In this episode, we are joined by Dr. John Spencer to talk about his new book, The A.I. Roadmap: Human Learning in the Age of Smart Machines. He helps us unpack a wide range of topics related to artificial intelligence (AI) in K–12 education, including some potentially poor choices, a better roadmap forward, implications for teaching and learning, and some practical examples of what AI looks like when it is integrated well into a classroom.
The blended approach embraces the strengths of both the human and the machine.
Dr. John Spencer, from his book, The A.I. Roadmap: Human Learning in the Age of Smart Machines
- AI in the K–12 Classroom (article collection)
- AI in the Secondary Classroom, with Julie York (podcast episode)
- MagicSchool, an AI Tool for Educators (podcast episode)
- Adobe Firefly: AI Image Generation (podcast episode)
- Khanmigo, a Personal Tutor and Teaching Assistant (podcast episode)
- The Power of Empowering Students, Part I, with Dr. John Spencer (podcast episode)
- The Power of Empowering Students, Part II, with Dr. John Spencer (podcast episode)
A Blended Approach
There are two extreme approaches to artificial intelligence that educators should avoid. One is a techno-futurism approach, where technology is seen as the answer to everything. In this scenario, tech is adopted for the sake of the tech itself. The other extreme is the “lock it and block it” approach. This strategy assumes that a technological opportunity is inherently harmful and should be blocked completely in order to avoid any potential problems. However, this approach unfairly deprives learners from the opportunity to learn about, and from, this technology.
Dr. John Spencer argues that the best approach is somewhere in the middle. He says, “We want to avoid both of those two extremes and instead go with a blended approach that centers it on the human aspects, and then ask, ‘What exactly can the AI add to it so that it’s still human-centered?’” Throughout this episode, we talk about a balanced approach and unpack the complexities of effectively integrating artificial intelligence into a K–12 classroom. The following are a few highlights from this episode:
- About Our Guest: Dr. Spencer is a former middle school teacher and current college professor, who is on a quest to transform schools into bastions of creativity and wonder. He wants to see teachers unleash the creative potential in all of their students so that kids can be makers, designers, artists, and engineers. He is the author of several best-selling books, including his newest title, The A.I. Roadmap: Human Learning in the Age of Smart Machines.
- Poor Approach #1: The techno-futurism approach is problematic because it focuses primarily on, “What is the newest, greatest [tech]?” Dr. Spencer says, “It doesn’t center it on anything human,” and it “ultimately leads to chasing novelty.”
- Poor Approach #2: “The second trap that we want to avoid—the second dead end—is the ‘lock it and block it’ approach.” Dr. Spencer explains, “The problem with that approach is that it doesn’t teach students to use it ethically and wisely. It doesn’t prepare them for what life will be like outside the walls of the school, but more importantly, it doesn’t leverage some of the capacity of the AI.”
- A Blended Option: Dr. Spencer believes that we need to find a space of compromise in the middle of the two poor choices—one that leverages AI with thoughtful intention and focuses on the human in the equation first. He shares the example of how AI will always beat a human chess expert, but a human using AI will almost always beat the AI. He adds, “Life is much more complex than chess. It requires contextual knowledge. It requires empathy. It requires all of these dynamic variables, and because of that, we will never want to outsource that to the technology, but we can leverage it in a way that is still human-centered, and that’s my hope with the AI roadmap.”
- We Need to Do Some Things by Hand: “There’s a lot of things that technology does really well that we still need to do in order to engage in deeper learning,” says Dr. Spencer. “Part of why we write is to make meaning. . . . We write in order to learn.” He also adds, “Handwritten notes, as old-school as that may be, lead to deeper learning and more retention of the knowledge, which in the end, it makes it sticky.”
- A Concern: Dr. Spencer doesn’t think that AI is going to take over systems, but he is “concerned about what this means for misinformation, what this means for democracy.” He adds, “My biggest concern is deepfakes and misinformation.”
