3 Core Skills to Build Before AI Use

We've all either known—or been—kids who just don't seem motivated to do well in school. They may not complete assignments, might be disruptive, or even say outright that they don't see the value in their time in class. We also know students that definitely want solid grades but are keen to take whatever shortcuts they can because the positive report card—not the learning along the way—is the ultimate goal.

These issues aren't new by any stretch, but generative AI has highlighted them in ways that CliffsNotes and cheat sheets never could. And AI plagiarism detection just isn't reliable or equitable. Still, AI is the ultimate shortcut to so many assignments and subjects: If kids want to finish homework quickly and aren't convinced they're missing out on anything, learning will suffer.

If you've been reading about AI's impact on education, you've probably seen similar conclusions. They tend to argue that we need to rethink some fundamentals of traditional schooling to include AI as a reality while also avoiding some of its pitfalls. It may take a while for more schools to shift to standards-based grading and revamp curriculums. And hopefully there will be robust research into best pedagogical practices in the age of AI. In the meantime, it may be helpful to evaluate our own approaches and assignments through three building blocks that generative AI simply doesn't have, even though it can fake them: critical thinking, creativity, and connection.

There are multiple best-practice models emerging for AI (and we have a list of resources below), but the aim here is to keep it simple and ground pedagogy in fundamentally human abilities that will also make learning more relevant and engaging. This doesn't mean we ban or ignore AI! AI literacy is important and necessary. By teaching kids about the capabilities and limitations of AI, we can help them see the potential to extend and enhance our abilities, rather than replace them. And along the way, we can perhaps convince them via their own experiences that learning for learning's sake is actually worthwhile.

Critical Thinking

Image

Thinking critically is a huge consideration, and it influences all other areas. There are prescriptive frameworks to define critical thinking, but in the context of humans vs. AI, Terry Underwood's explanation of critical thinking works well:

"It's a whole-brain assault on passive acceptance of cultural models, absurd assertions, unwarranted assumptions, and faulty inferences. It's intellectual dissection, idea autopsies, life and death sense-making, vigilant ethical reasoning, disciplined skepticism, and analytical excavation. Above all, it is participatory and social, requiring engagement with diverse perspectives and collaborative meaning-making." – Learning to Read, Reading to Learn

Getting kids to engage with ideas fully—including others' voices so they can make meaning of it all for themselves—is a worthy goal of any assignment. Sure, generative AI amalgamates info that it's collected, and can come up with impressive responses from time to time, but it's unable to make its own meaning from a participatory, social process. 

First, kids need to have foundational knowledge to pull from (and that AI can extend). Without a knowledge and skills base, there's not much to work with. Ultimately, allowing AI to make meaning for us is offloading a skill we need to extend learning contexts and combine conceptual frameworks from our own unique neural networks. And imagine how this process deepens within a lively group discussion: AI can't—and shouldn't—take this process from us.

Below, you'll find ways to embed the seeds of critical thinking into core subject matter. Of course, standards and district-mandated curriculums may limit your flexibility. And the provided examples may or may not resonate with your pedagogical practices. But at the very least, perhaps these ideas will get you thinking about ways you can alter your approach and assignments to account for generative AI.

Focusing on the roots

• Foundational skills and info: Think about the surface-level information that AI can provide easily, and then go a level deeper. Then another. Rinse and repeat until you get to the skill or foundational knowledge that kids need to have to keep learning and thinking critically.
  • Example: Having an orientation of the world, its landmasses and bodies of water, gives kids important context for other topics. But memorizing every capital city of every country may be less essential—unless you want kids to explicitly practice short-term memory skills.
• Crucial context: It's particularly helpful to learn this foundational information when it's embedded within the context of why it's important.
  • Example: It's more meaningful to learn about where certain landmasses are while also studying where one's family is from, where current events are happening, where favorite animals live, etc.

