House Standing Committee on Education

Good afternoon everyone. We are now convening the Joint Committee on Higher Education and Education for the purposes of an informational briefing on the concept on the study of Artificial Intelligence in education. I am Andrew Takuya Garrett, Chair of the Higher Ed Committee.

To my right is the Chair of the Education Committee, Representative Justin Woodson, and to my left is the Vice Chair of the Education Committee, Representative Amato.

So in terms of the informational briefing today, we do have a slew of very well qualified presenters on this topic, excited to hear about the discussion today and before we turn it over to them, just had a few opening remarks on behalf of Chair Woodson and myself and then we will transition right into our first presenter.

Okay, so thank you everyone for being here today as we examine one of the most transformative developments in education. That's the rise of artificial intelligence and its implications for our schools, universities and students. AI is no longer a distant concept of the future. It is already reshaping the way we teach, learn and assess student progress.

From personalized learning platforms to AI-assisted tutoring and even automated grading, this technology holds incredible promise. But along with these opportunities come pressing questions and challenges. As policymakers, we must ask, how do we ensure AI enhances learning while maintaining the critical role of teachers? How do we protect student privacy and prevent bias in AI-driven systems?

What skills will students need to thrive in an AI-driven workforce? And how do we ensure that all schools, urban and rural, well-funded and under-resourced, have equitable access to AI's benefits? Today's discussion is not about endorsing or rejecting AI. It's about understanding its implications and making informed decisions for our education system.

Our job as legislators is to ensure that innovation serves our students, not the other way around.

Again, we are very fortunate to have experts with us today who can provide valuable insight into AI's role in secondary and post-secondary education, its impact on academic integrity, workforce readiness and digital equity, and the steps we can take to craft policies that foster innovation, responsibility and inclusion.

So with that, I look forward to a very productive conversation and working together on solutions that prepare our students and educators for this rapidly evolving future. And with that, we will invite Gabe Yanagihara to the table. Gabe is an educator.

I was going to hide his employer, but he has a big name tag that says where he works. Unfortunately, it's not a Department of Education school, but he is an innovator in the AI space and has been-- he actually does consulting work as well too, preparing organizations to get ready for AI adoption.

So with that, Gabe, the floor is yours.

Awesome. So welcome everyone. Today I'm going to be talking to you all about my experiences with AI in the classroom and as a teacher to try and illustrate as much as I can in the time that we have.

Just so you can get a slice of life of what it's like to be a teacher. Let me see if I can click to the next slide. Which computer is it on? That one. There you go. Awesome.

So I fed it all my information, all my old lesson plans and curriculums and everything when this tool first came out. And this is what AI illustrated for me. So my name is Gabriel Yanagihara.

I am essentially the guinea pig over at my school, anytime new technology comes out. I'm our head coach for esports program, 3D printing, laser cutting, VR, AR headsets. We got one of the first Oculus Rift headsets to let other students play with it.

So when ChatGPT first came out two and plus years ago, our head of school turned to me and basically essentially promised he wasn't going to fire me to let me figure this all out. So within two hours of using the tools, pretty sure I'd replaced two of my co-workers, entire jobs.

I had been writing essays and lessons. It was generating all this work. And this is 3.5, the older version. And at that moment we knew this is something that was going to change the educational world, change the very fabric of what it meant to be an educator and teacher in this future.

So what I did, what any self-serving teacher would do, I threw it all care to the wind. And I completely replaced myself with AI as an experiment.

My goal was to try and make all the mistakes as quickly as possible because we knew there were going to be hundreds of them and everything was going to have to change. So I'm still employed. I learned a lot of lessons, so I'm here to share some of that with all of you.

So the second thing that we did, once I was able to explain and try out all these tools for myself, I then did an AI boot camp with my students.

So we took students from fifth to eighth grade, six hours per day for a week and we dove in, simulating and practicing what would it look like with full AI use of every assignment and curriculum and system that we had in place to see what broke, what went well, where there are bright spots and exciting opportunities. And from those opportunities we went off and trained everyone we could.

So I've been on a mission. We've trained over 2,500 teachers across the state. Pretty much every school that will hear me talk, I've gone over there. I've done training, I've done workshops, everything from Sacred Hearts to Maui.

We even flew to Moloka'i with STEMworks to kind of share what these AI tools are able to do and get them in the hands of teachers and students to see what projects they could do.

So I want to share some of the bright spots that have come up from these two years of craziness that have been going on with the world of AI in Hawaii. First one being-- So I'm originally from Maui.

I was born in the backyard out by Wahoo, and I went to Pukalani Elementary, Kalama Intermediate, and King Kekaulike High School. Recently, after the Lahaina wildfires, I had the opportunity to return.

And when I was talking to teachers there, what we saw is in the redesign and in all this work that was being done, the students weren't the ones showing up and telling and putting their thoughts out on paper of what they want the town to look like.

We have this big challenge with brain drain in Hawaii and the challenge of how do we build a future where the students actually want to move back to the town and stay there and have their families and contribute to their local economy?

So in that initial prototyping everything, Yayoi from the Jodo Mission Temple on Front Street came up with the idea of doing postcards from the future. But that project was hamstringed by the idea of you would need a multimillion dollar consultancy firm development side to get all these architectural renderings done.

So what we were able to do is just put the students work and their ideas in front of AI. The students were able to give me their prompts, and I typed them into the tools. This is ImageFX, which is free by Google.

And they were able to create images of what they wanted to see in their town. So I was able to bring to light, bring to vision what they wanted to do, and they were able to collect them.

We hung out for a couple of lessons, taught them their interview skills, taught them all those things that they needed to do, and then they built the postcards with images and visuals and graphics of what they wanted. Obviously, you see the lifted Yota trucks there and everything, right? But they had a blast for it.

This is the first time that they saw their visions being shared. And as a result, for 10 cents of printed cards, they were able to send these to the decision makers, to the people who were making the decisions for their community. Obviously, not everyone's going to get a lifted truck. But they had stories from the students.

My family would always meet at this park and we would have multi-generational meetings where the grandkids would meet the grandparents and stuff like that. And I want a space for that. Other places like I want to be more local shops. My family has a business there. This is how I'm able to live here.

I want to make sure there's still space for that so they're able to generate and make their voices heard. They then took those skills and then they interviewed the older folks in their community as well.

So once they were able to have their ideas, they then took those skills and talked to their grandparents in the kūpuna, in the community to get their ideas as well. So yes, AI was really cool. It made these cool pictures for us and everything.

But I do want to reiterate it was the teachers and the teachers ability to use these tools that made this happen. It was the nonprofits, it was Yayoi. It was the rest of the volunteers who came together to make that project happen.

So the next one we want to work about. I've done a lot of work with STEMworks, which is Maui Economic Development Board. They've been doing some amazing work in the after-school programs, free after-school programs for all the different schools.

And they really saw very early on that AI was something that they wanted their students and their populations to have access to. So we've been going to Maui, Molokai and beyond with wherever they have their programs. And I've been doing AI trainings for them and their students.

So one of the projects that we did was one a little bit of the AI literacy, right? Just teaching them, just giving them that first permission to look at these tools and say what's possible but then to turn that into stories, right?

Each of these students, if you all have kids who draw stuff and put it on the, you know, on your, on your fridge at the back of the room kind of thing or you know, illustrations on the fridge.

But now what we're able to do is get them published and printed for less than three or four dollars per book shipped to Hawaii, right? So now they're able to make books and stories about the things. The most popular one, and the first one is a book called Kaiona.

