren song.

43:33
Amanda Bickerstaff
It is, it is taking you into this space where it feels safe and you're doing good job, you're getting all those A's. But is it actually undermining your learning? Is something that I think is really important and couldn't it. Would it be amazing, Michelle Lindsay to see a gen AI tool that ed tech provider, they actually asked the kids did this help you learn or not? So that it's actually built into the refactoria process of the actual developers work. But okay, Michelle, again, anything I'm missing?

43:59
Michelle Soriano
No, this is great. This is great. I was just going to say nowadays, you know with AI you can see the process, you can see the learner's process of how they interacted with the AI. So that with the rubric. Right. How did you transition it? How did you change it? How did you adapt it? How did you add to it?

44:16
Amanda Bickerstaff
Right.

44:17
Michelle Soriano
That's also another way that we can look at that learner's learning.

44:20
Amanda Bickerstaff
Absolutely. And I think that. But what I love about this is none of this is crazy. None of this is crazy at all. It's all like just very, very like you know, it's taking apart. It does require gen literacy because if you don't know that the tools are predictive that they that are based on training data sets that have bias built in and have gaps in it, then you don't know. So it still requires gen A literacy right from the individual. This wouldn't just happen. Like kids would not be able to do this on their own. But as Lindsay said so beautifully like so many times is that like the thing they can control is the learning design. And some of the learning design should be AI literacy and really reinforcing these practices. So the next one is that emotional capacity.

45:02
Amanda Bickerstaff
We thought that was just really about trust. Right. And support and that's that transparency that be transparent modeling norms about classroom expectations and how to disclose AI use, stopping putting it into the shadows that it's bad. God, we have made kids fear this technology not because it's undermining their learning, but because of everything else. Like we have made it such a big bad and kids are internalizing this in such a meaningful and potentially detrimental way. Modeling your own use. I will say so Lindsey and Michelle, our favorite thing in the world to do with general BI is to model like we, you know, we model almost everything we say this is where I made a mistake or I didn't push back or here's how I thought about context. I'm writing like that modeling of that reflective thought.

45:50
Amanda Bickerstaff
This is when I did this one time and it took over my voice and you know what? I just let it happen cause I was busy or you know what? I was lazy. This is a true story. I was lazy doing a project and I just like did it the. I just did it in a lazy way and it took me like way longer to do. And I was just like, if I had just stopped, either done it on my own or recalibrated and came back to it with AI, would have been a lot better for my mental health and wellbeing. But you just get into these places in which that reflection can really happen. And then finally that emotional capacity is. These tools do have an impact on other people.

46:23
Amanda Bickerstaff
It might work great for you, but what about your friend that is using the tools and it doesn't work as well for them or that it can have these biases? We know that these tools can be biased around neurodivergent young people, non native English speakers. It also has a climate impact and an impact on others. And so the idea of that emotional capacity of building that level of nuance and transparency, of talking about it and not being afraid to, I think is really important. So Michelle, anything I missed here?

46:53
Michelle Soriano
You covered this. All right here. This is great.

46:56
Amanda Bickerstaff
Perfect. Okay. And the last two are probably. I will say I'm definitely going to call Lindsay and Michelle for this last one to keep learning. But the act response to like the welcoming interest in identities, like we want this to be for all of us, which means that we need to as learners and as individuals, that we are always vetting tools for accessibility, for privacy, for bias. We must always surface tool limitations. If you are a builder in the room and you have a tiny little thing that says this could be wrong, it is not enough to have on ChatGPT. And I'm gonna say this very strongly on edtech tools that this might be a mistake. And actually I'm going to share my screen because I just want us to take a moment to look at something that could be so wrong that where.

47:38
Amanda Bickerstaff
Nothing here says anything like that. Where, where does it say. It says it right down here. ChatGPT can make mistakes. Check important info. See cookie preferences. Great, thanks OpenAI. But the idea here of like literally this map has mistakes in it. But that's not intentional. But the idea that's not enough if you are designing these tools, I just can't say that strongly enough. And then finally, like, what is the impact on your personal voice and autonomy? So out of all of these, I would say the one that's most important to the conversation we've had today, especially from students, is like, is this replacing my voice? We know that these tools are predictive, which means that they are going to be pattern predictors and they will automatically flatten voice to a, essentially a mathematical mean unless you do a really good job of training this.

48:25
Amanda Bickerstaff
Like, you can still get around that, but the majority of our users, it will happen without you even realizing and it might look better or sound better and be nothing like your voice. And so it's really important to always have that, that responsibility to yourself. Right. And your responsibility to your community. And then the final one is that sustaining effort of persistent and to keep learning. These tools are weird guys. They make mistakes, they're strange, they don't work as expected. And you know what? They can also be harmful without the right approach. So the idea of embracing that productive struggle we talked about more than once today, creating space to create and to risk take to do things wrong and have them. Like, one of my favorite things you could do is have kids.

