By Matt Renwick
In a text message exchange, my daughter referred to me as “DadGPT” after I responded to her question about her computer. What I had sent her was my original thinking (I swear!). However, I’ve referenced AI enough in my offers to help with her pre-calculus homework to arouse her suspicion.
I am a frequent chatbot user, but I’m not sure if joining the chorus of other non-experts offering their opinion on AI will be helpful here. That said, I do have a few thoughts as this technology becomes commonplace in our society and in schools.
AI is here to stay, whether we like it or not.
Concerns about AI use are valid. In her book Empire of AI: Dreams and Nightmares in Sam Altman’s OpenAI, journalist Karen Hao discovered that “under the hood, generative AI models are monstrosities, built from consuming previously unfathomable amounts of data, labor, computing power, and natural resources.” For example, underpaid workers in Kenya were hired by OpenAI to flag explicit and inappropriate images generated by users. Access to the constant barrage of harmful visuals led to an increase in mental health problems for the workers.
It’s clear that whether or not to use AI goes beyond a technical decision. These choices are made within the context of our values. And yet, in the near future, so many of our jobs will come with an expectation to use it, even if we do not already. How do we balance this reality with our personal beliefs and values?
What’s helped me is understanding our individual impact. How much electricity and other forms of energy are used interacting with it? A technology expert I consulted informed me that ten prompts (our queries we submit to AI) consume about the same amount of energy as watching one hour of a show on Netflix. I also learned that the majority of energy expended goes into training AI versus using it. It’s this kind of information that can help us and help our students make informed decisions.
TRY IT: If you have never used AI, put a prompt into a chatbot. Ask a question you have been pondering, one that is important to you. As an example, my wife asked ChatGPT what plants belong in an apothecary garden (she is making healing oils). We now have plans to order calendula seeds, apparently a “must-have” plant according to AI.
After your query (and you check the AI’s response using other sources), reflect on the experience. How do you feel about using it? Is the trade-off worth it? No wrong answers here. Consider sharing this experience with your students to facilitate a conversation.
AI will only make us dumber if we let it.
A recent study from the University of Pennsylvania and Wharton found that students who used AI for studying performed worse on exams than students who didn’t use AI. Students who did make gains in their studies while using AI often did so because they had “outsourced” their thinking. They solved 48% more practice problems with the AI. But because they hadn’t built the neural pathways themselves, when they were given an exam on the material with no AI tool available, they performed 17% worse than the group that didn’t use AI to solve the earlier practice problems.
Not receiving as much attention are the findings that when students used a “tutor” version of the AI (designed to guide students with hints rather than directly giving answers), they performed better on the practice problems than the control group (127% increase) and scored the same on the exam as students without access to AI while studying.
Also worth noting is that “students who used AI assistance were overly optimistic about their learning capabilities – even the high-achieving students. Teachers, on the other hand, seem to be overly concerned and tend to dismiss the advantages of AI.” The researchers think this mismatch is because “students and teachers aren’t yet trained on how to use AI effectively to augment traditional teaching methods.”
I’ve experienced this in my work as a writer. AI will produce an entire essay on my behalf if I am not clear that I only want to brainstorm around a topic or generate an outline. For a busy student with multiple assignments due the next day, it can be awfully tempting to have ChatGPT take over the reins and do all the work.
TRY IT: Be intentional around your use of AI. For example, show students how to use the “guided learning” mode when submitting a prompt to Google Gemini. Another strategy is to post an idea you have and ask AI to critique it. The tools are effective in helping you see the other side of your position, applicable for teaching persuasive writing or preparing for a debate.
AI will be everywhere.
Growing up watching Star Wars, I assumed the interactions depicted between living things and droids were a fantasy. Now it is our reality. People are having conversations with their phones and even their appliances as if they were with other human beings. Some of these interactions focus on deep topics, even when we are aware that AI is not a person.
Some people are using “AI companions” to help them navigate stressful issues in their lives. A Stanford study learned that when researchers impersonated teenagers in these programs, “it was easy to elicit inappropriate dialogue from the chatbots – about sex, self-harm, violence toward others, drug use, and racial stereotypes, among other topics.” AI tools tend to reinforce our biases, even when we are wrong, and not challenge us as a good friend or a therapist would.