Curiosity, relevance, and reflection

• Wondering: By couching foundational knowledge in curiosity, we keep the pilot light of learning lit. At some point in school (often around sixth grade), kids start to lose this natural curiosity, at least in classroom contexts. Helping kids hold on to it by allowing them to start from things they wonder about (as often as possible) can truly create lifelong learners.
  • Example: If a student is curious about how football players get to be great, help them assemble a project to investigate different contributing factors, do the research, and demonstrate their learning. AI might help them organize different factors into areas of research.
• Personal investment: Allowing kids to address issues that have meaning to them is also key in helping kids make meaning for themselves. If they don't feel that what they're learning is relevant to them, they're more willing to offload it to AI.
  • Example: Learning how your country's government works is a key element in being an active, informed citizen. Since kids can easily offload memorizing facts about how many senators there are, you can start with the taxes that will be taken out of their paychecks. Where does that money go? Why? How is that amount determined? Once they feel that relevance, they can create a model—maybe learning about charts and percentages in the process—that represents their understanding.
• Reflection: Have kids reflect on their own learning and what role AI could—or shouldn't—play in the process. When they examine what's gained and lost by using generative AI, they can think critically about its use.
  • Example: Give kids an opportunity to express something meaningful without AI, and then with it. What are the differences between the two outputs? Is there a point in a creative process where AI might be helpful, or no?
• Metacognition: Along those lines, asking kids to reflect on the learning process and their confidence in what they know is another avenue to think critically about the learning process.
  • Example: Before a bot quizzes kids about a topic, have kids reflect on how confident they are in their knowledge and what learning strategies seemed to work best. The quiz results can inform further reflection.
     

Creativity

Image

Another key element in the learning process in an evolving AI era is creativity. Studies have shown that it can help extend human abilities when used with specific prompting strategies, but generally it falls short. And though generative AI can produce images of all kinds, it isn't coming from a place of lived experience combined with knowledge and emotions. Allowing students to express themselves to both learn and demonstrate learning is a sure way to bypass AI and help kids retain agency and true learning.

Self-expression and choice

• Choices and agency: As we audit our lessons and assessments, we can look for opportunities to allow kids to choose how to learn and express learning. When they connect core concepts to things they care about and then create something, they're invested in what they're learning in a different way.
  • Example: If students are learning about the water cycle, giving them options to illustrate their understanding—maybe even ones we wouldn't immediately think about, like a Minecraft creation or vibecoded game — allows more agency.
• Extending and inspiring vs. replacing: If we're coming from a place of authentic creativity, AI can extend our abilities in some ways. Helping kids understand the difference between offloading and extending illustrates AI's possibilities and limitations. And when they feel stuck in a creative process, turning to a bot just might unlock new ideas.
  • Example: Kids might have a creative fundraising or community-building concept but aren't sure how to build a website. AI can help them with those logistics to make their ideas a reality.

Making meaning

• Unique representation: AI can write music and poems, and can create visual art. However, even with very specific prompts, it can't fully embody a human being's experiences, including sensory ones.
  • Example: AI can lie and say that it has sensory experiences, so testing and comparing those hallucinations to students' personal perceptions is an interesting exercise. Have kids identify sensory experiences connected to their own memories and create representations with art or words. How do those differ from AI outputs?
• Unmasked bias: Without causing further harm, it can be useful to have students examine the biases that AI has assimilated and creatively compare them to their own experiences.
  • Example: Ask students to create something that represents an element of their cultural backgrounds and what it has meant for them in terms of experiences and values. How does AI portray various backgrounds, occupations, and life experiences? What's missing from those representations? 
     

Connection

Image

No matter what a bot tells us, it can't care about us. It can't bring its experiences into a discussion to help illustrate a point. It can't care about our well-being. Among many other things, these inabilities make AI unable to forge two-sided, human connections with students that can deepen learning and development in immeasurable ways.

Teacher-to-student

• Deeper personalization: Really capitalizing on the complex connections we create with students is more important than ever. Discovering what makes our students tick and then finding threads of connection to the curriculum is part of the art of teaching. But if you need specific examples or explanations about a concept that align with niche interests, AI can be a helpful tool.
  • Example: An accomplished soccer player is struggling in math, and you've already provided support and encouragement. What you need now are very specific examples of the order of operations aligned with the rules of soccer! Sit down with the student and show them how to craft prompts that can deliver examples that speak to them.

Community-building

• In-person connection: There's no easy stand-in to satisfy the human needs of belonging, being seen, and feeling valued. And cellphones are already impacting in-person academic and SEL skills in schools. Prioritizing a positive school climate and community can help solve all kinds of problems, and AI can't create that. However, it might be able to help you facilitate them.
  • Example: You need ideas for lunchroom conversation starters and games, now that students can't use their cellphones in common spaces. Ask AI for some activities that are age-appropriate and even based on pop culture. Place whatever parameters you need, provide translations, and generate new ones each week or month (without spending hours doing so).

Resources

• AI Literacy lesson collection
• Offline Digital Citizenship: Soft Skills to Support Strong Online Habits
• Games for Building Critical Thinking Skills
• STEM Apps for Higher-Order Thinking
• SchoolAI
• Decrease Digital Distraction and Elevate Attention Spans
• Search Engine podcast episode: "What Should We Do About Teens Using AI to Do Their Homework?"
• More Than Words: How to Think About Writing in the Age of AI
• AVID Open Access: AI in the K–12 Classroom
• Cambridge Life Competencies Activity Cards