Jenny Anderson, a journalist, is author of the Substack “How to Be Brave.” Rebecca Winthrop is director of the Center for Universal Education at the Brookings Institution and author of the newsletter “Winthrop’s World of Education.” Together, they wrote “The Disengaged Teen: Helping Kids Learn Better, Feel Better, and Live Better.”
Whenever a new technology comes along, schools often rush to build “literacy” in the new domain. That has certainly been true of artificial intelligence.
Many schools are training teachers and students how to use AI chatbots. They’re showing them how to design a good prompt and guard against hallucinations. They’re doing so under pressure from school boards and parents to protect kids from generative AI, but also because they want to ensure students are not left behind when they graduate.
This “master of AI tools” approach is tangible. It’s straightforward. It appears to future-proof kids. It’s also a massive mistake.
Kids don’t need to learn prompt engineering or proficiency with a never-ending menu of commercial AI tools. They need a holistic understanding of AI — not only what it can do in the moment, but also how it works and how it affects learning. When they deeply understand AI and then have time to practice with it, they develop the capacity to know when AI supercharges their learning and when it stunts it. None of this requires more time on screens.
That’s how to build agency over the tech, not just agility with what it offers.
Consider Seckinger High School in Gwinnett County, home to the largest school district in Georgia. Seckinger opened in August 2022 as America’s first AI-focused public school. Its students don’t spend lots of time playing with prompts; instead, teachers help students understand what AI is built on — data, algorithms, machine learning — so they can use it throughout their education. Social studies classes track disease patterns using data science. Math students use machine learning to identify shapes. English teachers debate the ethics of deepfakes.
Seckinger students outperform other schools in the district on English language arts, science, math and social studies. The school had so many requests for visits that the principal had to designate a day for guests to come to observe.
A similar approach can be found at the Kinkaid School in Texas. Sixth-grade teacher Claire Osborne starts her AI literacy unit by asking her students to complete the following two prompts: “I think AI is interesting because …” and “I’m worried about AI because …” Kids write down their answers before ever seeing AI tools in classrooms.
The goal is to explore how they think and feel about AI before they start using it. From there, Osborne teaches kids the vocabulary of AI (data, machine learning, bias) and helps them set expectations for responsible use. Finally, she gives them time to “practice with purpose” with the guidance of adults. This is what developing agency looks like.
The alternative approach — teaching a student how to use AI tools — is akin to teaching biology by focusing on instruments in a science lab: mastering the microscope or engineering the perfect petri-dish experiment. These skills are important, but they miss the bigger biological forces that shape the living world.
Teachers also need rigorous, well-designed training to understand what AI means for learning. For instance, Ed3 DAO, a nonprofit that trains teachers in emerging technologies, begins by discussing Bloom’s Taxonomy, a commonly used educational framework that ladders thinking skills from lower-order ones like remembering and understanding to higher ones like evaluating and creating. Which skills are AI good at? Which ones are humans good at? When might they work productively together? AI seems to be good at evaluating, but the training emphasizes the importance of ethics, contextual analysis and judgment in human decisions.
This agency-first approach is central to the AI Literacy framework developed by a coalition of prominent education organizations. Helping students understand not only what AI can do but what it should do is a central feature of its guidance.
Educators are eager to avoid mistakes schools made with social media, when they let kids bring phones to school without teaching them about addictive algorithms, the attention economy or the downsides of a 24/7 comparison culture. The internet posed similar problems when it went mainstream. Many schools embraced a tools-first approach to web literacy, racing to teach students what websites were and how to spot credible sources. That was a reasonable effort, but it ultimately did little to help students discern truth from fiction online.
A more holistic approach to AI won’t necessarily result in more use of the technology. In fact, research shows that the less you know about AI, the more likely you are to use it. People “can’t exercise agency over systems that are illegible to them,” writes Amarda Shehu, chief AI officer at George Mason University. Good AI literacy will accelerate smart use, including knowing when not to use it.
Parents are still trying to figure out how to view this technology. A poll conducted in the fall in Massachusetts found that while one-third viewed AI positively, another third viewed it negatively, and the rest were unsure or had “mixed” views. No one wants a wild west, but everyone wants their kids protected and prepared. Students easily get around simple bans. Outright embrace of AI so early is also likely to be harmful, and a new Brookings study finds the risks overshadow the benefits.
Emphasizing agency will help empower students to engage with AI from a position of strength. The choice is clear: Let them be pawns of AI, or masters of its vast potential.
The post Schools are thinking about AI all wrong appeared first on Washington Post.
