Public Education Has an Anti-Artificial Intelligence (AI) Bandwagon Problem.
So, I have a question for the anti-AI celebrants: Are we creating yet another technology and educational-opportunity gap?
Because while some public schools debate whether students should be permitted to use artificial intelligence at all, some elementary and middle school parents will be teaching their children coding at home, enrolling them in after-school or weekend coding programs, or sending them to summer coding camps where AI technology will absolutely be incorporated into the learning experience.
And therein lies the danger.
Too often in public education, the least-informed—and frequently the loudest—nonprofessional pedagogical voices are allowed to drive educational policy, while professional educators with deep theoretical knowledge and practical experience in a particular field are pushed to the sidelines.
We have also entered an age in which simplistic, non-thinking solutions to extraordinarily complex problems are increasingly treated as preferable to thoughtful ones. Instead of carefully examining a problem, identifying its multiple dimensions, consulting knowledgeable practitioners, anticipating unintended consequences, and then developing a measured response, we reach for the sledgehammer when a scalpel is required.
We can see versions of this phenomenon far beyond education. Consider America’s immigration policy—or lack thereof. Rather than developing a sensible, comprehensive, humane, and well-contemplated approach to a profoundly complicated national issue, political pressure can produce sweeping, ugly policies and practices that disregard both the humanity of the people affected and the long-term best interests of the nation and its citizens.
Public education should not operate that way.
Everything cannot be rushed into policy because the loudest activist voices demand immediate action. Historically, some of public education’s biggest mistakes have emerged from precisely that kind of reactionary policymaking. (Do any former NYC high school principals remember the 1994 Sequential Mathematics Program disaster?) And those mistakes have often done the greatest damage to students from economically disadvantaged communities and, disproportionately, to students of color who are already positioned on the disadvantaged side of educational-opportunity gaps.
I remember, as a superintendent, meeting with a contingent of parents from a non-Title I school who wanted to eliminate those “bad” standardized state examinations.
I listened patiently to their arguments—talking points that, interestingly, closely mirrored those of another political entity in the district. But I digress.
Two important issues emerged from that meeting.
First, the parents made a legitimate point: state examinations did not—and could not—measure the full range of their children’s intellectual capabilities, talents, creativity, and possibilities.
I agreed.
But I also explained that the responsibility of school-based educators—and certainly my responsibility as superintendent—was to ensure that the state curriculum standards upon which those examinations were based represented the ground floor, not the ceiling, of what students were expected to know, understand, create, and intellectually accomplish.
A standardized state examination should not define the totality of a child’s education. But neither should we confuse the limitations of an assessment with the absence of a legitimate need for common academic standards.
The purpose of state standards and assessments is, in part, to establish a common expectation that every child in the state—and therefore every child in our district—should demonstrate sufficient proficiency to master grade-level, school-level, and course requirements while acquiring the prerequisite knowledge and skills necessary to succeed at the next learning level.
What is the alternative?
Should every classroom, school, or district simply create its own subjective definition of proficiency?
We would never accept that approach in other professions. Dentists, plumbers, lawyers, firefighters, airline pilots, and, yes, even school superintendents cannot simply declare themselves proficient, licensed, and certified. Some legitimate external authority, informed by expert practitioners, establishes knowledge and performance standards that must be met before someone receives a professional designation.
Why should we believe that determining whether a student has mastered essential fourth-, eighth-, or twelfth-grade knowledge and skills requires no common external standard whatsoever?
Can state assessments be improved?
Absolutely.
Should they constitute the entirety, or even the most important part of a child’s educational experience?
Absolutely not.
But eliminating them completely is another matter, particularly when our public schools already operate across enormous gaps in expectations, educational opportunities, access to resources, and parental capacity to independently supplement what schools provide.
Then came the second revealing moment in that parent meeting.
I asked:
“Are you willing to go on record as opposing all standardized examinations—the Gifted & Talented (G&T) admissions exam, eighth-grade Algebra screening exam, SAT and ACT, Advanced Placement examinations, International Baccalaureate examinations, the LSAT, MCAT, teacher-certification examinations, professional licensing examinations, and so forth?”
Silence.