We ended up partnering with HIEMA, so Hawai'i Emergency Management Department. And then for them we created a book teaching tsunami safety. So instead of a PDF and a pamphlet and a PowerPoint, the kids aren't going to watch any of that stuff.

It's very good information, but the students said, "Oh, I weren't going to read any of that, but if you put it in a cool storybook, that would be awesome." So then the students worked with us and built a book. We published it, we printed it, and that entire project from zero to printing and ordering the book was $0.

These tools are free. You can download them, run them locally, and they have access to make all their visions come to life.

So that's one of the projects, the most popular ones and the easiest, first tool that I've seen teachers adopt is the idea of using the AI tools with the students to have them illustrate their own children's books and stories. But in the same vein that we're like, AI can write any compare/contrast essay nowadays, right?

But what it can't tell you is your personal life story or your family's story.

So each of the students can now generate perfectly professional quality images and visuals to tell their family's history, learn about their family's history or even local myths and legends from just around their school area, and then get them up on Amazon or Shutterfly or whatever printed. So the format was super easy, right?

They just come up with an idea. This was one of the students. It was the story of a coconut that could. The stories were a little rough, like the coconut fell over, its brain fell out, and now it's trying to figure out what to do.

But the kind of crazy out there stories that you wouldn't spend 20 grand of illustration work to get published, but a student, that's their first published book and now they're published author by seventh grade.

The next capability that I've seen come out of these AI tools in the classroom is not specifically with students, but with our communities as well. So this is Greg Bowman and Alan Suemori. They've been teaching way longer than I've ever been alive. And they are legends kind of in my small teaching circle.

They're the teachers who wake up early. They greet every student at the door with a handshake and ask them a specific question about a family member and I have no idea how they do it.

But as a result of that selflessness and their willingness and desire and drive to be a teacher, they're not the type of person who's going to pat themselves on the back and say, I should hire a ghostwriter and tell a book about my life story and sell it online, right? That's not who they are. So I forced them to with AI.

We sat down, they had their kind of ideas of what they wanted to tell the next generation of teachers. We used an AI tool to listen to our voices and take notes.

So we did a four hour conversation with them, turned it into a table of contests, worked with our English Department and were able to publish a 200 page book called "Teaching Joy" based on their teaching philosophies and histories. This example has repeated itself several times over. I was able to present to several other teacher groups.

There's a group called the International Society for Key Women Educators. And I envision a future with AI where every one of those women, every one of our local leaders has a published book of their story and how they got there in the library which they graduated from.

So that students in that library can not just read about Harry Potter or like some long, far off person, but someone who actually grew up and went and sat in the chair where they sat in and had those tools for $0, right?

The expensive part was the time and effort it put in to learn the tools and be able to break down those barriers and try all that stuff. So one thing I want to say is that right now with all this deep seek and all these crazy, like, financials happening in AI, really all eyes are on us, right?

We are the educators in the space, right? Everyone behind me here, everyone's working on these things. So we're in a very unique role in that. We have so many stories that haven't been told that should be told. We could be the ones in the national spotlight showing what these tools are capable of in good hands.

So I'm really excited to see what everyone else is able to do with the tools. I want to share a final couple examples from everyone. The first one being intro computer science. I teach a seventh grade video game design course where the boys, very boy-heavy class, they like making video games.

And for the last 15 years since I started teaching in 2009, I've had a lot of fart games and like poop games and silly games that they have. And for years I thought that was just because boys were kind of immature and hanging out and everything.

And then I introduced AI tools where they could put their rough kids sketch in and it would turn it into a final quality. So this is a tool called Whisk by Google, again free, and the games changed overnight.

So it wasn't so much a fear of just I don't want to take it seriously, I just want to do a silly game. It was a peer view. My skills haven't developed enough yet, but I have these amazing ideas, and I don't want to be seen as someone who wasn't able to bring that to light.

So being able to access the tools to bring what they had in their mind onto the screen in front of them greatly increased the quality of their work. This is a creative writing Dungeon and Dragons class. I taught this over the summer and after school.

This was a girl's project, so she's in the fourth grade, and I don't think she's ever rewritten something in her life. She's struggling with writing skills. It's one of the things that she was dealing with.

And what they had to do is they had to describe their fantasy character, their own Lord of the Rings, Harry Potter type character, and bring it to me, and I'd type it into Image FX or Midjourney or any of the tools, and then it would illustrate it for them based off what they wrote.

When I first did this heroine for her, she snatched the paper away from me and said, that's not what my princess looks like. And in my head I'm like, yes, that's because your writing kind of sucks, right? So we're working on that, right?

So then that's the first time she ran back to her desk, scribbled it all out, rewrote it, and that was the first time I'd ever seen that student rewrite work. So, again, an innovative use of these tools to get the students to engage with their writing.

Not having the AI write for them, but encouraging them to write again and again and again until it's to a quality that they like. It happened again and again and again. So one of the things I wanted to leave us with today, I know we have a little bit of time.

I want to leave some time for conversation and stuff as well. In our hands, the people in this room and the educators and everyone here, AI can become a force for good. We're not going to use it to do sales force. We're not going to try and sell people more stuff in Temu or anything like that.

It's kind of giving power and voices to all these cool stories that we can do with the resources of an Internet connection. A $20 disposable cell phone and the connection to the Internet. Okay, so I did want to wrap up with a final one. There are some of the challenges that we have integrating AI tools.

So we had two years of change management at our school, doing full PD days for teachers, having working groups, doing a lot of change management with everything. One thing that we're seeing is that it's pretty much ubiquitous in that all students are using these tools already.

They are developing not necessarily expertise in these tools, but potentially reliance upon them because they're not being guided through proper ways of using these tools. So the sooner we can get literacy going with the students, the sooner what we can do.

So what we've been doing with to address AI, we had full faculty professional development and tools training. We have ongoing workshops, so I've been doing consulting work and PD work pretty much to anyone who will have me to make sure everyone is aware of what these tools do, how they work underneath the hood and so that we can start doing that, right. And it's not a one and done situation.

We're doing like almost every week I'm doing a different professional development training, whether it's new AI tools, concepts, projects, stuff like that just to keep that conversation going on.

For students, starting at the seventh grade, we have middle school students learning with dedicated AI literacy training and lessons so that all the students are going to be starting to build an awareness of what the tools are, get hands on with them, and how to ethically evaluate these tools on whether they should use them or not for different projects.

And then finally we're starting with a student advisory group. So these are students who either were nominated by a faculty or someone in their department for someone who knows AI stuff or has been experimenting with it. And now that group is serving as an Advisory Committee to us as well.

So they can come back to us and say, "hey, here's what I want." Just yesterday we did a big panel discussion where we had five of them in the front of our big conference hall with 50 something teachers. And they basically told the teachers what it's like to be a student today with AI.

And that opened many eyes and we were able to continue that conversation with them as well. Okay, so two final statements. The first one we're-- me personally, I anticipate AI to be as ubiquitous as cell phones and email ongoing. And we all know how boring checking your email is.

My job is to make AI as boring as that. The sooner we get past this hype cycle, we figure it out, we do all the change management we need. It's not any different than how we've had to adapt to all these previous technologies. One of the frameworks that we use, there are many free ones online, we do one where the students understand how the technology works, how it's trained, how the data and statistics work.

Then they use the tools. Right, we have sanctioned tools that are school appropriate. I know there's Magic School, School AI, there's the ChatGPT for Edu coming out. There's many different wrappers and tools that are out there.

And then we have the students have thoughtful discussions and evaluations on all of it as well. All right, I think that wraps up most of it. I wanted to open it up for any questions or feedback or anything like that. What do you guys want to know?

