The real AI curriculum is not using it, but knowing when not to

By University of Hawaiʻi System News · July 8, 2026.

Three Hawaiʻi educators—Kawika Gonzales and Leah Aiwohi, both graduates of the College of Education at the University of Hawaiʻi at Mānoa, along with their colleague Chloe Sato—have been recognized in Washington D.C. as national finalists of the Presidential AI Challenge, after winning the regional round. The award honors their work leading Global Innovation Race Hawaiʻi, a statewide program that connects high school students across the archipelago to solve real problems in their communities using design thinking and artificial intelligence responsibly. They received a cash prize and access to digital resources to keep developing the program.

The detail worth pausing on is not the prize itself, but Gonzales's phrase that sums up the program's philosophy: "Teaching our students to use AI is no longer enough. We also need to teach them when to use it, why to use it, and whether it should be used at all." It is a subtle but important distinction, and it marks a leap beyond the first wave of AI digital literacy, focused almost exclusively on handling the tool (how to write a prompt, how to use this or that app). What this program proposes is a curriculum of judgment, not of operation.

This connects directly with one of the theses we have been maintaining about AI's impact on employment by sector: in education, the winner is not the teacher who simply transmits content, but the one who orchestrates AI with pedagogical judgment. The same logic, applied downstream, holds for the students themselves: in a world where knowing how to "use" generative AI will be a skill as basic and commoditized as knowing how to use a word processor, the differential value—and future employability—will lie in knowing when to delegate to the machine, when to trust human judgment, and when to simply abstain. It is exactly the kind of competence that neither a language model nor a rote-learning curriculum can replace.

That this initiative is born in high school classrooms and not in a university master's program is also significant. As context for the sector, most debates about AI and higher education revolve around detecting academic cheating or reformulating exams; programs like this one instead bet on moving the conversation about responsible use to younger ages, when habits of critical thinking toward technology are still forming. It is a modest effort in scale—a handful of educators, a small state—but it is exactly the kind of local experiment that, if it works, gets replicated: the national recognition and the resources accompanying the prize are designed for that.

Our reading is that these educational initiatives, though they don't grab headlines like the big model launches, are the ones that will determine whether the transition toward a society with an abundance of AI tools results in citizens capable of harnessing it with judgment or in passive users dependent on it. AI's long-term promise—freeing up human time for the work people are passionate about, not just automating tasks—demands generations that can distinguish useful delegation from the abdication of their own thinking. That is, at bottom, the true curriculum being put to the test in Hawaiʻi.