Building My HI about AI: Part 1 with Eric Nentrup
Introduction to this Series
It seems like I can’t go a day without hearing about AI—artificial intelligence. It shows up daily in my education blogs and e-zines; in the general news; my friends and colleagues bring it up; and it’s in books, movies, and TV shows. When my friend and colleague, Dr. Caitlin Howley, asked me if I would write about AI, I had to pause. I have tinkered with some AI engines, but I readily admit that I am not an expert. Why should people believe me about AI? I’m a newbie just getting my toes wet. Luckily, I know several people who are not newbies, and in fact, can probably be called experts in a field where expertise is changing every day. They have what I will now call HI—human intelligence—about AI. So, this is the first in a series of blog posts about AI intended to help me—and hopefully others—build some HI about AI based on advice from different experts in the field.
While AI has seeped into almost every industry, this blog series is geared toward AI in education. All three experts I interviewed are current educators or education consultants with much greater experience and knowledge about AI than me. I already had respect and appreciation for their work, and it was fun for me to learn some more about this rapidly evolving topic from people I trust. I hope you find it helpful as well.
An Expert at the National Level
I started my HI journey with Eric Nentrup. I know Eric for his filmmaking skills as we worked together on a project in a large urban school district on the East Coast this past year. We got to visit some great teachers in their classrooms; I interviewed them and Eric created fantastic videos. Eric calls himself a storyteller, and he truly is. His films prove that. A former English teacher and administrator, Eric also goes by the more formal titles of educational consultant and writer/producer.
Many people in the education sector know Eric through his work following his teaching career, working broadly in educational technology (EdTech), and more recently as an education consultant focused on AI’s impact on the profession. He is part of the team that wrote Artificial Intelligence and the Future of Teaching and Learning: Insights and Recommendations for the Office of Educational Technology (OET) at the U.S. Department of Education. This team is currently working on additional guidance documents for the OET, so be on the lookout for them.
AI is Not Just ChatGPT
One of the first things Eric cleared up for me is that the now familiar AI engines that some of us—along with our students—have found and use are not all there is to AI. Eric notes that these tools, such as ChatGPT, Pi, Claude, and others rely on AI components and give us a way to interface with it. As Eric points out, when we’re working with ChatGPT, “We’re experiencing the superficial layer—and often not aware of the large language model beneath it and how it came to be.” There’s much more to AI than a search tool.
Eric shared that when people refer to “AI” these days, they are usually talking about generative AI (sometimes called GenAI), tools that can create—or generate—text, images, audio, and other content. In fact, while writing this piece, OpenAI introduced their next product, Sora, which can accept text prompts and turn them into realistic videos that are just as—if not more—stunning than what other recent tools have produced. GenAI has been made possible by large language models (LLMs) that are “trained” with an inconceivable amount of information so that they can then respond to your prompt, craft a picture, or create that video for you. He describes AI as the “big umbrella,” and underneath that umbrella are multiple disciplines, which are complex in their own right. These include machine learning, natural language processing, computer vision, deep learning, neural networks, and more—far too many things for an AI newbie to explore.
The takeaway for me, however, is that these components of AI are everywhere, not just in education. There are a lot of moving pieces that are changing every industry under the sun. AI is changing medicine, transportation and shipping, marketing and advertising, coding, the arts, and even political discourse. It’s impacting the tools, processes, and information we use every day.
Eric makes a great distinction between fads, trends, and paradigm shifts. He said, “Ignore the fads. Trends are where we do most of our work. Paradigm shifts do most of their work on us.” Generative AI is the latest paradigm shift, but he notes, “You don’t have to be full-time into AI. You’re already benefiting from it. You’re already using it. You’ve already been exploited by it.”
Ack! Exploited? What am I supposed to do about that? Eric’s advice: “Every time you’ve been surprised by an astute music or product recommendation from your favorite platforms is an example that your contributions to training Apple or Amazon or some other organization’s algorithms have made that technology more useful to you—and conversely, more apt to give those companies your money! So, stay curious. Don’t be afraid. Don’t dig in your heels. It’s better to learn to swim, even if it’s in the shallow end.”
Eric referenced a helpful metaphor he used in the report for the OET. “We should think about AI in education like we have other automation examples. As teachers, we can harness AI-enabled tools that are more like an electric bike and less like a Roomba. One amplifies the efforts you put into your craft, while the other is a delegate for your most menial tasks. Though the latter is helpful, the former is more exciting to ponder.”
Educators Aren’t Looking at the Right Thing
Eric is worried that the initial knee-jerk reaction to AI by educators is drawing away focus and energy about the potential for AI to support and even improve teaching and learning. “Initially, people were not paying attention to the right thing. They were obsessed with trying to prevent things like plagiarism instead of seeing this as a chance to usher in teaching and learning strategies we’ve desired for decades,” he said. Another early reaction to ChatGPT as an example of mainstream AI that could affect education was the fear of replacing teachers. This is an ongoing complaint about new technologies that was levied against technologies such as radio, television, and, of course, personal computers. Eric notes, “We can acknowledge such fears or frustrations without promoting them as foregone conclusions. The tasks and the job are going to change just like they always have. This is a reminder that as educators, we are purveyors of change.”