Thank you, Gabe. I think just in the interest of time, what we'll do is go through all three presentations and then we'll save the end portion for questions. We'll ask all the presenters to come back up, and hopefully we get a few more Committee Members to join us at that point.

I'm sure they're watching on the live stream on YouTube as we talk here. But thank you very much, Gabe, and we'll have you back up here in a minute. Up next, we have the Department of Education presenting. Ms. Heidi Armstrong, our Deputy Superintendent of Academics.

And we will give her a second to settle in and queue up her slide deck. Give us one second.

Let's just recess briefly. Recess.

Okay, thank you everyone for your patience. I think we got the IT issues squared away. So as I was saying, we have Deputy Superintendent Heidi Armstrong here with us to talk about the Department of Education and how they are tackling AI. So with that, Heidi, the floor is yours.

Thank you. Good afternoon. Chair Garrett, Chair Woodson and Members of the Committee, thank you for this opportunity. And I think it's very timely to follow Gabe in this presentation because he showed so many examples of ethical and responsible use uses of AI.

And what I'm going to do is share how we're introducing, supporting and growing our teachers and students in learning about AI and what it can do so that we can use it as a tool to support learning. Absolutely not replace any teachers.

Teachers are the most important role in this whole rollout because it's through the teachers that the students learn how to use the technology and how to embrace it to make their learning experience more enriched. And so what you have in front of you is really a PDF of our website, which is our guidance for our employees.

So you should have this document. Okay. So really this is a website, but it's easier to show you on paper for this presentation. And so I just like to start from the very basics.

So we're very well aware that the implications of this rollout are critical and we need to make sure that we do it right Department wide. We have many examples and you saw some. Some of those might even been our students that Gabe showed. Where the teachers are taking off and the students are taking off.

We have elementary schools already winning competitions. You know, our middle schools are embracing it. We have many classes in high schools. But as a Department, how do we do a smooth rollout where we can prevent any negative implications of AI from happening?

When we're talking about AI, it means it's a machine based system and it can make predictions, it can make recommendations, it can also make decisions. There's three common types of AI. If you look at your handout on page two at the bottom, you're familiar with Alexa or a game of chess that you can play on the computer.

That's an example of reactive AI. Predictive AI you might have been shopping on Amazon and it gives you welcome back. Based on your previous purchases, you might find these items interesting. That's predictive AI. For this presentation, I'll be talking about generative AI.

Generative AI is when you give a prompt and then that prompt is going to produce an output. Okay, so if you on our training for teachers and students is ongoing and for teachers, how this Generative AI can assist in helping with their grading, helping them with lesson plans.

Many of our schools have their academic plans that are due and they're using AI to help put out the first draft of that. So AI from the employee side can be very beneficial and such a time saver.

But we still need the critical eye and the human brain to determine if what AI put out is the product that we want and what criteria are we using to measure that product for its effectiveness and for its use for our students. Some of these were mentioned through Gabe's presentation. AI has many wonderful personalized tutoring programs.

AI can also be used to help students with their assignments. And this is the area where we can continue to navigate. For example, AI gives on demand assistance. So a student has an assignment compare and contrast two articles. Well, AI can compare and contrast that two articles for them.

If we're using an AI to enhance and advance education, part of our learning and our teachers learning is now what do we use AI for? What do we still need? What skills do we still need to teach the children?

And then how do we change our questions or our assignments so that we can harness the power of AI, yet still allow students to develop their skills to further their learning? We're currently navigating that right now. We don't have all of the answers for that. If you turn to page four in your document.

As we navigate the evolving landscape of AI, we have to be clear. We need to provide clear guidance for our employees regarding data protection. It's very easy to let personal information, sensitive information get loose on AI and we want to prevent any of those issues from happening. So we do help.

Our training includes helping employees navigate data protection. We also emphasize the human involvement in the use of AI to validate the information. And that would be the training on human oversight for fairness and equity. We want everyone to uphold the ethical use of AI and prevent plagiarism or misuses of AI.

Our training also includes best practices using AI tools along with that focus on ethical use and that's our accountability. Evaluation of the use of AI tools will be ongoing as they continue to evolve. And then our biggest focus and our biggest challenge is the professional development, because all of this is accomplished via a robust professional development portfolio.

And if we go to slide 4, our professional development opportunities include guidance for staff and students. And that's the website or the document that I'm walking you through. Also, I've given you a sample of our professional development opportunities that have been offered to our teachers. And that's the AI Literacy in person training handout.

So examples of AI trainings for our teachers include Art, Creativity and generative AI, Deep dive Prompt engineering, Generative AI for educational leadership, AI for effective assessment, Notebook LM your power, teaching assistant, the power of conversational AI, et cetera. And so this handout gives you a snapshot of the number of teachers trained just in these areas.

And these are our in person trainings. We also offer asynchronous AI courses and if you look in your handout on page 8 and 9, these are available for all of our employees at any time and they include. zero, I'm sorry, that's on page six. Sorry, page 6 and 7.

The topics of these asynchronous courses include the power of conversational AI, overview of AI in the HAIDO system, artificial intelligence in the Educational Summit. We also have a toolkit available that helps them with citing styles and formats and also acceptable use guidelines, technology use guidelines, and of course our code of conduct.

So we integrate all of this into a series of professional development opportunities that are both in person and asynchronous to build the capacity of our teachers. We have teachers who are just exploring this tool and we have teachers who have learned from Gabe and participated in many of the exciting projects that he showed in his presentation.

You go to page 5 in your handout. Key topics in our AI guidance for students include approval, private personal information, bias and misinformation, safety and transparency, and citation. If you look at number one, this is a poster that we'll have available in every classroom.

Number one is our, not our biggest challenge, but an area where we continually are navigating and that is the use of AI for schoolwork. You'll need your teacher's approval, but this is our responsibility to develop the skills of our teacher Teachers. When is it appropriate to use AI and when is it not?

We also focus our students attention on the importance of private and personal information. It is so easy to get entrapped with giving information that should not be given to AI. And so that is another focus of ours. We also are striving to teach our students about bias and misinformation.

Just because something comes up from AI, is shown from AI or generate generated from AI does not make it right, does not make it correct. And so how do we help students navigate that arena?

We also of course want them to be safe as they use it and also ensure that they properly cite any work that is generated through AI and avoid any plagiarism concerns.

Our state distance learning program, starting in School Year 2223 utilized Khanmigo to act as a personal tutor for students using AI, and we had about 400 students in that pilot. The pilot allowed us to test Khanmigo and create a pathway for schools to be able to use so that we could scale.

The pathway setup allowed schools to receive a template for their data sharing agreement as well as implementing tips and lessons learned. And I think this is a model that we will continue to use as more AI tools come. Our way is to pilot the have lessons learned and then scale.

Magic School is another AI program that allows teachers to pull AI tools such as Chatbox Chatbot, which is a writing assistant, to utilize directly with their students. Another tool that teachers may use for Magic School might be a unit or a lesson planner. There's a wealth of tools on here.

And again, this is a pilot program, so we're piloting with a group of teachers. And this pilot will allow us to learn and successfully expand the use of this tool within the Department.

In addition, teachers are being trained to use Gemini and Google Notebook LM to be able to use these generative resources to support and enhance their practice. An example I gave earlier, they're working on academic plans or unit or lesson plans. They can put the information in and a draft plan will be generated.

They can put standards and core expectations in a certain content area into these tools, and these tools can generate their lesson plans or their unit plans. And so our goal from the department is to scale this.

We want to embrace it, but we also want to use it responsibly and ethically and ensure that all of our teachers and students are equipped to do so. I have with me Dr. Winston Sakurai, who leads this effort. And we're available for any questions afterwards as well. Thank you.

