The classroom experiment that was supposed to showcase the future of education has instead triggered a much bigger question: how much of a child's education should be handed over to machines?
The future of education was supposed to look exciting.
A humanoid robot walks into a classroom.
It speaks to students.
It answers questions.
It demonstrates technology.
It helps children learn robotics and artificial intelligence.
The teacher remains nearby, while the machine becomes a new kind of teaching assistant.
That was the vision behind a nearly $60,000 humanoid robot approved by the Salamanca City Central School District in rural upstate New York.
The robot, nicknamed "Sally," was supposed to help high-school students studying robotics and technology.
Instead, the experiment triggered backlash from teachers, residents and state education officials.
The school district eventually paused the plan.
And then something even more significant happened.
New York lawmakers began moving toward legislation that could effectively keep AI-powered humanoid teachers out of classrooms across the state.
This is no longer simply a story about one controversial robot.
It is becoming a referendum on what we think a school is supposed to be.
The robot was supposed to represent the future
The Salamanca proposal was not necessarily an attempt to replace teachers.
The stated idea was to expose students to emerging technology and help those studying robotics and related fields.
That sounds reasonable.
Schools should prepare students for the technologies they will encounter in adulthood.
But the controversy became much bigger because of the particular machine involved.
The robot came from Realbotix, a company known for developing highly realistic humanoid robots.
The company's connection to the adult-oriented robotics industry became one of the concerns raised by critics.
Suddenly, the question was no longer simply:
"Can a robot teach children?"
It became:
"Should this kind of machine be interacting with children in an educational environment at all?"
And that distinction matters.
Because once a machine is designed to look, speak and behave like a human, the psychological relationship between child and technology changes.
A laptop is obviously a machine.
A chatbot is obviously software.
A humanoid robot is different.
It can look at a child.
Speak to a child.
Move toward a child.
Respond emotionally.
And potentially create the impression of a social relationship.
That is where the technology becomes much more complicated.
New York lawmakers are now asking a bigger question
The backlash did not end with Salamanca.
New York State Senator Shelley Mayer and other lawmakers introduced S10671, a bill specifically targeting AI-enabled humanoid robots and AI avatars in instructional settings.
The legislation would require school districts, charter schools and boards of cooperative educational services to establish policies prohibiting AI-enabled humanoid robots from performing instructional functions.
That includes delivering classroom instruction, tutoring students, providing educational support and even assessing or grading student work.
An Assembly version, A11693, was introduced in September and is currently before the Assembly Education Committee.
Another proposal goes even further.
Assembly Bill A11662 would prohibit schools from purchasing, leasing, deploying or otherwise using humanoid robots in instructional settings, with limited exceptions for robotics courses, demonstrations, exhibitions and research.
It would also require instructional settings to remain under the direction and control of certified educators.
AI could still be used as a supplemental resource under teacher supervision.
In other words:
New York isn't necessarily saying no to AI. It is beginning to draw a line between AI as a tool and AI as a teacher.
That may become one of the most important distinctions in education policy.
The machine isn't the problem. The replacement is.
There is a legitimate argument for using robots in schools.
A robotics student should be able to build and program robots.
An engineering student should understand autonomous systems.
Computer science students should experiment with artificial intelligence.
Children should understand how machines perceive their environments and interact with humans.
There is nothing inherently wrong with that.
In fact, refusing to expose students to robotics could leave them unprepared for the future.
The problem begins when the machine moves from being the object of education to becoming the educator.
That is a completely different proposition.
A robot used to teach students how robotics works is a learning tool.
A robot used to replace a teaching assistant is something else.
A robot that becomes responsible for tutoring children, evaluating them or providing social-emotional interaction crosses another boundary.
And New York's proposed legislation is essentially saying:
There are some things machines should assist with, and there are some things humans should continue to own.
We are discovering that AI has an education problem
The humanoid-robot controversy arrives at exactly the moment when concerns about AI in education are exploding.
The latest PISA data provides an uncomfortable backdrop.
The OECD's latest global assessment involved more than 760,000 students across 91 countries and economies.
And the data suggests that students who use AI for certain schoolwork tasks can perform worse than students who do not.
The issue is not necessarily artificial intelligence itself.
It is cognitive outsourcing.
When a student asks AI to summarise a chapter, write an essay, solve a problem or conduct the thinking required for an assignment, the student may receive a finished product without developing the intellectual ability the assignment was designed to build.
That is why the humanoid robot debate matters.
Because AI is no longer simply sitting inside a computer.
It is beginning to acquire a physical presence.
Imagine the classroom of 2030
Imagine a classroom five years from now.
The teacher is standing at the front.
Twenty students are working at their desks.
And beside them is an AI-powered humanoid assistant.
One student asks the robot for help.
Another asks it to explain an equation.
Another asks it to read an essay.
Another asks it to provide feedback.
The robot responds instantly.
It never gets tired.
It doesn't need a lunch break.
It doesn't take sick leave.
It can theoretically interact with hundreds of students.
And unlike a human teacher, it can potentially maintain a detailed digital record of every interaction.
At first glance, this sounds incredibly efficient.
But education is not simply an efficiency problem.
A school is not a factory.
The objective is not merely to produce correct answers at the lowest possible cost.
Children are developing language, confidence, judgement, empathy, social skills, curiosity and relationships.
A teacher does more than deliver information.
A good teacher notices when a student is frightened.
A teacher notices when someone has stopped participating.
A teacher understands the difference between laziness and exhaustion.
A teacher knows when a student needs encouragement rather than another explanation.
