OPINION:
It was a bit of news last month that most people probably paid scant attention to: In 2025, the United States lost ground in education — again — in reading and math.
The decline has continued for at least a decade, but last year’s drop was particularly steep. The Organization for Economic Cooperation and Development’s 2025 Program for International Student Assessment, known as PISA, put American students’ reading scores near a 25-year low and math scores at their lowest since the test began.
Undertaken every three years, PISA measures the capabilities of 15-year-olds worldwide in three core academic areas — reading, math and science — along with a new computational problem-solving domain in 2025.
People were interested in these podcasts
Last year, American students’ reading scores were near their lowest point in a quarter century, having fallen sharply since the last PISA was conducted in 2022. Math scores were among the lowest recorded since the test began in 2000.
It is worth noting that scores have been declining for years across industrialized OECD member countries, with overall performance declining last year. Still, scores rose in a handful of places, including the one that matters most to us: China. In its four participating provinces — Beijing, Shanghai, Jiangsu and Zhejiang — students ranked first in math and science and second in reading among all 91 participating countries and economies.
As part of his administration’s strategy to ensure the U.S. beats China in the artificial intelligence race, President Trump signed an executive order last year — “Advancing Artificial Intelligence Education for American Youth” — establishing a White House Task Force on AI Education and directing federal agencies to expand AI tools and curricula in schools.
Yet that may be the wrong way to approach the problem, and it could be one of the main reasons our children’s math and reading proficiency is plummeting.
In the U.S. and elsewhere in the West, teachers often use AI to “make things easier for” students, Andreas Schleicher, OECD director for education and skills, told The New York Times last month.
Meanwhile, in Asian countries, it is the teachers, not the children, who tend to use AI, he said. When the teachers use it, they do so “to make students think harder, work harder.”
Students need real teaching in foundational subjects and skills, not flashy tools that may end up obscuring a failure to comprehend material.
“Artificial intelligence runs on linear algebra, probability and optimization,” George Mason University mathematical sciences professor Padhu Seshaiyer wrote in a May op-ed for U.S. News & World Report. “In the age of algorithms, mathematics is not a niche discipline. It is national infrastructure.”
So here is how we “win” at AI. Stop making elementary schoolers haul laptops home every night. They are not tiny accountants. In fact, other than designated technology-learning times, let us cut way down on school-led computer use by children in general.
Let us replace it with a return to teacher-student interaction in which educators demonstrate to and drill children in core — and yes, sometimes dry — aptitudes. Assign paper homework. No Wi-Fi? No problem.
For the younger ones, bring back the focus on handwriting, cursive and multiplication tables.
Prioritize reading — of physical, full-length books, not misspelled, slang-riddled social media blurbs. That is not reading.
Go back to having students write out long-form test answers — not on a keyboard, but with pencil and paper. Emphasize proper grammar and spelling. If they make mistakes or change their minds, that is what the eraser is for.
Teach debate so that young people learn how to form and organize arguments and speak and write persuasively.
Reinstitute memorization — of word spellings, formulas, theories, dates, capitals, geographic locations and more. None of this requires a phone or a tablet.
Cutting back on technology in classrooms is not Luddism. It is the opposite. For us to achieve dominance in AI, we must ensure that its underpinnings are in place, and those are the core subjects on which we are now failing so spectacularly.
Proficiency — which we currently can only hope to attain — is not enough. We need excellence. In some ways, getting it will require turning back the clock and doing what we know works.
That is how the U.S. will win the AI race.
• Anath Hartmann is deputy commentary editor for The Washington Times.
Follow the author
Please read our comment policy before commenting.