Thank you very much, Deputy Superintendent. Up next, we have three heavyweights from the University of Hawaii at Manoa. Campus Provost Bruno, VP Syrmos and VP Yoshimi.

I don't know about that.

Intellectual heavyweights. Floor is yours.

Good afternoon, Chair. Chair. Thank you for having us. We'll have no presentation. We're three engineers here, two WS electrical engineers. Michael is an oscillation engineer. So we're going to be the most boring presenters you've ever seen. And they are my two agents, because generative AI is a thing of the past. So we're going into a genetic AI.

So my. These are my two AI agents. So before I start, I want to start a little bit about the history of AI just to put things in perspective. It is a great tool, great technology, Everybody does it, very few know what it is. So it's nothing that is new in concept.

Actually it started in the school that Michael did his doctorate degree at MIT in the late 80s, mid-80s, it was a famous, it was called the LIDS, the Laboratory of Information Decision Sciences. And then it died. And in the mid-90s, early-90s, they decided, oh, it's useless.

And the reason that it was useless is because there was no computing power behind all these great algorithms, great decision making processes. So we're coming 3035 years later and you see that explosion of computing capability and computing power you have. Nvidia chips now are being built just for AI applications. It has become a household name.

That's a little bit of a history. So the University is not coming into this totally new.

There are a lot of our faculty that they have been working on this technology from the R and D part of it for the last 15, 20 even maybe some of them 30 years when it wasn't as sexy as it is today. So at UH, we have put our AI tools or AI concepts under five pillars.

One of them is governance and ethics. The other one is outreach and engagement with respect to AI, our bread and butter as a University, a lot of that, especially at Manoa as a research. One University is the research innovation piece of course.

Another huge piece for us is the workforce development and that includes the educational piece within all 10 campuses. To me the most exciting one is the economic development of engine and opportunity that AI is on. Industries, high growth industries. We're here in the State of Hawaii. I think we're very good and some of them are health.

Health care is going to change very soon. Actually it is changing as we speak with the insertion of AI in that technology, creative industries. Gabriel talked about esports, gaming. This is another area that explodes and the state actually has a lot of potential to bring it further. We're also using AI in manufacturing and manufacturing, advanced manufacturing.

Now it's called AIM, Artificial Intelligence Manufacturing. So that is becoming another big area. We have large research grants from the National Science Foundation. Hopefully we'll keep those grants going forward.

We have an NSF large grant, around $20 million which is actually insertion of AI and data science in climate change, climate resilience and how we monitor these effects and how we make decisions going forward. And then let me see if I hit them all in agriculture.

We have some really good things going in agriculture and that's how we are organized. The way we see governance is very important. I'm going to start with a boring one so we end up with the more exciting ones.

Governance is very important, not just because of plagiarism or other issues around AI, but in the University we have research that involves human subjects, it involves animals, it involves FERPA, it involves HIPAA, so clinical trials. So all of that is as important as AI itself.

It is whether it's generative agenda or then you go in the physical systems of AI. So this is a big piece. Garrett is leading that and his group is leading that effort. And then the engagement piece and how AI, the great thing about it, most of these tools are free.

So it's an open source community, which is a great thing for us and for our kids and for our teachers. Right. Because when we do that engagement, we are actually trying to change the way our own faculty present their curriculum and they teach.

You take all the fluff out of that and you actually elevate the critical thinking in the class and in the curriculum.

Because there are a lot of things that you can do through generative AI, but one thing you can do it is the critical thinking and how we actually change the conversation from our teaching, starting from arts and humanities to physical science to life science. In order to do that, it is a big conversation.

When I hear the teachers don't always teach the students. Actually the students sometimes teach the teachers too. This is an example. Some of our brightest young kids actually are more competent and more fluent in all that AI data science components that our faculty, however we work through these issues.

Michael has actually a pretty robust effort within Manoa on their curriculum and he will talk to that. Then finally again to me personally, there's the opportunities in the economic development. It is an engine that is going to be unleashed. You're going to see agenda AI in the next couple of years exploding.

So it's all about agents that they help you to do your job better. And then after that, the last piece of that, of that AI is going to be on the physical AI. When you have humanoids, you have flying cars, you have self driven vehicles, and that is the Holy Grail at the end of the road.

And I don't think we're very far from that. We're entering the agency era, but we're not very far from the physical area of AI. So with that I would like to turn it first to Michael. Do you want to speak a little bit about the great stuff you guys are doing in Manoa?

Thank you. Chairs, Members so this time is not unlike a number of different crossroads we've passed in both higher education and in K12, I think our colleagues did a great job earlier setting the stage. My unscientific survey suggests that maybe Garrett and I are the only people who remember slide rules. I think they're in museums now.

There was a really cool gadget that allowed you to do a bunch of mathematical manipulations. Calculators were already there when Garrett and I were learning this, but we had some old school faculty in each of places. We learned where they wanted you to put everything in your bag and solve problems by hand.

Of course that made no sense because all of us were going to leave school and be in the workplace where we had opportunities to use all of these advanced calculators and whatnot. So we're at that place now.

So a few years ago in the classroom, what AI meant was how do I as a faculty Member make sure that my student didn't cheat and use AI to help write this paper to help solve this problem again, just like 20 years ago, 30 years ago, that was the wrong direction.

This is a national phenomenon where higher ed has had to quickly adapt and turn away from a notion of cheating. And let's catch you using AI to embracing AI in the classroom. Recognizing that those young people, the moment they leave our campus, are going to be in the workplace using artificial intelligence in their day to day jobs.

And we are doing a disservice to them if we don't fully prepare them for using AI, generative AI as a tool. And so that that pivot has taken place.

And so now it's a matter of okay, how do we equip the faculty with the knowledge and understanding to do that, to give the students entree into learning AI, incorporating it into their work, understanding the ethical and other complications, but also importantly, understanding so called hallucinations you might already know.

Hallucination is a colloquial term for when the AI algorithm returns something that is not true. Maybe one of the highest profile examples of this was when an attorney was caught submitting a brief in the court that contained case law that did not exist.

AI, the machine behind the AI system they were using, could not basically put 2 and 2 together to make four, so invented something. Obviously any user of generative AI needs to understand that. So that's an example of what we put to them. And our training of faculty continues.

We've had a number of different seminars led by administrators, led by other faculty, led by industry experts. We will soon be giving faculty, all faculty and staff access to online, online tutorials, online courses, to AI. I'm going to end with one really cool example of the power of this technology for our teachers.

So a few of the Members were with me when we did a tour of the Hawaii Space Flight Laboratory.

So that group and some others were asked to prepare an Executive course for folks from the continent who would come to Manoa and then also be online for something like a 12 week course on satellite design, launching the various uses of satellite technology, et cetera.

A couple of our, let's say more advanced faculty got together to develop a course. So what they did was they used generative AI. They took the descriptions of all of the courses at MNOA that could possibly be related to this topic from across the entire campus, hundreds of courses. They took examples of Executive courses at various universities.

They then prompted the AI system and also said, zero, and build in in person site visits and where we might go and where those might be placed in the, in the course. It generated an entire sequence of courses, syllabi, course descriptions, the order interspersed with in person tours.

They handed that to me, not telling me at all that AI was involved. I was blown away. This was in two weeks they had this. I thought it was the greatest thing ever. Then they told me it had been done by AI, but with a few tweaks. That is the course that we will be providing.

It's just an extraordinary tool. We probably would have had to pay a few faculty summer support to work on developing something like that. Just an indication of the power. Thank you.