Together we lamented this same cycle that, historically, has caused people to react defensively and try going the route of banishing the new technology rather than exploring its potential. I shared my experience of creating a performance assessment for whether schools should or shouldn’t allow cell phones in schools—in 2010(!)—and the issue is still alive today. More and more school districts and even some states are exploring banning a part of their students’ daily lives that they have grown up with and have always known and used in every aspect of their life … except in school. Can we not do that with AI, please?
Eric had a great suggestion to shift the focus from obsessing over plagiarism and instead leveraging AI to promote teaching and learning—for all students. He suggests that educators “should instead invest that energy into redesigning their assessments.” Remember when we used to say, “If you can Google it it’s not a great assessment?” AI has upped the ante on that one. Eric notes that we know a lot about how to design more relevant instruction and assessments. We have proven design practices such as backward design. We just don’t always follow through with best practice. Now’s our chance.
I’m all for it. I taught music. I never gave my students a multiple-choice test. They always had to perform, at least something. Teachers in the arts and career and technical education understand this. The best way to show you can wire a house is to wire a house! I carried this practice over to the graduate students I taught in an educational technology class. I told them the class was based on the philosophy, “You won’t be asked to fill in a bubble sheet to teach a kid. So multiple-choice assessments aren’t appropriate in my class. Instead, you have to show me what you know.” Maybe AI will help us finally get to the point where assessments become relevant teaching tools rather than an audit of low-level knowledge and skills.
Promoting Equity with AI
I mentioned that AI is not just an education sector thing. In fact, education is probably one of the last sectors that everyone in the generative AI disciplines is worried about. AI has the potential to change fundamental business models. It has the potential to eliminate formerly lucrative careers and spawn new ones. That all means money—so people are moving full steam ahead to try to figure this out! Educators and their AI considerations may be on the back burner.
That’s an issue for educators like us. Because these LLMs and the other resources being developed and honed are not necessarily being developed with all audiences and populations in mind. A foundational pillar of our education systems is the concept of equity—equitable access to educational experiences and resources for all students. Eric warns that equity was not always at the forefront for everyone who generated these LLMs, and educators need to be aware of this when incorporating AI into teaching and learning as well as other areas such as discipline and even surveillance. “Employing emerging technologies with good intentions doesn’t equate to protecting a student’s civil rights. I’ve learned to consider impact over intentions and it directly applies to AI in education,” he said.
There are many people who raise the concern of bias in AI. A quick internet search about bias in AI finds dozens of scholarly articles and reports from respected organizations. They note that these engines “learn” by using data and relying on algorithms to make decisions about that data. If the data used wasn’t representative of multiple viewpoints or populations, the system generates biased output. Bias and stereotypes can also be perpetuated through the way AI has been coded and the algorithms it is programmed to use, whether those doing the coding realize it or not. Humans can also perpetuate bias and stereotypes when they take this content forward in their thoughts, products, and actions. So, what do we do about that?
Eric suggests that reducing bias has to start with the vendors. Educators need to be savvy when talking with vendors about where the data comes from in the AI products they use. Does it take multiple viewpoints into consideration? Does it take into consideration underrepresented populations? And then how is the new data that users generate being used? Educators should require significant transparency about where the data in AI tools comes from and how it is used.
How Educators Can Get Started with AI
While educators may not have much sway over determining how GenAI is trained, especially with an eye toward equitable representation, there are things Eric suggests educators at different levels can do to explore AI and to begin to harness it in their own work. I asked him to break it down into three levels: state, district, and school/classroom.
State Level
Eric said that at all three levels, there is already a growing body of resource materials about AI. In addition to the report Eric contributed to, the new National Educational Technology Plan has also been released. It’s the first revision in seven years and truly necessary because of the rapid changes in technology—not just technology’s capability but also how the way we use technology continues to change. Eric referred to the new plan as “your broad syllabus” and it includes information from the AI report he helped produce.
Eric noted that AI and the Future of Teaching and Learning is just the start of guidance from the OET. There’s a developer's guide available and a toolkit forthcoming. And there are education-friendly organizations providing advice to educators at all levels. He suggests paying attention to the work being conducted by EDSAFE, which is focused on the issue of equity and supporting research on AI to generate greater trust in its use in education. He’s also a fan of educators like Amanda Bickerstaff and her AI for Education work. As always, it’s good to find some experts you trust so you can keep your thumb on the pulse of what’s happening.
District Level
In addition to keeping informed about the development of AI through trusted organizations and experts, Eric suggests district leaders focus specifically on what they need to know to employ AI—or any new or emerging technology—to become more effective and efficient leaders. He stresses that AI should promote teaching and learning, and district leaders have to build a background to understand how AI can support and even improve teaching and learning.
Eric recommends district leaders ask themselves three questions:
- Are we doing everything we can to empower teaching and learning to increase the academic offerings we have, regardless of age, grade, or learner pathway?