49:06
Amanda Bickerstaff
Kids think that they're doing a Gen AI assignment and they can do whatever they want and then be like, the assignment doesn't matter, but the ways in which you interacted with the bot do. So let's look at your chat. Did you push back? Did you ask questions? Did you accept the output without reviewing it? Like that itself can be the entirety of that learning experience. And then finally, AI literacy is durable. It is critical thinking. It is all the mindsets that we have here that it doesn't matter if you're doing it with generative AI, you should do it with everything. If we can get to that level of critical, engaged thinking, that responsible thinking, that transparent thinking of any technology or non technology, I think it's going to be a better world.

49:41
Amanda Bickerstaff
So I'll stop there and I'll hand it over to Lindsay and Michelle. And then we just hope that this is a really like really good resource for you all to bring back to your organizations. But Lindsey, I'll let you take it away.

49:51
Lindsay Jones
Yeah, and I think a great place to end is actually a question in the chat. At what age can students start to learn about how they learn? And we would say very early. But do you want Michelle to talk about that? I mean like immediately, immediately.

50:07
Michelle Soriano
And I saw that pop up And I immediately thought of my grandson who just turned two. Right. And I was thinking about learning colors and things and how I was watching at such a young age. You asked me what color is blue? And he goes and chooses one. That's incorrect. H. Is that blue? No. Right. He can assess his own learning at the age of two and I would say even before that. So I think immediately, I think it's prompting and we have to question. Yeah, absolutely, there's going to be failing forward. But I think that analyzation is happening at a very young age. I mean, immediately. That's how we're able to learn as humans.

50:45
Michelle Soriano
And so I definitely think the younger that we can just start pushing this and start asking and analyzation and reflection, it's going to help build all learners as we grow.

50:57
Amanda Bickerstaff
Absolutely. I mean, things like asking like a five year old, an eight year old was this hard. Why do you think it was hard? Can you notice when things feel harder or easier, like those types, like, why do you think that they're easier? Like, what happens if we do it this way? Does it become less hard? And why do you, like, it's that metacognitive, like, role of like asking about thinking. But it also, it doesn't have to be complex. Not like, oh, what is your cognitive load? Or like, it could be something as simple as, like, you know what, I notice that I have, I struggle with writing. I notice that you don't. Why do you think that is? And like, what do you.

51:34
Amanda Bickerstaff
Or I notice that you struggle with writing a little bit like me, but not when you're writing about your little brother. Like, what? Like, is it that you need to be inspired? Those are the types of things that we can do at a very young age that can lead into significantly more robust metacognitive work as they get older. Right. And it doesn't mean that like, you know, in some cases it might mean that kids avoid things, but it's a lot harder to like, if I know I am avoiding something, it's a lot harder for me to justify avoiding it. It's very easy for me to justify avoiding it if I don't understand why. Oh, it's just, oh, I just don't have time. Like, we're really good at like, you know, like not telling ourselves what's really happening.

52:10
Amanda Bickerstaff
But I do think that it is very important to like have that built in. And the worst thing that happens is kids look at you like you're crazy and you move on. And as teachers, you know what we get looked at like we're crazy all the time, but you get better at it. You know, if you care about it enough and you're. You're about it, there's an opportunity. But we're coming up on time. And first of all, I just want to say thank you to everybody in the chat. We like, this is why we do what we do. To have such a diverse, beautiful group of perspectives that are deeply thoughtful, that actually align with UDL really well. And the mindsets that we talked about is really important for us.

52:46
Amanda Bickerstaff
But also, I just want to say thank you to Lindsay and Michelle, first of all, because as always, I took us off track, but also because I think the quality of conversation and the importance of you all doing the work that you do, but also not just UGL on everything. Like, it has to better. Like we owe our young people something better with everything that we do, but with Genai specifically, with the risks and the moment that we're in, it has to happen. And if, whether you're a tool developer, whether you're a researcher, whether you're an educator, this is so important. So I just appreciate Lindsay and Michelle, you guys, and don't forget, we have the updated resource which the tagging will be there for you by tomorrow. We're also going to share the recording if you want to share it with everybody.

53:27
Amanda Bickerstaff
But I just want to say thank you to everybody and thank you, Lindsay and Michelle.

53:31
Lindsay Jones
Yes. And thank you. It's great to be in partnership with you. We appreciate it.

53:34
Amanda Bickerstaff
Absolutely. Well, I hope everyone has a beautiful rest of your day wherever you are in the world. We appreciate you all. Bye, everyone.