Definitely. Thanks Chairs, Members of the Committee, thank you very much. I'll fill in some of the pedestrian stuff you heard about some really cool things that Gabe has been doing for a long period of time, as well as some of our various researchers. Our students.

Our students are really upfront in this stuff and it is all about providing them with the ability and the landscape to do it. And in some cases just setting them loose. So let me dial it back a little bit to the pedestrian stuff.

As Vasilis noted, one of the things that's very critical to us, I mean implementing technologies like AI revolve around data, data governance, privacy, information security. So that's in my suite of services that we provide for the institution.

Fortunately for us, as we started to see this tidal wave coming back in 2022 and AI has been around for literally 4050 years, but the acceleration and the ability to take it over the top is really the computing, the availability of computing, high powered computing, not cheap by the way, but high powered computing.

To take it over the top and to have the abilities of these consumerized applications to take hold and literally go wild.

But in that context, we have to make sure that the data that the AI engines get and the AI models get is properly vetted and fed to the engines, including the data that should not be fed and released into the engines, so that we can protect the data that's critical to the University, critical to our individuals that are Members of our community.

So privacy, data security, data governance is hugely important. And one of the main things that we've gone in early on, we've got a fairly sophisticated data governance, privacy and information security framework that was already in place before 2022. We've been working on this for years.

But what we were able to do is take that same governance framework, apply it against the AI tools that were being released, taking a look at the principles and the kinds of data that potentially could be dangerous to release into these models, dangerous from the standpoint of exposing private information.

As Vasilis noted, FERPA, HIPAA, we've got a bunch of regulations as education that we all have to comply with. And you know, it's really cool. You get this cool open AI thing. You have ChatGPT, it's free. Nothing's ever free. It's free.

And you just like, I'm just going to put in my tax returns from the past 10 years, I'm going to put in my intellectual property I've been working on for decades to try and create something cool and new and put it into those free services.

And by the way, you just posted it on a public bulletin board and these models will simply consume it, collect it. And by the way, the way the models are trained, you actually cannot erase it, you can't go backwards.

So it was critically important for us to make sure we establish that governance layer and that security layer from the front, provide guide rails, not restrictive guide rails, but guide rails, to make sure that we establish the right principles and practices for the members of our community, our faculty, our staff and our students, so that everyone is aware of what you should and shouldn't put in, what you can and absolutely cannot put in, depending on what position you hold within the University and what kind of data you have access to.

So it is really important that we establish that as the basis. And again, fortunately, we had a fairly robust structure to start with, but we're able to apply it to these AI tools as they broke onto the scene, again, consumerized tools, back in November 30, 2022 give or take a few days.

But this is really important to us to make sure that we took a hard look at it and made sure we had that base in place.

Fortunately, we took a look and also did not have to create a standalone specialized AI policy at that point in time because we had this fairly robust system of data governance, data privacy, information security already set. And the existing policies gave us sufficient guidelines as a starting point to be able to build that experience.

On top of, we also hire ED. Fantastic industry. We share everything across the industry. We share practices, we share examples. We've been able to work with our peers across the nation, in particular about 70 or 80 of the largest research one universities across the nation. We have my peers, the CIOs from those institutions.

We actually get together on a regular basis and talk about topics like this, in particular topics that may have substantial impact to the industry. And AI just literally popped as soon as ChatGPT popped on the scene.

Some of the early adopters, you may have heard University of Michigan, very early on, they wanted to be in front, they wanted to be first.

They put a pile of money in the pot to be first, did a lot of experimentation and learning out front, and even in that first year came back with a bunch of metrics that were astounding.

Increased student performance, increased faculty performance, saved time, time savings by faculty in terms of being able to do more with the folks that were there.

And then also turning the power of some of these engines and some of these tools loose in the hands of both faculty and students to see what they could produce that could be of value not just to the classes, but for the University as an industry as a whole.

They spent a lot of money and they had to create this walled garden where the data was protected because that wasn't absolutely important to them as it is to us. But we're able to learn from examples like this as we go forward.

Besides University of Michigan, University of Notre Dame, Vanderbilt University, University of Florida, UC San Diego, the some of these large institutions, they got resources, but large institutions that are able to help lead the way that we can work with together to bring this stuff forward.

A couple of examples in the fall, late fall of 2023, Notre Dame held an open workshop to build some of these capacities among and beyond and throughout higher ed. So we were able to participate in that as part of this effort of bringing the research universities together.

They held the second one a couple of months ago, so we were also able to participate in that. And in fact, one of our staff Members co led the workshop there, which is pretty cool. Next week, working with Cornell University, they're doing an AI hackathon in New York City.

We've got very cool stuff going on in this space, developing both capacity as well as product that potentially could be used industry wide. Let me come back to the University of Hawaii.

A couple of things that we're doing that we're already doing in that space include the deployment, like the Department of Education, the deployment of the Gemini tool in our Hawaii EDU domain.

We were able to, working with our partners at Google, we were able to add the appropriate licenses with the necessary data protections, but add those licenses to our existing workspace for Education plus enterprise license at no additional charge. Good price, right? Good price. Not anyway good price. They said nothing is free. Yeah, yeah.

The workspace for Education plus is not free anymore. So it used to be like 5 years ago not free anymore. But they added it on as value add. And one of the things that we're actually announcing, so spoiler alert, we have a AI tools webinar being run by some of my team.

Tomorrow we're announcing the release of the Google AI Essentials tool for our entire community. So it's an online training course, certification at the end delivered by Coursera, but we're hiding it inside the, inside our domain so that our folks can hit the page, hit the splash page, go ahead and sign up, take the course.

Supposed to be about eight or 10 hours. Pretty decent. I actually was one of the guinea pigs. I ran through it, I passed the test. I took it a couple of times. I took it a couple times. But very, very solid baseline course.

Of course it mentions some of the Google products, but it also mentions other products in the space.

Doesn't force you to use it if you don't want to, but gives you a good base of concepts, principles to dos, not to dos, examples of some of the tools and in particular as it goes through some of the ethics principles, human in the loop. Absolutely critically important because that's how we make this stuff work. The agentic.

So the movie scenarios, the doomsday movie scenarios, the power of the products are really there, but the ability to really get value out of it is you keep the human in the loop. So I'll stop there and maybe we can continue conversation.

Yeah, that sounds great. Thank you all for the wonderful presentations. Very thought provoking things you guys shared with us today. Chair Woodson, I know you have to leave real quick. Do you want to ask a quick question or two? All right. He stayed on as long as he could.

If we could ask Gabe to come back up to the table and then Heidi as well, then maybe uh, you guys want to do a three way junk in poll there and figure out who can fill in the third seat?

Okay, so Members will open it up to questions at this point. Don't all speak at once. Yes, Rep. La Chica. Go ahead, please.

Thank you so much, Chair, for putting this together. Question, Deputy. As mentioned by Vice President Garrett, on the date, the governance structure, data governance and security structure. Does the Department have somebody or a plan to have something similar? I know there's a new office. Right. Of information Technology. And like, how does that all play into the.

Yes, we do. And to explain it, I'm going to call Dr. Sakurai up. And that has been key to us and even any tools that we use. And I mentioned, I think I mentioned about any agreements that we have with those programs like the Magic School or conmigo.

We need to be clear that we know what they're doing with our data.

Collecting our data.

Absolutely. And then what we get in return.

Thank you for the question. Winston Sakurai. I'm in the curriculum office, but we've actually been working cross office with our Office of Technology and our Data Governance office. That's actually been our biggest conversation so far is protecting student data, protecting our employees data.

We've been doing everything possible to ensure those protections, including working with a lot of these corporations are distributing the AI tools to us. We have people on our staff that are actively looking and pinging the servers.

