American education is overdue for disruption. For generations, we have educated children in roughly the same way: grouping them by age, following a standardized schedule, ringing a bell, and sending them to the next class. Whatever the merits of that model, it was built for a different era, and as Americans, we should be asking whether it is still the best way to educate our children today.
It makes sense then that Alpha School, an innovative new education model, has received so much fanfare, including Education Secretary Linda McMahon recently praising it on CNN. Alpha School flatly rejects the “teacher-at-the-front-of-the-room” model. It rejects teachers completely, for that matter, while claiming that its AI can teach students “2x in 2 hours.” For a minimum of $40,000 a year, your child will spend an average of two hours each morning completing core academics through customized, AI-driven tutoring, then spend the rest of the day developing skills like entrepreneurship, leadership, public speaking, and financial literacy.
There is much to laud in this model. It embraces the very principles that the homeschooling movement has understood for decades: the traditional 7-hour school day is a relic, most children can be autodidacts with assistance, and parental ownership of the children’s education is non-negotiable.
Alpha School has released non-independently verified data to back up their claims, and their standardized scores are impressive. But measuring the success of their method on the motivated children of the top 1% is a little like asking Michael Jordan if his shoes helped him dunk a basketball. When the outcome is more or less inevitable, the interventions are cosmetic.
Replacing human teachers with software in the name of disruption is problematic, as is Alpha apparently surveilling children in their own bedrooms in the name of accountability. And I worry that Alpha’s AI-as-teacher approach and American educators’ broader embrace of tech in the classroom rest on an unexamined premise: that mandatory hours of screen time over traditional methods have no downside. Given what we know about how the brain processes reading on paper versus screens, as well as the benefits of writing over typing, both of these processes are crucial to recruit more of the brain to work together through neural integration, which then leads to more lasting memory.
Alpha’s gamified, metric-heavy approach is also concerning. Think Khan Academy, but infused with an Orange Theory Fitness metric-based vibe. The constant assessment, while motivating for some, can easily grind down the wonder inherent to a child’s constitution. And hotwiring a child’s internal dopamine regulation system to achieve results may also have long-term negative effects. Learning is not meant to be easy or even fast. Confusion, frustration, and failure are not bugs in the process. They are part of how competence is built, and technology that circumvents the human element is likely to be a long-term detriment.
And while knowing the etymology, spelling, and meaning of every word helps you ace standardized tests, should that come at the expense of getting to sit with your elementary school classmates, the windows open to the sweet spring breeze, while your teacher reads you Jack London’s “Call of the Wild” or E.B. White’s “Charlotte’s Web?” Can an AI replicate that experience with a screen and headphones?
By uncoupling judgment, depth, and emotion from knowledge through their choice of software over humans, Alpha seems to be arguing that knowledge in and of itself can beget judgment. But human teachers naturally pass on much more than sterile knowledge; they pass on how and when to apply that knowledge to the broader world through empathy and wisdom.
Every generation gives children new tools. In the broader education landscape, the question is not simply whether the tool is powerful, but whether the child has first developed the judgment required to use it well. That becomes especially important when the tool promising to accelerate learning can increasingly perform the thinking itself.
My concern is not what we don’t know about AI, but rather what we already know about human beings. We tend toward the path of least resistance. LLMs in the hands of experts, people who have already done the difficult work of learning, failing, struggling, and developing judgment, is a reasonable proposition. It can augment abilities that already exist. AI in the hands of children is an entirely different proposition. LLMs effectively hobble their users’ ability to originate thought. Outsourcing that capacity before developing it should give us all pause. Descartes rightly connected thinking to human existence itself, so why then surrender even a part of that most essential human function?
As a civilization, we are now grappling with the “Matrix” problem: do we prefer a sterilized version of reality to the real thing? “The Truman Show” or Solzhenitsyn? AI-composed jazz or Miles Davis? Is there a difference? And if so, do we need our children to be able to discern the difference? And if we do, does the difference even matter? I would answer in the affirmative to all three.
In addition, an LLM is predisposed to flatter its user to the point of sycophancy. It imitates confidence while devoid of it, emulates originality without possessing any, and does so because it was told to, but not because it wants to. In the words of a Christian apologist, “How can an object that wants nothing, fears nothing, enjoys nothing, needs nothing, and cares about nothing have a mind?”
And this leads us to the following paradox: that the best preparation for an AI-infused world may be a largely analog education. In short, a disciplined, sound mind is the best way to delineate the superficial from the real. The student with a grounding in history, a command of philosophy and mathematics, and a developed capacity for independent judgment can use AI as a tool without becoming dependent upon it and will most certainly not be technologically incompetent.
Conversely, the intellectually ill-equipped who have outsourced their thinking may become the most dependent upon AI while simultaneously being unable to discern its errors. That asymmetry should matter enormously to educators because intellectual skill is not just producing an answer but knowing whether the answer deserves to be believed.
As McMahon herself articulated later in the same CNN interview, speaking about higher education, it is open debate and the “collision of ideas” that move education forward. She’s exactly right. But so is Princeton professor Lauren Wright, who has found that conservative students at progressive universities get “sharper and sharper” because they are forced to encounter arguments they reject and defend their own beliefs. This friction builds intellectual grit.
Our children need to be challenged, not placated. We have no longitudinal evidence showing what happens when children outsource learning to artificial intelligence, especially in their earliest years. Our children are, in a very real sense, the experimental group, and by the time we understand the consequences, it may be too late.
The Trump administration has an extraordinary opportunity to rethink an educational model that no longer serves many American children. It should be ambitious enough not merely to digitize the old one.
In one telling of the legend, Arthur was not worthy because he possessed the sword, Excalibur; he possessed Excalibur because he had proved himself worthy. Before students are given a tool capable of doing so much intellectual work for them, they should first develop the intellectual and moral capacity to wield it.
Perhaps Alpha’s most important lesson is not that children need more AI, but that children need more life. More conversation, competition, failure, responsibility, and all the messy human experiences through which children develop. Teach children to think before a chatbot thinks for them. Teach them to write before AI writes for them. And above all, teach them to look for the “why” instead of just the “what.”
The goal of American education should not be to prepare children merely to operate the technologies of their time. It should be to prepare them to judge those technologies wisely.
***
Dr. Benjamin Chacko is an Emergency Medicine Physician and the father of two sons.
.png)
.png)
