Kristen’s Corner now comes to you on its own, rather than inside our quarterly newsletter. You can refer a colleague to Kristen’s insights by sharing this link.
By Dr. Kristen DiCerbo
August 2026
The rapid growth of generative AI has produced hundreds of tools for students, teachers, and school systems. These tools promise to create instructional materials, provide feedback, tutor students, support teachers, and address challenges ranging from literacy to mental health. Yet schools already have a long history of adopting exciting technologies that lead to scattered pilots, tool sprawl, initiative fatigue, and ultimately little change in outcomes before they are abandoned.
The key question when starting to choose technology is this: What problem are you trying to solve?
Schools face no shortage of problems, including low academic achievement, teacher shortages, absenteeism, discipline, scheduling, funding, student well-being, and graduation rates. None of these problems is simply a “lack of AI.” Because schools cannot address every challenge simultaneously, leaders need to prioritize problems.
One way to do this is to think about each problem related to four criteria: importance, tractability, controllability, and evidence base. A worthwhile problem to address should affect meaningful outcomes, have a realistic chance of improvement within roughly six to eighteen months, fall at least partly within the school’s influence, and have credible evidence suggesting that an intervention could help.
Applying the prioritization framework across school challenges often leads to a focus on learning outcomes: the knowledge, skills, and ways of thinking that allow students to progress toward their future goals. Learning outcomes are the important job of school, they can be tractable and controllable. There is an evidence base about what impacts learning outcomes.
There are certainly many interacting causes of achievement outcomes. Some are out-of-school headwinds, including poverty, housing instability, unequal background knowledge, and mental health challenges. But schools can acknowledge these while focusing on causes that sit within the instructional system, including curriculum quality, instructional practices, practice opportunities, feedback, implementation capacity, and organizational coherence. Focusing on controllable factors helps concentrate effort and restores educators’ sense that improvement is possible.
Our experience at Khan Academy illustrates the importance of clarity in the problem to solve. Districts that begin use of Khan Academy with a defined academic problem, such as lagging math achievement, are more likely to set specific goals around its use and integrate Khan Academy practice into their instructional materials and daily routines. Districts that simply want “an AI tool” tend to use isolated AI features, producing fragmented experiences and negligible changes in outcomes. Educational technology, including AI, works best as a solution to a specific educational problem.