One of the examples we found is that they started moving some of the data offshore to countries that are potentially bad actors. And we actually had to stop the agreement with the organization because we were concerned that data could be seized and actually used against our employees and our students.

So that's the level of, I guess, caution and security we want to take to make sure that we protect all the people in the Departments.

How does that work? If you get some ping of like, potential data breach or data. Right. Like, how does. Cause the amount of data you could store is so huge. Yeah.

Right. We actually have very good people within the Tech Department that actually look at that on a daily basis. And so once they detect it, the first thing is, they do is, they either talk to us or they do, go directly to the source to say, you know, what's going on with where the data's heading to.

And again, we've had multiple conversations with this one organization, one corporation with that. And they said that that' that's their, that's their plan moving forward. And so we actually had to get discontinued services on that one product at this time.

When they, when they actually have other services that are going to be in areas that we feel are really going to be protected, then we can actually reengage conversations with them.

Thank you. And then final question is, have you started evaluating, you know, the growing technology of the, the chatbots? The particular.

So we have legislation where we're looking at just the regulation of these chat bots that can mimic human-like conversations or they use language learning to have actual conversations with us in, in voice form, in audio form, and people and kids trust them.

I listened to my son have a conversation with ChatGPT, telling him, telling ChatGPT tell me a story about, tell me a story about my dog, about a dog. And ChatGPT returned with the names of our three pets without him even prompting that information. That was just like, it's very personalized. Right.

How have you been able to look at some of that emerging kinds of technology?

Yeah, that's actually part of the training that we have.

When we talk about the ethical use of AI and making sure that as our teachers and our administrators are working with students, they tell them the dangers of it, but also the productive nature of it, the introvertedness of some students, they actually find some comfort in having those conversations.

But they can be something that can be either addictive, which we found in some studies, or also very harmful because the information that they give is so personal that it really creates a fantasy type of world that we need to actually help them recover from.

But that is part of the things that we talk about in our asynchronous courses and our in person training to make sure that people are aware of that, but also making sure that is used productively by students.

Thank you. Thank you, Chair.

Thank you, Rep. La Chica, Members. Rep. Kapela, go ahead please.

I have a couple of different questions. I'm not actually sure who the best person is to answer this question, but earlier it was mentioned that there's no. That nothing's free. Right.

So within talking about this ethical AI, is anyone talking about or researching the environmental impact of the use of AI, specifically in some of these, like, larger institutions at the University level, but also at the DOE level. How is that being monitored or measured? Is that being shared with students? It's a loaded question. Yeah.

Can you clarify the environment?

Well, AI uses a lot of energy, specifically a lot of water.

Oh. Okay, okay.

The University might be able to explain this more, but yes, the heat generated, the amount of power, the lack of silicon to actually do all these things are going to be very important to discuss. Maybe the University has a better answer.

But we've been talking about that as well because there is going to be environmental impact because of the power that AI uses.

Okay, thank you. Thank You.

Yeah, definitely. So that's actually on the international scene in terms of the impact of the data centers that run these things.

So the training models and where these queries actually get executed typically are in large data centers that are located, usually you'll locate them in areas where power is cheap or where there's the availability of land to build those data centers. So in particular, one of the challenges that we have here is power is expensive here.

So we never would build a large cluster here just because of the cost of running at scale. But from the standpoint of both national and international impacts, that is a concern in particular at the scale that some of these models run.

So you have very large, so you've heard in recent news, very large investments, dollar investments to build data centers to support AI and AI model generation and model training. So it's really important to make sure that the industry also looks to be more efficient when these chips get generated and designed.

But they get, they also get designed for more efficient power usage because you need the compute power to make this stuff run. And that's really the magic, the magic behind the scenes.

But in order to do it in a responsible fashion, especially when this scales up, the need for those chips to be very energy efficient is very high on the big picture things that we have to take responsibility for to make it happen. So nothing magic. There's a bunch of these large data centers all over the place now.

And the current models do require that power. Some of it's getting distributed. So the phones, everybody's carrying around lots of compute power in the phone. Not quite enough to run these models at scale, but enough to run the edge pieces. And then the big models have to be run in those large data centers.

So the necessity for the industry, especially the silicon industry, to push for energy efficiency is very high on the radar. Not just from a cost standpoint because it's more efficient economically, but also from an environmental standpoint to make sure we don't have that impact. It's a big deal though.

So are we just a follow up question on that, are we tracking like this usage?

I'm just concerned that maybe down the road as this really blows up, are these, some of these places that these things are being, that these data centers are being located in, could those people potentially charge us for the potential use or what we're drawing down from that location? I don't really know how this works.

So if you can explain that and if we are tracking that, is that something we need to be tracking?

So there's a bunch of industry folks that do track that on a regular basis. In particular because of the cost of energy. And like anything else that's mostly free at the consumer level, at some point we're paying.

So right now most of us are paying by the data that we volunteer into the applications and that the data is being farmed in some cases, pretty reasonable and useful for us.

But that's the place we have to be careful in particular on privacy and security to make sure that the data that gets used that way is mutually consented to for benefit. Right. So the Amazon shopping thing, perfect example. But the need to be aware.

Heidi and Winston and Gabe just said you have to be aware so that you know what you're volunteering into these applications and how you're using AI to make sure that it doesn't, like, abuse you back.

Yeah, thank you. I have some other questions here, but I can wait.

Okay, let's just see if anyone else has questions first. Rep. Kila, go ahead, please.

Thank you, Chair. Gabriel, I am thoroughly impressed. And I made a whisper over to Chair Woodson. Let me say this. Obviously, I think that's just a benefit of when you're putting $30,000 a year into your child's education and then you can do things like that, no fault of your own. And I just.

But obviously, as a state Legislator, our fiduciary duty is overseeing oversight of public funds and public entities like the state DOE. So I think you've shared how you folks have had collaborated. And can you just expand on what that has looked like with working with DOE?

Yeah, so. So all the examples that I showed you today of the. The student books and the stemworks. So stemworks is. So this is all work I do outside of my day job as a teacher. So this is stuff that I do where I just work directly with stemworks.

And they were like they were looking for someone who knew, who had tried these things. I had made all these mistakes before. I'd kind of figured out a good starting point for where educators could start playing with these tools at a very safe and entry level spot. Like these make your own book project.

So all the stories of like the children writing their own stories and all those were at school. So El Intermediate. We did Maui Way Now. We did. I went to do a tech talk at Baldwin High School.

So anyone who's just willing to just have me come and talk story with them, I've been sharing those more as like, personal hobby type thing. For some of them, it's like they're trying to do change management as well. Like, how do we do that?

And just bringing in another educator instead of like a consultant or like someone from like corporate to be able to say like, hey, I as a teacher have been playing with these tools. They're free $0. Try playing with them as well. Here's how I use them. Here's the guidance that the DOE put out.

Let's do something cool together. I think that's been the biggest partnership that we've been doing with everything. And I'd love to just keep doing that until everyone's got it figured out.

Well, then to your comment too, right? When you talked about the fourth grade student and in concept, they never had to actually rewrite something, but they had a visual form to see actually what they're writing to manifest it into that. But in that simple technique, you challenge them to then rewrite.

And those are the kind of things, when I think about AI, it can be super beneficial and helpful. And so I want to look at Heidi and I would love to invite you to the Nanakuli-Waianae Complex.

I presented at some of their district area PDs in the past. I would love to come back and do some.

Yeah, let me get you in front of the teachers. Yeah, look to our Leeward chapter facilitator as well. But then to that point, I think I would oppose this to DOE. Can you folks give me just an insight on. Obviously, a private school is much more flexibility on curriculum.