Can an AI eventually imitate some of these behaviours?
Probably.
But imitation is precisely where the ethical problem begins.
Should a machine simulate empathy?
This may be the most uncomfortable question of all.
If a robot says:
"I'm proud of you."
What exactly is happening?
The machine does not experience pride.
If it says:
"I understand how you feel."
Does it understand?
Or is it predicting the most appropriate linguistic response?
For an adult, that distinction may be obvious.
For a young child, it may not be.
That concern is explicitly reflected in the reasoning behind New York's proposed S10671.
The bill's justification argues that humanoid AI systems can simulate human appearance, emotion and social interaction, potentially blurring the distinction between genuine human relationships and artificial ones, particularly for young or vulnerable students.
This is not a science-fiction concern anymore.
The technology is arriving before society has finished deciding what psychological boundaries should exist around it.
The privacy problem is even bigger
There is another issue that receives far less attention than the novelty of a robot walking around a classroom.
What does the robot know?
A humanoid educational robot may potentially interact with children through microphones, cameras, sensors and software systems.
Depending on how it is designed and deployed, it could collect enormous quantities of behavioural information.
Who talks the most?
Who struggles with mathematics?
Who becomes frustrated?
Who asks certain questions?
Who behaves differently around other students?
Who appears distracted?
Who has difficulty communicating?
Who needs additional assistance?
That information could become incredibly valuable.
It could also become incredibly sensitive.
And this is where the AI-in-school debate connects to a much bigger question about children's data.
Should a technology company be building detailed behavioural profiles of children simply because its product has been placed inside a classroom?
New York's A11662 proposal specifically includes student privacy protections and would restrict AI systems used by schools from collecting, retaining, selling, licensing or transferring biometric information.
That is a signal that lawmakers are beginning to understand something important:
The AI classroom is also a data-collection environment.
New York is not rejecting the future
This is where the debate needs nuance.
It would be easy to portray New York as anti-technology.
That would miss the point.
The proposed legislation does not necessarily prohibit schools from teaching artificial intelligence or robotics.
In fact, some of the proposed rules explicitly allow humanoid robots for robotics, engineering, computer science and AI courses, as well as demonstrations and research.
What lawmakers are challenging is the idea that a machine should become the primary instructional authority.
That is a very different position.
New York is essentially saying:
Teach children about robots.
Teach children about AI.
Let them build robots.
Let them experiment with AI.
But don't quietly turn the robot into their teacher.
That is arguably a much more intelligent technology policy than simply banning artificial intelligence.
The irony is impossible to miss
The AI industry has spent years promising that technology will solve educational problems.
Teachers are overloaded.
Classrooms are crowded.
Students need personalised attention.
There are teacher shortages.
AI can supposedly provide individual tutoring at scale.
And those arguments are not completely wrong.
AI could genuinely help.
But there is a dangerous temptation in technological thinking:
If a machine can perform a human task, assume the machine should perform it.
That is not necessarily true.
There are things humans do inefficiently that we should happily automate.
But there are also things humans do precisely because the human relationship is part of the value.
Education belongs partly in that second category.
Africa should be watching this debate
For countries such as Nigeria, this debate is particularly important.
Nigeria needs technology.
There are enormous gaps in access to quality education.
AI could help students access tutoring.
It could translate educational materials.
It could assist teachers.
It could help students practise difficult concepts.
It could provide educational resources to communities where qualified teachers are scarce.
But there is a danger in importing educational technology without importing the debate surrounding it.
If wealthy countries are beginning to ask whether AI should replace human instruction, African countries should not blindly accept the idea that the solution to teacher shortages is simply to install machines.
Nigeria needs to ask a different question:
Are we using AI to strengthen teachers — or using AI because we have decided not to invest enough in teachers?
Those are not the same thing.
A robot cannot become an excuse for underfunding education.
The classroom should not become a testing ground for corporations
This may ultimately be the most important lesson from Salamanca.
Technology companies have an obvious incentive to demonstrate their products in schools.
Schools have an incentive to appear innovative.
Governments want to be seen as technologically progressive.
And students become the people inside the experiment.
But children cannot be treated as an endless beta-testing population.
Before deploying AI systems in schools, governments should demand answers about:
Who owns the data?
What information is collected?
Where is it stored?
Can it be used to train commercial AI models?
Who is responsible when the system gives a child dangerous or incorrect advice?
Can parents opt out?
Can teachers override it?
Can the system be independently audited?
What happens when the company disappears?
And perhaps the most important question:
What educational problem are we actually solving?
If the answer is simply "because the technology exists," that is not good enough.
The robot may have arrived before the rules
That is the bigger story emerging from New York.
A school wanted to introduce a humanoid AI robot.
The experiment met resistance.
The school paused.
Lawmakers responded with legislation.
And now New York is debating whether humanoid AI systems should have any instructional role in schools at all.
The irony is striking.
The machine was supposed to demonstrate the future of education. Instead, it helped trigger a debate about the limits of that future.
And perhaps that is exactly what should happen.
Technology should not be allowed to define the future simply because it can.
Humans should define the future and decide where technology belongs inside it.
AI can be the tutor.
It can be the assistant.
It can be the research partner.
It can be the laboratory.
It can even be the subject students study.
But there is something we should be extremely careful about turning into a machine:
the human relationship at the centre of education.
Because if the purpose of school is only to transfer information, a robot may eventually win.
But if the purpose of school is to help human beings learn how to think, communicate, question, cooperate and become human beings themselves, then the teacher is not an outdated piece of technology.
The teacher is the point.
Comments
Post a Comment