Building My HI about AI: Part 2 with Maria Stavropoulus

Working at the District Level

In the first part of this series on building my HI (human intelligence) about AI (artificial intelligence), I interviewed a friend who is having an impact at the national level. Eric Nentrup is an independent education consultant and writer/producer who is part of the team that crafted Artificial Intelligence and the Future of Teaching and Learning: Insights and Recommendations for the Office of Educational Technology at the U.S. Department of Education. I next turned to someone I know who has been working in a school district that has found support from national organizations.

I’ve been collaborating with my friend and colleague Maria Stavropoulus for a few years on a project for the North Carolina Business Committee for Education (NCBCE), located in the state’s governor’s office. We’ve helped the NCBCE establish a network of student technology teams across the state and have provided guidance, coaching, and some curriculum development. I’m fine with the curriculum stuff—even for topics new to me like Information Technology Fundamentals—but when it starts getting highly technical I’m lucky to have Maria on my side. She has a whole lot of HI in a lot of areas and is a great thinker. She’s always able to come up with a different way to see an issue or a task.

Maria is an educational technology (EdTech) leader in Township High School District 214 in the Chicago area. She brings a range of knowledge and expertise from both the EdTech and IT sides of the house, and for much of the time I’ve worked with her she’s been sharing her growing interest and activity in AI—even giving me nudges now and then. Maria is a former board member of Illinois Educational Technology Leaders (IETL). IETL is the state chapter of Consortium for School Networking (CoSN), and a professional association for K–12 EdTech leaders. She recommends the toolkit and guide that CoSN co-authored as a valuable reference for having conversations with educators in her district about AI.

It was fun to catch up with Maria and share my first steps with AI in a way many first-timers do—by giving ChatGPT a whirl. I told her that my experiences weren’t as fruitful as I had hoped because I was specifically working on learning activities for kindergarten and first-grade students. ChatGPT knew there were content standards, but it had a hard time describing activities that aligned with them or were at an appropriate developmental level. It also struggled to reduce the reading level of some passages I wrote to be appropriate for 5- and 6-year-olds—something I heard AI could do.

“Were you using the free version?” Maria asked. As a newbie, of course, my answer was, “Yes.” “The free version is based on older data and algorithms,” Maria shared. “You’ll get different results with different tools.” Lesson learned. It also echoes advice from my colleague Eric Nentrup: Use different tools because you’ll get different results. I’ll remember that going forward.

Districts Write Procedures Based on State policies, but Not Too Many States Have Created Policies to Support/Monitor AI

I know about Maria’s work with CoSN in her state. I know she has an understanding of issues related to AI in this broader context as well as at a more micro level since she works with teachers and others in her district. She’s also teaching a course for Quincy University called Exploring AI and Understanding Your Human Potential. I began by asking her what some of the best advice is she has for state leaders about AI.

“Policies are critical, because districts look to the state for policies so they can draft procedures for implementing new resources like AI,” she said. Her concern, however, is that not many states have crafted policies to govern AI or to guide educators in its use. Based on an upcoming presentation she was working on about AI for local educators, as of September 2023 only two states provide guidance on AI—and only 11 were developing guidance. AI is moving quickly in every business sector, few states have taken steps to consider guidance for education.

State AI Guidance

In just a couple of weeks since talking with Maria and as of this publication, we’ve now found AI guidance from seven states, so state guidance is growing.

California: Learning With AI, Learning About AI
North Carolina: North Carolina Generative AI Implementation Recommendations and Considerations for PK-13 Public Schools
Ohio: AI Toolkit
Oregon: Generative Artificial Intelligence (AI) in K-12 Classrooms
Virginia: Guidelines for Generative AI Integration Throughout Education in the Commonwealth of Virginia
Washington: Human-Centered AI. Guidance for K-12 Public Schools
West Virginia: Guidance, Considerations & Intentions for the Use of Artificial Intelligence in West Virginia Schools

There are educators across the country who rely on state departments of education for guidance on issues like this, just as they have for past technologies, such as the internet and cell phones. In the meantime, school and district leaders are turning to a hodgepodge of resources, including vendors and even the press, to try to figure out how to implement AI in their classrooms. Maria’s recommendation, if you’re in this boat, is don’t wait for the national release of any overarching policy. Instead, first go to your acceptable/responsible use policy and see how it might or might not support AI and work from there.