But if a teacher is trying to institute something like what you were saying, seventh grade AI media literacy. Are one, the teacher is given the flexibility, or two, is the DOE exploring some comprehension of this tool? Because ultimately too, when I think about AI, it's almost in concept.

If you can find a really good way to leverage it, it can actually be a tool of equity to the point of Iolani School and Nanakuli High School. Right. I might not have access to some of the things that are available at Iolani, but ChatGPT has no discrimination.

So it can be accessed in Nanakuli and Iolani in the same way. Does that make sense?

Absolutely. And if you break down some of the work from the samples that you saw, it's addressing the standards because we follow the standards. And so it's how do we use AI to make the learning fun and exciting and provide opportunities for students. Students to rewrite willingly. And that's part of our lift with our training. Right.

How do we allow AI to be a tool to enhance the learning, not replace the learning? So can it be done in Your Waianae Nanakuli. Absolutely. And we do have teachers who are doing that. We just need to continue to scale that.

So I think that in the question about curriculum, obviously, you folks obviously have to report to the board of education as well. But is DOE having these internal conversations? Because would media literacy around AI be constituted as a coursework or elective or is that a subcontext of probably basic English?

I think it can be both. It can be all. It just depends. We can integrate it into a course. We can have a. A separate course. It could be a credit course. It could be, you know, a half credit elective course. So I think there are opportunities.

As we get better at the implementation and as we get better at moving forward from the mistakes, then we'll be able to better design a course or implement within an existing course.

Okay, one more question. Sure. So in that context then, would you say this was ground zero for DOE policy curriculum on AI?

I think this is our. We need a foundation and we need guardrails and we need expectations.

Okay. So would you say this school year is year one with incorporation of AI or?

Yes, this is probably year two. When ChatGPT came out in kind of December of 2022, we actually started working with our board strategic plan to incorporate a lot of that and then started some of the pilots that year.

So we have some early adopters who are really gung-ho and we wanted to give them the ability to work on that. So yes, we've. This is probably year two and a half actually.

Great. And I think then to give context to my question, it allows us to gauge if we're going to consider this year two of adoption. It can help us as we continue the years and dialogues to see how we are progressing and where we can progress more. Because, Gabe, what year are you on at ChatGPT for Iolani?

2. 2 as well.

Okay.

Everyone's been kind of writing at the same time, then that is also did. A very similar year of grace. So all the teachers were able to use the tools, experiment, try things out, and then from that find the bright spots and to build a more system based approach.

Okay. So if we can mark this, then as year two, it gives me a good base on how to watch us develop. Yeah.

Yes. And Gabe is exceptional. And it's a compliment. I can match that exceptionality with some DOE teachers as well. Yeah.

Oh, I totally believe you. I totally believe you.

Okay. I just had to. I had to put in that plug.

Yeah, I believe you, Superintendent. Thank you, Chair.

Thank you folks for that line of questioning. I think, Rep. Kila, you touch on this issue of the digital divide. I think we saw that during COVID when a lot of the private schools went back to school weeks, if not months before the public schools. I think this kind of bared out the lack of equity in this.

So I think this does hold a lot of potential to address that and really make it borderless. If you will have a few questions for Gabe, if that's okay. I imagine the administrators at your school are very supportive in this sense.

But do you sense any resentment or concern from fellow teachers within the campus about what this might mean for the profession of education if teachers might one day get replaced by this technology?

So I've been talking to teachers from all across the state. Like, we've been just going and talking to everybody in general. There's a general sense of excitement for the, for the technology, but a need of support for, you know, I don't want to do something that's against what guidance or anything else says. Right.

So I'm looking for permission. So me, I can only speak for myself as an individual educator and consultant. But just being given that permission to experiment that first year gave me the wings to kind of do what all teachers like to do is teach. Right.

Once we're able to do that and we found out what worked, being able to share that with other teachers, in general, it's been excitement. I've never been to a campus where the faculty or staff are not happy that I'm there to talk about AI there.

And even when we did, tackling the really hard conversations about what is plagiarism like in this day and age, like, what are all of these things that we have to deal about? It's not from a perspective of like, let's hide this away so it doesn't affect what I do day to day.

It's always from a situation of, well, we got to figure this out, let's do it. So I haven't felt any of that animosity individually, personally, as an individual. Right.

Okay. And you mentioned plagiarism. Like, how do you deal with that in the classroom?

So right now, a lot of the work, again, speaking just for myself as an individual teacher, again with the year of grace and the year of what is the impact of the work that you're doing? Right. So I have. We adopted a whole level system from one to five, so there's an easy framework to communicate with kids.

Like, this is a level one assignment. No Internet. I just want to test your meat popsicle, right? All the way to level five, which is full AI.

I don't care if the whole essay and document and everything is written or what was the impact of the work that you did in your local community or what was your final product look like? Right, because we're seeing this in the workforce already.

It's not that AI is going to take your job, it's that the students who know AI are going to become the workers who will take your job.

Because the end product of what they can produce as a worker is going to be a much higher quality and everything when they're using all these tools versus if you were to like shelter away from them. And I think teachers really understand that.

How do you maintain teaching students how to think critically? Do you have a concern about an over reliance on AI where kids just automatically just default to, you know, looking up on ChatGPT or whatever?

In the very beginning I did have that like, anticipation of like, okay, well this is, I've seen what like social media has done to my generation. Like, is this going to do something similar to these students? But that group we put together, the Digital Literacy Ambassadors group.

So these are all students from all walks of life across campus. The input and the feedback they've been giving us is way more critical of these tools than even I've been seeing on the academic and adult level.

So for me, the critical thinking part is going to be as soon as they understand how these technologies work, it's going to be on par with having the same conversation, but swap the words out. AI with email.

Like, are you worried that email is just going to replace all their writing skills because they can just type in a computer versus write out the letter. We're anticipating similar transitions, just for me, and anticipating similar transitions as we adopt these new tools.

So they'll be using the critical thinking, but supported by all these agents that might be working for them. Like, oh, for this agent, create me a bunch of social media posts. You create me some letters of recommendation that I can start drafting and giving out to people. Here's my website. Go make my website.

That kind of stuff that they could do.

Okay, just one other question before we open it back up here, but I think you said you're teaching AI literacy in the middle school level. So what is the next phase when they get to high school? How do they move beyond literacy into preparing themselves for post high school life, for example?

So there's been amazing work by teachers from all across the state that I've been seeing on this and they've just been working with the students on individual projects.

So the same way you wouldn't take a typing class senior year, right. As they matriculate up to a system where I would anticipate they're gonna be once they have a good start. It's just gonna be as natural as, okay, everyone, take out your iPads, use the iPads.

It's gonna be same as like, take out your calculator, take out your AI, use that, put it away when you're done. That's kind of how I'm individually addressing all those. And with.

And because the students have that literacy base knowledge, when their advanced data sciences or their AI class that we've rolled out comes out, they already have that knowledge base in them. So there's not like a two week intro to AI in every class. And once they get into their older ages, they've already built into that capabilities.

Okay, thank you. Just stop there and see if Members have any other questions. Yes, Rep. Kapela.

Thank you, Gabe. First off, I just have to say it's so wonderful to see teachers on the stand, mainly because you're so animated and it's just really enjoyable. But I guess my question is how do we support women and specifically young girls who are-

Civil Beat did an article about how they're missing out on these, like, as you call them, the training for these high jobs that we're going to see that one day these students will have. How are you keeping girls in these classrooms?

How are we attracting young women to be able to take some of these to partake in these types of classes? What do we need to do in order to make sure that they stay there?