My response was mental, not verbal:
“Oh, I see. Not those standardized examinations. Just the state examinations.”
And that brings me back to artificial intelligence.
The New York City Schools Chancellor made what I believe was a good and sensible interim decision. Having served as a superintendent in that city, I understand that sometimes educational leaders need to deliberately create a pause-and-think working space, particularly when outside political voices become so loud and distracting that thoughtful policymaking becomes almost impossible. And in New York City, educational decisions are inevitably complicated by political actors who will seize almost any opportunity to undermine a mayor or chancellor they oppose.
But a pause cannot become a permanent policy.
The next step should be the creation of a well-informed task force composed of experienced educators, curriculum and instructional specialists, developmental psychology experts, STEM practitioners, researchers, parents, students where appropriate, and—yes, absolutely—people from the technology industry, where many of our students will one day seek employment, establish professional partnerships, or perhaps create their own technology-related companies.
Its mission should be to develop a thorough set of meaningful, pedagogically sound, developmentally appropriate, and grade-specific recommendations to the Chancellor that can inform the development of guidelines governing how students should—and should not—use AI.
The question should never be reduced to:
“AI: Yes or no?”
That is far too simplistic.
The more useful questions are:
At what age? For what educational purpose? Under what conditions? With what safeguards? With what degree of teacher supervision? What should students be required to know and do independently before AI assistance is permitted? When does AI enhance intellectual development, and when does it replace the productive struggle necessary for intellectual development? How do we teach students to verify AI-generated information, recognize bias and fabrication, protect their privacy, and use these tools ethically?
Those are worthy and complicated educational questions that require the thoughtful work and informed input of highly skilled, experienced educators—not political performers.
And there are other realities that we cannot ignore.
Wearing my principal’s hat, I recognize the practical limits of what a central-office staff—especially in a large school district—can realistically monitor and enforce when it comes to carrying out official directives at the school and classroom levels. The reality is that there will be some NYC students, in some NYC elementary and middle school classrooms and schools, who are going to learn how to use AI applications regardless of what the official policy says.
And,
Public schools cannot control everything children learn through informal education outside the school building and beyond school hours. So, while one group of adults celebrates keeping AI out of a school, another group of parents, in that same or another school, may be enrolling their children in outside-of-school coding, robotics, engineering, science, and technology programs where those same children are learning how to use AI productively.
Five or ten years later, we may suddenly discover that we have an “AI achievement gap” and act mystified about how it happened.
We should not be mystified.
We created it.
I hope educators, STEM practitioners, parents, and community organizations—particularly those serving historically underserved communities—are already thinking about after-school, weekend, and summer programs that teach children the responsible use of AI in coding, robotics, mathematics, engineering, scientific inquiry, and other intellectually productive activities.
Technology, like so many human innovations, is not inherently the enemy. Students can benefit greatly from thoughtfully designed software-based instructional tools. The point is not that technology should replace teachers or student thinking. It should not. The question is whether technology can be intelligently designed and responsibly used to strengthen teaching, learning, practice, exploration, and intellectual independence.
We should therefore resist jumping onto every educational bandwagon that comes rolling down the street—whether that means enthusiastically and thoughtlessly going all-in on AI or reactionarily and reflexively going all-in on being anti-AI.
Instead, parents, educators, and community leaders should continually ask:
How will this policy decision affect my child’s ability to become the fullest intellectual and skilled expression of themselves?
How will this policy decision affect my community—particularly those children whose families may not have the financial resources to purchase privately what the public schools decide not to provide?
And, paraphrasing a warning I learned from my former Columbia University professor and mentor, economist Eli Ginzberg:
When major public policies are put into place, we had better ask very carefully whether Black people will once again receive the short end of the stick.
Finally:
What community-based educational defenses, enrichment opportunities, and support systems must we build now to prevent today’s policy decisions from strengthening tomorrow’s educational disadvantages?
Those are the questions we should be asking before we prematurely celebrate the demise of AI. Because when it comes to artificial intelligence and public education, the greatest danger may not be AI itself.
The greater danger may be making poorly informed educational decisions about AI—and discovering years later which children were given the opportunity to master the technology while other very capable children were prevented from learning it.