Sherry Turkle documented these concerns with “artificial intimacy” years ago. In her book Alone Together: Why We Expect More from Technology and Less from Each Other, she found that “technology is seductive when what it offers meets our human vulnerabilities. And as it turns out, we are indeed very vulnerable. We are lonely but fearful of intimacy. Digital connections and the sociable robot may offer the illusion of companionship without the demands of friendship.”
I admit that I have used AI at times in this capacity. For example, I am not always confident about how to best discuss a topic with one of my teenage children. Claude (claude.ai) has helped suggest changes in my tone or approach, avoiding unnecessary conflict. Yet I can see how this resource can be a slippery slope: for example, not going to my wife first with these issues. Do we really want AI mediating our relationships? Where is the line between support and dependency?
TRY IT: Turn your preferred AI tool into a coach. Through the settings, add directions to your personal preferences so that the chatbot paraphrases and poses questions instead of offering blanket advice. Demonstrate this process for your students while explaining the reasons. (Claude has a helpful article on personalizing your preferences here.)
Remember that we have agency.
More money has been invested in the infrastructure and development of AI than the Manhattan Project and the Apollo program combined. (This visualization from Reuters is illuminating.) Data centers are being built everywhere, sparking debates and resistance regarding energy and water usage. A lack of land and resources on Earth has led technology companies to explore building data centers in space, which would use the sun’s limitless energy (plus, space has no environmental regulations).
As this new technology spreads throughout society and in schools, it can feel like we don’t have a say in how this will play out. But what seems to be its strengths – an enormous infusion of corporate dollars – may also be its weakness. Technology companies and their investors want to see a return on their investment.
They need us to buy what they are selling. That gives us power. As Karen Hao concludes in Empire of AI, “only through… powerful organized resistance can a technology transform from enriching the few to lifting the many.” Our classrooms can be a laboratory for this work. Learning with our students about the benefits and costs of using AI in our lives is a starting point for shaping how this change occurs.
TRY IT: Ask your preferred AI tool to develop a unit of study that engages students in the ethics of AI. Google’s NotebookLM allows you to upload various files to use as knowledge sources to create this type of project. Include speaking and listening standards, some of the resources shared in this article, and curriculum guides on project-based learning to guide the work.
You can empower your students by including them in the development of this unit of study, modeling appropriate AI use in the process. Here’s a prompt you can use to start this process:
“I teach ________ (subject area) to ___ graders. You are Jay McTighe, curriculum expert and creator of the Understanding by Design framework. My students bring a variety of backgrounds, experiences, and values to my classroom. I want to facilitate productive conversations about the costs and benefits of using AI in school and in life. Using the Understanding by Design framework, along with the resources I have uploaded, generate a draft unit of study that we can start with in this curriculum development process.”
Some closing thoughts
To close, we need principles to help guide our journey as we explore AI together. What the technology companies want isn’t 100% aligned with our own purpose and values. Consider instituting three rules for using AI in your classroom (and maybe in your life):
Human first. Before using a chatbot for advice, talk with another person. Use AI as a second opinion if you are still unsure about the issue.
Question deeply. Assume whatever response you receive from AI is likely biased and inaccurate. Don’t accept a response without at least one question that pushes back. You may find AI correcting itself.
Be intentional. WE don’t need to ask AI what we should have for dinner tonight. Use it if the value added in the response is worth the trade-off in terms of time, self-reliance, and depth of learning.
Matt Renwick is a systems coach for CESA #3, a state-supported education service agency in Southwestern Wisconsin. During his 20+ years in education, he has served as a principal in two Wisconsin elementary schools and as a middle school vice principal, teacher and athletic coach.
Matt is the author of Digital Portfolios in the Classroom: Showcasing and Assessing Student Work (ASCD, 2017) and Leading Like a C.O.A.C.H.: Five Strategies for Supporting Teaching and Learning (Corwin, 2022). Subscribe to his substack and follow him on Bluesky. Read all his MiddleWeb articles here.
Feature image by Pete Linforth from Pixabay