So one of the things that I really like about. They were talking about how their student was able to talk to these tools. Giving students the base literacy and how to engage with these tools makes it so they don't see AI as like a real friend that can talk to them every day.

But at the same time, no one in my family has graduated from college. I want that to be my dream. I want to talk to you every day. Until that's true is the most powerful mentor thing I can imagine.

And that little girl who did that rewrite, if she keeps working with these tools, that can be the entire life story right there. Same with the stories of the books that should be written of the people whose stories should be told.

That's work that we could kick out in two or three months, find the female leaders in our community, just record them for an afternoon, turn that into a table of contents book, go through the editor process, get those books published.

Now we have a whole section in our state library of the stories and examples for all these women to follow. And that I think would be a good starting point. And then just showing that that is being done by other people will give permission for people to mimic and copy. Right.

Like a kid sees us do cool stuff and then because I'm doing cool stuff, the kids want to do cool stuff. So just that kind of feedback loop with female leaders we have in our community, like you all on this board should get your life story published. Do it. It's a fun weekend project.

And then now you can share that with all, you know, all the young women that you meet and say, hey, here's how I did it, here's my life story. Get inspired and get involved your way as well.

I appreciate the homework.

Yes.

Sorry, it's like ingrained in me. I got to keep seventh graders entertained all day.

So you're doing a great job here.

Thank you.

The other question I think along that though is we talk so much about AI and these classes becoming very like computer heavy or video game heavy. And a lot of that sometimes really gears more towards young men.

And I think that sometimes that could be a weave out for young women who don't feel like maybe that's the avenue are there.

I guess my question really is how do we make the classes that are going to be created more, include more of these like life stories like you're talking about or something that's more exciting and broader than just video games being developed through AI?

Mm, I think so. One of the coolest projects I've been hearing about. And these are all other teachers, right? The Office of Civic Engagement. Right. Any sort of the one mile projects, projects where you get involved with the local community, AI is going to have way more of an impact in that. Right?

Like just watching the social media posts of like Ala Wai Canal and the cleanup and everything and then just seeing videos and contents of those people of the state workers going out and cleaning up everything to get it up.

Those kind of engagement projects or student projects like beach cleanups and stuff like that, the posters can be AI generated, right. All the proposals to their local governance to get engaged with stuff can be AI generated.

So as long as we meet the students in where they're interested, AI is one of the crazy tools where it's adaptable to so many different industries. It's not like, oh, it's just going to be in the video game class for boys. It can Just be anywhere.

So as long as there are teachers engaged with these tools and have the skills and literacy for it, that's not a conversation for me to be leading. That should be the female teachers saying, I know how to do this. This is what gets me excited.

This is how I can role model these engagements and these tools for people. And when they see their role models using those tools, they see you using these tools, they're going to feel that permission and go at it that way. Right.

How are you leaving space for those female teachers then? Because you're the person here at the table.

Ooh, for me here. Well, I will meet with literally anyone and talk to you about AI stuff. So that's the, that's the starting point that I can have.

I appreciate that. Thank you. Chair, if I may, one more question. I'm so sorry. For, for AS Armstrong or for DOE, how are you measuring the impact of social emotional learning that AI is having on our students?

I. At this point, I think we're too new into the game to measure the impact. I think the examples, the rare examples or there's positive examples too. For example, the chat box where constantly. Or that is one of the responsibilities of the teacher. And this is only when it's within the teacher's purview in the classroom.

We couldn't monitor at home, but what are the students doing with that chatbot? And is it appropriate or is it not? And then how do we intervene? And I think Dr. Sakurai gave an example for that. As far as statewide data on that, I can't give you statewide data, but at this point we can give individual scenarios.

And I think more are going to come out the more tools that are generated or the more opportunities that the kids have. But we are aware and it goes back to the whole use of technology. Right. It goes even prior to AI, we've seen the impact of cell phones or social media.

You know, we've seen the impact of sometimes the disengagement when the involvement is, a child is so involved with the computer losing that social interaction. So I think it fits in that realm.

But we are noticing some positives and we're noticing sometimes when we need to step in and intervene and how do we correct it so that the AI is not a detriment to the child, but rather an asset?

Thank you. Thank you, Chair.

Yes. Rep. Muraoka, go ahead, please.

Thank you, Chair. So I don't have a question as much as I have a comment for the DOE. Can you make it a point to hire 250 more Gabes? That Energy. That energy, the commitment. I mean, I. I can only wish I had a teacher that engaged. That fun when I was in school.

Yes.

I mean, that alone will keep the kids involved. Thank you, Gabe.

Just, I don't want our friends from the University to feel left out with all these questions. So just a few questions for them, being that I am the chair of the Higher Education Committee, maybe. Yeah. If you folks don't mind coming back up. And I'm sharing this from personal experience.

I have a daughter who's a senior in high school right now, just went through a very stressful time applying for colleges. Maybe. Provost, how's the University like, screening? Like, you know, a big part of the college application process is a personal statement.

How do you screen that statement to ensure that it was not like, AI assisted or generated and trying to just capture the authenticity of the students that are applying to the University.

Wow. That is a. That's. That's an email I'm going to send when we're finished with this.

And for the record, my daughter did not use ChatGPT for her application.

That's. That's important information. No, I do know that our admissions folks try to be in contact, personal contact with each and every applicant. So their review of the applicant and their preparedness for University is far more than the essay. And we've long done away with the standardized tests. They're welcome to submit them. They don't have to.

So we look at their performance in high school, and then there is always a conversation with each and every student. But, yeah, that is a good question. I don't know of any existing tools that could. You have them?

So under my office, we have research integrity. So it is monitored by the Federal Government. And we have tools, and there are several tools that actually can check proposals, papers, how they have been generated, whether they're being plagiarism in the traditional way or they have been plagiarism with respect to being ChatGPT, whatever tool you use.

So we pass that whenever we have an allegation. However, even that is going to change how you monitor research integrity. And actually, integrity of a product is going to change because what if I write a proposal which I have my first introduction. It is a literature review.

How important is whether that has been done by the researcher or by an AI tool as opposed to the body of the actual critical work? These things are new to us. I think they're new to any kind of office of research integrity. So these are really cool topics and difficult topics going forward, Chair.

Thank you. We do have a bill that we will hear in the Higher Education Committee regarding the AI Institute within the UH.

Can you give us a sneak preview as to your testimony, when we schedule that, how you see this AI Institute assisting the University in this space.

Yeah.

So the way we want to, the way we put that Bill together, it is it. It has couple of. For a couple of reasons. First is to actually formalize the AI Institute within the University. We want to actually co-fund that Institute.

We're going to invest some of our money but it always gives a lot, much more gravitas if it is a partnership between the University and the Legislature and the state. So we're going to put it under for now the office of the Vice President for research, my office.

We going to tailored around these five imperatives which I talked about and we're going to go out there and hire people to run those imperatives and it's going to be a little bit different. We, we think that AI now is driven not just only by traditional faculty, but by professionals and faculty practitioners.

So, Manoa and actually the University, we thought of faculty classifications. So now we have faculty of practice. So we're going to try to recruit some faculty of practice that they were in the industry in order to come in and actually educate us and faculty and the students. So that's the way we see going forward.

Okay, thank you. Members? I know we're coming up on the two hour mark so I want to thank you for everyone's patience so far. Final call for questions.

Good.

Okay. So with that Gabe, Heidi, thank you our friends from UH, thank you all for joining here today. I know this is just the beginning of a very important discussion and hopefully we can continue to have this as we proceed. So thank you very much. Thank you. Informational briefing is adjourned.

Sa.