A key consideration is data privacy, which more states do have laws about. AI is in your schools, whether you have procedures or not. How do you know data and information from your staff and students are being kept secure as they’re using AI? How do you know your staff and students are managing to keep data secure? Do they even know? How do you know a vendor that incorporates AI bots or widgets into other products is keeping that data secure?

She notes that all too often IT departments get requests from teachers and are pressured to purchase resources without adequate scrutiny. Sometimes these things are purchased at the building or classroom level without IT being involved. That throws out a lot more questions for Maria like, “Can it be supported in your technology environment? Will it run on the devices and network resources? Who’s going to provide technical support for it? With equity as a concern, will the resource work for every student? Will students who have an Individualized Education Program or other accommodations be able to use it?” Yeah, a lot of questions.

Of primary concern is what does it do with the data that students, teachers, and other staff are feeding into it? School districts that have already investigated technology and curriculum purchases through the lens of data privacy and security may have a leg up on others, but now AI has to be woven into those purchasing decisions.

Another recommendation is to go to your vendors right now and ask them what they’re currently doing about AI. Have they started to infuse it? Are they thinking about it? Maria talks to vendors for her district and she’s had responses from “It’s in the pipeline” to one that has “21 separate AI resources running in the background.” Twenty-one! Others are slowly releasing it—by tomorrow or by the end of the school year. It’s likely your students will soon have an AI chatbot as a tutor that “lives” within a learning management system (LMS) you’ve already adopted and it’s just going to be added as a feature without your district having the opportunity to go back and scrutinize the whole service. Or teachers will have AI lesson plan support within or outside of the tools they currently use. “The train,” she notes, “has already left the station.”

From an operational standpoint, it’s critical you have these conversations with your vendors. How do the vendors talk to each other? How do the tools talk to each other? What data are they sharing? What assurances do you have that vendors are keeping data secure? If they are relying on a third party to provide some of these tools and resources, how are they keeping on top of them to ensure secure data? She notes that you can build on experiences from past curriculum adoptions. Ask similar types of questions. It just looks different, because it’s AI.

I asked if Maria had a good checklist or rubric to support decision-making when reviewing AI resources. She couldn’t think of one right off the top of her head, so she generated one to help you get started. Thanks, Maria!

Table
Table

District Leaders Should Partner with ALL of Their Departments, Not Just IT

Once again I learned that vendors play a crucial role in the development and deployment of AI resources. Education leaders can talk to vendors, but they can’t always control the actions vendors take or the decisions vendors make. So I asked Maria for advice on what district leaders can do in their district with the people they can influence.

Maria didn’t hesitate a second and noted that district leaders need to partner with their departments—all departments—to consider how AI might impact their work. It could be teaching and learning, transportation and maintenance, or the business department. AI is not just the purview of your technology department. She said that all of the district’s departments need to be having “the AI conversation.” That’s because it’s already impacting your systems—whether you have considered it or not.

Like me, you may be more familiar with the AI in our daily lives—Google Maps plots your route home, spell-checkers clean up your emails, word-prediction suggestions speed up your text messages, or maybe your rely on Siri or Alexa to tell you the weather or start your favorite playlist. The data we have been feeding into those systems has been training and customizing them and the way they interact with us for years. Maria said, “We all have that responsibility to start understanding what it means for us personally, as individuals first, and then consider how that translates into our work.” She suggested expanding those considerations to your department, your team, and even globally for people like you. It’s all moving so quickly that she encouraged due diligence to stay informed and keep up with it.

While all departments should be having “the AI conversation,” it is vital to keep technology departments looped in. They should be at the table, for example, when doing curriculum adoptions through teaching and learning, because they’re likely to include technology-based resources that contain AI moving forward. Take the popular Khan Academy that just released its AI tutor Khanmigo. And if the technology department can’t always be at the table, they might be able to check in at critical junctures and provide guidance and tools to support decision-making. Working in a technology department, Maria jokingly noted that technology teams are often viewed as the “No, nope, just say no” team. She says they don’t try to be that way. Instead, they need other departments to understand privacy and security issues and that the entire system is focused on protecting the PII—or personally identifiable information—of students and others.