- Can we find evidence-based solutions that provide finer offerings for all interest-based aspirations?
- Can we personalize learning to the extent possible without increasing the burden on the teacher or student?
Eric warns, “Don’t let the tail wag the dog!” Focus on teaching and learning, and doing so in a way that keeps everyone and their information safe.
School and Classroom Level
Reflecting on his early classroom days teaching at Indianapolis Metropolitan High School in Indiana, Eric recalled an internship program with a local IT company, the Kinney Group. One statement still resonates with him from the group’s leader, Jim Kinney, who said, “If you automate the mundane, you liberate the people.” Eric notes that’s a practical place to start for educators who want to begin using AI in their daily work: delegate what you safely can, and recover instructional time and relational growth with students.
Eric said that if he could “automate the mundane for himself,” he’d give himself the liberty to reallocate those energies and resources elsewhere—whether that’s time, energy, or money. Classroom teachers can find AI resources to do that in their daily practice right now. One example Eric shared was simply making information available to a young person, which is the bulk of what most teachers do every day. Eric notes that for some time, we have not needed to be information disseminators. Many teachers do enjoy delivering content directly, but there are now multiple resources that can do that for us, especially through GenAI. Although, he does not discount the “value of an experienced human sharing what they know.” Like Linda Darling-Hammond and her colleagues at Stanford (2014), Eric notes the most critical element in a technology-enhanced learning environment is still the teacher.
But what if you’re completely new to AI? Maybe you even have your doubts or concerns about it? Eric says a good place to start is to play around with some of the apps built on top of the most popular LLMs. Many of us may have heard about ChatGPT, but it’s only one resource. He suggests starting with ChatGPT, but then you should also try out some others. If you’re “risk adverse or feel intimidated” by the buzz you’re hearing about AI, Eric recommends trying one that’s “a little bit more warm.” Try Pi, which he says makes an excellent thought partner and sounding board to generate momentum on a task.
“And if that was fun, get in the middle and try Google Gemini or Anthropic’s Claude,” Eric said.
He suggests newbies, like me, should work on building a baseline, rudimentary understanding of what interacting and playing with these LLMs is like—and that means experimenting with several in a safe, personal way.
He had one last suggestion that is especially salient for those who might be a bit reluctant to explore new technologies. “Ask yourself, are you a more hands-on heuristic learner, or are you happier being a spectator?” If you’re a spectator, find peers you can work with and “look over their shoulder to see how they approach this stuff.” Eric notes it has been most informative for him to watch others write prompts to get different results from LLMs and see how they might apply to his own work. See, even with his experience and expertise, Eric is still learning with others. We can all do that.
Next Up: Maria Stavropoulus, then John Daniels
In the next post in this series, I get to share advice from another good friend and colleague—an EdTech director in a school district. Maria Stavropoulus not only helps teachers in her district explore the uses of AI but also collaborates with other EdTech directors across the Midwest through work with the Consortium for School Networking. After that, I’ll talk to a new friend, EdTech Coach John Daniels, who works with teachers daily at his elementary school in Carteret County, North Carolina, and is exploring the value of AI with teachers in his school.
Want to know more about Eric Nentrup? Eric is an independent education consultant and writer/producer specializing in emerging EdTech at the intersection of policy and practice. You can find Eric on LinkedIn or through Advanced Learning Partnerships.
Build Your Own HI about AI by Exploring these Resources Eric Mentions
Please note: not all resources are free
Artificial Intelligence and the Future of Teaching and Learning: Insights and Recommendations f and the National Educational Technology Plan from the Office of Educational Technology at the U.S. Department of Education
Using Technology to Support At-Risk Students’ Learning (2014) by Linda Darling-Hammond, Molly B. Zielezinski, and Shelley Goldman
EDSAFE focuses on equitable outcomes for all learners
AI for Education from Amanda Bickerstaff
ChatGPT (if you can, check out different versions to see what they offer)
Pi (also comes as an app for your mobile device)
Anthropic’s Claude
Create video from text with Sora
About the Author
Dr. John Ross has spent more than three decades in education, and even though he taught music for one of them, he has always been interested in technology. He wrote his Master’s Thesis on evaluating the then-new educational software entering many classrooms on 5-1/4 floppy disks using a just-released Mac Classic, so you can figure out when that was. He’s been investigating and using technology since then. He is co-author of the first and only textbook on the ISTE (International Society for Technology in Education) Standards for Educators and published a best-selling book about Online Professional Development for Corwin. He is the former director of the federally funded Institute for Advancement of Emerging Technologies in Education and has worked with multiple state departments and dozens of districts across the country on topics such as strategic planning for and integrating just about every new educational technology that has entered schools in the past 25 years, instructional design and authentic assessment development, coaching teachers and leaders, blended and online learning for learners of all ages. Most recently he has been helping schools, districts, and states across the country develop student Tech Teams that provide IT support, professional learning opportunities, and community outreach while earning course credits and industry certifications. You can find out more about him on his website TeachLearnTech.com or connect on LinkedIn.