- Information Literacy: Information literacy will need to evolve as the technology evolves. Dr. Spencer says, “In the future as we approach information literacy, it’s going to need to be a mindset, and it’s going to need to be a habit, and it’s going to have to be one of skepticism, and it’s really going to have to be looking at multiple sources, multiple angles, and it’s going to look different.”
- An Evolving Approach to Creativity: Dr. Spencer is a huge proponent of fostering student creativity. In terms of how AI will impact this, he thinks that the art of curation will become increasingly important. He says, “I kind of view curation as this bridge between creativity and consuming, and so I think curation will become an even more important human skill that students will have to develop, and the better we are at curating, the better we will be at finding our own voice. Because in the end, the truth is, in any work that we do, we’re all collage artists. We all sample.” Regarding the student role in this equation, he adds: “The role of voice, and tone, and style will be elevated a lot higher, and their ability to think critically about the ideas will sort of come from the AI, and they’re going to have to make it their own.”
- Project-Based Learning: In the world of AI, project-based learning will become increasingly important “because of all of those essential skills that you gain through doing a project.” Dr. Spencer believes that “in each phase of a project, generative AI becomes a powerful tool.” Among other tasks, AI can be used to generate questions, fill in gaps around prior knowledge, become a virtual partner and sounding board, summarize research, help with project management, and even edit final work.
- Assessment: AI has the potential to be transformative in this area of teaching and learning. Dr. Spencer believes, “It will automate a lot of the aspects of assessment that have a high percentage of user error.” It may soon be used to conduct fluency screenings with voice recognition, automate repetitive grading tasks, diagnose math errors, diagnostically screen student work, facilitate self-reflection and metacognitive thinking, and more. He adds, “On a basic level, it seems to be another set of eyes.”
- Cheating vs. Scaffolding: One approach to this challenge is to have students identify how they are using AI in their work. For instance, they might color-code what’s AI-generated from what they created from scratch. This approach allows the use of AI while leaning on transparency, truth, and trust.
- Personalized Learning vs. Adaptive Technology: Dr. Spencer says, “In beginning with the person, it’s focused on their curiosity, their questions, their interests, rather than students going through a program where everything is set for them and based on their answers. It just modifies it and changes it because [with] adaptive learning, basically, the student is invisible. They’re using the same algorithms for all students. They don’t really adapt to the individual students, whereas personalized learning says, ‘Look, who you are matters. Your voice matters, your curiosity matters, and we’re going to design it in a way that centers it on who you are as a person and on the relationship we have with our students.’ It’s messier. It’s harder, but I think what it does is it takes the notion of personalized learning, which in the past was often really hard to accomplish, and it just makes it that much more feasible.”
- AI in the Classroom: Dr. Spencer shares, “I’m imagining myself using it on a daily basis, but the bizarre thing is that if you walk into the classroom, so much of what you see is going to look a lot like the PBL-style [of] teaching that I did at the time, so it’s really just kind of happening, in many respects, under the hood.” He would use it in learning centers for students to ask questions, to facilitate imaginary conversations with historical figures, in creating checklists for project management, for scaffolding content, for creating leveled materials, and more.
- AI in STEM: In addition to effective uses in a humanities classroom, Dr. Spencer also sees a nice fit in STEM subjects. AI could be used effectively for Q&A, digital modeling, simulations, and virtual labs.
- FACTS Cycle: Dr. Spencer has formulated a prompt engineering model to help students and teachers effectively engage with generative AI chatbots. His model is called the FACTS Cycle, which stands for: Formulate questions, Acquire AI, Create context, Type a prompt, and Scrutinize results. Information about this model can be found on his website at spencerauther.com/facts-cycle/.
If you are listening to the podcast with your instructional team or would like to explore this topic more deeply, here are guiding questions to prompt your reflection:
- What are your thoughts about the two extreme approaches to AI?
- How might a blended approach work in your classroom or school?
- How might you balance the AI dichotomy of cheating vs. scaffolding?
- How could AI best be used to personalize learning?
- What role should AI play in student creativity?
- What would AI look like in your classroom?