“Sometimes I think the smaller districts do a better job at it if they have enough leadership that is proactive versus reactive,” she said. Larger districts can have more turnover and sometimes, she observed, things slip through the cracks. These issues can be swept away due to the “It’s not my job. It’s somebody else’s job” syndrome. She suggests this issue is likely at all of our jobs but we can all be diligent. “Use your resources. Find things in supporting documents. Identify and question things. Talk to your peers.” These are all ways we can be diligent about keeping on top of AI and how it’s being used.

If you’re in a district where you’re not exposing your staff to AI, Maria recommends you start at the top. “You can either partner with external consulting firms for structured AI training tailored to your district’s needs or leverage your internal talent to create a customized AI professional development program. Both approaches offer unique advantages, whether it’s accessing specialized expertise or fostering a culture of continuous learning and adaptability within your team.” But they have to have some understanding of AI first.

Focus on Instruction

Of course, your teaching and learning staff need to understand how their practice might have to change. For example, how do they shift the discussions they have with students? With other teachers? With parents? They need to understand how the tools and resources they are using might be incorporating AI. She notes that if you haven’t been focused on personalized learning, now more than ever you will be personalizing learning.

Personalized learning and the resources that support it have pushed data privacy laws and policies. AI ramps up the importance of following procedures to keep data safe and secure. AI is going to be embedded in all kinds of resources teachers use—their LMS, search engines, record-keeping tools, and even productivity suites like Google Workspace and Microsoft Office. If a student or teacher is using one of these resources and entering pirvate information or student identifiers, according to Maria, “we’ve got issues.” For teachers, that means what you have in your online storage, the information in your emails—perhaps information about students—all has to be kept secure.

Truly personalizing learning can yield benefits, though. Schools and districts that do it well have already seen those benefits. Just like when I was talking with Eric[DBH3] , we came across the “If you can Google it …” question. If you can, it’s not a great assessment. Now with AI, if you can get AI to answer it, consider turning it around. Asking students to write a definition of new vocabulary and use it in a sentence, write extemporaneously about a topic, or even complete a cookie-cutter science lab report are all tasks GenAI can do for them in a matter of seconds. But teachers may be surprised by the suggestions they might receive asking a GenAI tool to generate classroom activities that could replace these activities. Vocabulary practice could become collaborative storytelling. Students older than 13 might use AI to brainstorm ideas and create an outline for a writing assignment that includes peer review all supported on a collaborative document. Students might create their own labs to demonstrate their understanding of a topic. None of these ideas will be a stretch to most teachers and may just be reminders of more personalized learning opportunities.

When we truly personalize learning, we can move beyond auditing student recall and foundational skills. Maria is excited about the ability for AI to help generate opportunities to help students develop their critical thinking skills. It’s going to be necessary in some fields, as technology is not just replacing labor-intensive jobs but is supplementing or supplanting decision-making positions and creating the need for differently-skilled workers. Quoting The Future of Jobs Report 2023 from the World Economic Forum, Maria said many workers “will need to upskill and reskill to remain competitive” and there is a growing need for “soft skills, such as creativity, problem-solving and critical thinking.”

Teachers are often called upon to be creative problem solvers and to think critically. That’s because we work with humans—a wide variety of humans—and all their complexities. As Eric did, Maria sees the potential for AI to support education by completing some of the administrative tasks people don’t really need to do or by using it to assist in these tasks so we’re more efficient and effective. That leaves teachers to do what they do best, teaching those humans—but doing so with an assistant that’s going to take care of the little things that could otherwise sap your time.

Getting Started as a Teacher

Where to start? What if you are a teacher and already have all these other things on your plate and it seems like there just aren’t enough hours in the day? Maria echoes advice I received from Eric: “Jump in and try a few things, but use it for personal reasons first. If you’re comfortable with that, consider how those resources can help you with your instruction. Keep talking to it. Keep prompting it.” Maria said, “I love that it doesn’t get mad at me. I rephrase the question when I don’t get what I want, but that helps me get better at prompting.”

