As schools rush to integrate generative AI, Misryoum argues the most durable preparation is computational thinking—not tool-specific training.
A decade after “Hello, world!” classrooms chased coding as the future, generative AI is now restarting the same urgency—only faster.
The question on educators’ desks is no longer whether students should learn technology. but which skills actually hold up when tools change.. Misryoum frames this debate around a critical classroom reality: the most visible AI training often focuses on using the latest systems. while the deeper instructional use case still feels unclear for many teachers.
For the moment, adoption appears uneven.. In a Misryoum-led. two-year research effort working alongside teachers. many educators—including those teaching engineering and computer science—struggled to name a clear. repeatable way to use generative AI across lessons.. That hesitation matters because it’s not just about comfort with a new tool.. It’s about whether the tool supports learning goals in a way that can survive the next update. the next product. and the next shift in what students are expected to know.
The “learn to code” era offers a cautionary parallel.. Many schools expanded computer science access after coding gained attention. but tool-focused instruction didn’t consistently translate into stronger long-term workforce outcomes.. Misryoum’s takeaway is less about dismissing coding—and more about recognizing how curricula can drift toward interface fluency.. When students learn how to operate a specific environment. they can end up performing tasks that resemble standardized test preparation rather than building durable problem-solving strength.
That limitation shows up again in today’s AI wave. where professional development often highlights prompt techniques and “getting good outputs.” Misryoum sees the risk here as well: prompt strategies can become a moving target.. If the educational aim is long-term readiness for technological change. the subject may need to shift from “how to use this tool” to “how to understand what the tool is doing.”
Computational thinking is emerging as that steadier anchor.. Misryoum describes it as a set of problem-solving practices used across computer science and other analytical disciplines. including breaking complex problems into parts. recognizing patterns. designing step-by-step processes. and evaluating the outputs of automated systems.. In an AI classroom. computational thinking also supports a different kind of literacy: students learn to examine why a system produces a certain response. not just whether the response sounds right.
Importantly, this is not only relevant for programming classes.. Misryoum notes that these habits can carry into fields as diverse as engineering, public policy, and data-heavy decision-making.. Students who practice decomposition and pattern recognition develop a way of reasoning that doesn’t depend on a particular app.. Even more. computational thinking helps learners treat AI results as data to investigate—signals that may reflect assumptions. limitations. or errors—rather than as authority by default.
Teachers participating in Misryoum’s work often described approaches that fit this framing, sometimes without using the term itself.. When students analyze why a chatbot makes an error, they are practicing evaluation.. When learners compare training data to real-world processes, they are connecting inputs to outputs and building causal understanding.. Misryoum highlights that these strategies can be taught with AI positioned as a case study. allowing students to focus on reasoning while still engaging with the technology that is shaping information today.
This shift also has a human impact inside classrooms.. Students are increasingly surrounded by algorithmic outputs—recommendations, summaries, translations, and automated feedback.. Misryoum’s argument is that computational thinking equips them to navigate that environment with judgment: they learn how systems generate results. where mistakes can come from. and how evidence matters when answers are produced quickly.
For educators, Misryoum suggests a practical direction: treat AI tools as objects of analysis.. Students can evaluate outputs, identify failure modes, and investigate how models generate responses under different conditions.. Lessons can connect AI to broader themes like data quality. algorithmic bias. and information reliability—areas that overlap with long-standing goals around critical thinking and media literacy.. Instead of allowing AI to replace the thinking process, these activities make the thinking visible.
Misryoum also sees an opportunity for edtech developers.. The classroom experiences described during Misryoum’s research show that many AI tools were designed as general-purpose systems. then introduced into schools later—often leaving teachers to translate them into instructional value.. Future products could benefit from earlier collaboration with educators so tools align with classroom learning problems, not just technical capabilities.. In the near term. teachers are already experimenting: building course-specific chatbots. drafting AI literacy lessons. and creating small. testable classroom applications.. Misryoum interprets these efforts as early-stage classroom “product development,” driven by teachers’ need for tools that support learning reliably.
The next step is governance—rules that help schools decide when and how AI belongs in learning.. Misryoum’s research team is moving toward a follow-up phase with partnerships across school districts to develop guidance on AI governance. with an emphasis on identifying conditions under which AI supports teaching and learning. and on reducing harm when it doesn’t.. Until educators can answer the question teachers ask every day—what can I actually use this for in math or writing?—most classrooms will keep experimenting cautiously. adopting what works. and relying on professional judgment to decide where AI enhances learning and where it distracts from it.
For schools, districts, organizations, or edtech teams interested in joining Misryoum’s next phase on AI governance, Misryoum invites collaboration through its research work.
Sentences Of The Week: What Education Must Learn About AI, Belonging & Reading
Educator Voice: The Secret to Sustainable School Change
AI in Schools: What Works, What Fails, and What’s Next—Misryoum Brief