Prompting is an important skill in using AI. It’s strategic. In some tools, like Chat GPT, the user has to create the prompt. But there are tools like Magic School AI where it provides a template so users can build better prompts. It may ask, “What grade? What’s your goal? What are you looking for?” She compares that process to building a website from scratch versus using plug-and-play elements you can drop into a website template. Some people will need the template but others will not.

There will be users at different levels—from those who “geek out about AI” and want to have full control to those who will benefit from a template. Maria says the templates are a great way to get some people started. “Not everybody has to be a coder, or a developer, or a prompter of AI. They can use the supports that are built into some of the tools, whether it’s an adult, a child, or somebody in between.”

The course she’s teaching for Quincy University has 15 teachers and counselors exploring and using AI in their own practice. They recently completed an assignment anyone can do and it would be a good group project for any district or school trying to learn more about AI. She asked them to review all of the curricula that is being used in their line of work and all of the tools being used with their students. She told them to look for AI infused into tools. She challenged them to find AI elements and determine how it was being used right now. She also challenged them to think about how it might it be used. She asked them to evaluate how it might be used more effectively and efficiently. She encouraged her teachers and counselors to talk to their peers and also to students to see how they might be using any of these resources. “It’s important to talk to the students, because sometimes teachers assign things and don’t realize there are more ‘bells and whistles’ on there, but the students always find everything.”

She urges her teachers and counselors to build a more global perspective on AI and its uses. If they haven’t yet capitalized on these tools, she encourages them to go further to find out what is available. They are encouraged to start with something relevant in their practice and then go a step further with their students and create some acceptable classroom procedures that incorporate AI. By doing that, they are generating samples that can later be shared in their own schools and districts.

She shared one last activity that is not only fun but actually accentuates the idea that AI is more than just a text generator. There’s a lot of focus on plagiarized text right now, as my friend Eric noted, but AI can create a range of content in different media formats that can be used to explore, encourage, or reflect on new ideas. Maria had her students use a prompt to create an AI-generated photo. Using Adobe FireflyCanva, or Microsoft Bing and the prompt “A (color) (animal) eating (food) in (destination)” her students could generate images of a purple goat eating pizza in Rome or a pink giraffe eating macarons in Paris. She shared one of the images.

Make a Fake graphic using a prompt

It was great to have a fun ending to our conversation and to understand that not all is doom and gloom in AI land. I told her my next step was to dump our recorded chat into Otter AI so it could generate a transcript. “I’m so proud of you,” she beamed. “This is how it starts for you.” And, you know, I guess it is.

John Daniels and Eric Nentrup

This is the second in a series of three interviews on my path to learning more about AI. Next I will share what I learned from John Daniels, an EdTech coach at Newport Elementary School in Carteret County, NC. I have been working with the EdTech coaches this year and AI is a popular area of interest amongst them. John Daniels has taken it a step further and is presenting at the North Carolina Technology in Society (NC TIES) conference on how he uses AI with teachers in his school.

If you missed the first one, this conversation started  with an interview with my friend and colleague Eric Nentrup, who is one of the authors of Artificial Intelligence and the Future of Teaching and Learning: Insights and Recommendations and has more publications forthcoming.

Discover more about Maria Stavropoulos, an EdTech Specialist dedicated to enhancing education through technology in the Chicago area. Beyond her daily role, Maria is an active participant in the tech community, often sharing her insights at esteemed conferences like IDEAcon and CoSN. Connect with Maria on LinkedIn or via Advanced Learning Partnerships.

Build Your Own HI about AI by Exploring these Resources Maria Mentions

Please note: not all resources are free

Artificial Intelligence and the Future of Teaching and Learning: Insights and Recommendations f from the Office of Educational Technology at the U.S. Department of Education

The Future of Jobs Report 2023 from the World Economic Forum

Consortium for School Networking toolkit and guide

Illinois Educational Technology Leaders (IETL)

Khan Academy’s AI tutor Khanmigo

Magic School AI provides templates for prompting

Generate images with Adobe FireflyCanva, or Microsoft Bing[CH5]