The full white paper. Twenty-five pages on skill formation, the digital divide, and appropriate use of technology in the AI era. Read the designed PDF.
In June 2026, two of the largest school districts in California moved on the same day to restrict what students can do on district devices. The impulse is understandable. The pattern it creates is not new, and we know how it ends.
The pattern
Every time a technology arrives in schools, the same sequence runs. The technology appears. Some students use it badly. The institution responds by restricting access. The restriction lands hardest on the students who have no second device at home, no parent who can explain the workaround, and no adult modeling what skilled use looks like.
The students in well-resourced homes keep building. The students who depend on the school for access stop. Ten years later we describe the resulting difference in capability as an achievement gap, as though it appeared on its own.
The students most likely to be handed a locked-down device are the students who most need to be building with these tools.
Consumption and production are not the same skill
Most of the policy conversation treats access as a single quantity. It is not. There is a difference between a student who consumes what a system produces and a student who builds with it, and the difference is not time on device. It is whether anyone taught them to treat the output as something to interrogate, revise, and own.
A blanket restriction removes both. It stops the consumption we were worried about, and it stops the production we said we wanted. What survives is the least valuable use, on a personal phone, with no adult in the room.
The LEVER position
The LEVER Framework is a communal systems pedagogy built on the premise that better models come from communal design and networked feedback, and that the goal is sovereignty rather than success inside someone else's system. Applied to this question, its five dimensions say something specific.
| Dimension | The principle | What it says about access |
|---|---|---|
| L — Leverage | Small changes at high-impact variables create large effects | Access is a leverage point. Restricting it is a large change at a high-impact variable, made in the wrong direction, usually to solve a smaller problem. |
| E — Environment | Context shapes behavior; designed environments are pedagogy | A locked device is a designed environment. It teaches that this technology is something done to you, not something you do. |
| V — Velocity | Rapid iteration and feedback accelerate learning | Students denied the tool cannot iterate with it. Their peers are running feedback loops they are locked out of. |
| E — Execution | Output is learning; performance-based, not theory-based | You cannot assess what a student can build if the building is prohibited. Restriction pushes assessment back toward recall, which is exactly what we said we were leaving. |
| R — Repetition | Deliberate practice compounds | This is the one that matters most. Capability compounds. A year of denied practice is not a year of pause. It is a year of divergence. |
What the alternative actually requires
The alternative to restriction is not permissiveness. It is harder than restriction, which is why restriction is popular.
- Name the harm precisely. Most device restrictions bundle four different problems: distraction, academic dishonesty, exposure to harmful content, and data privacy. They have different solutions and only one of them is access.
- Change the task, not only the tool. If a tool can complete the assignment, the assignment was measuring the wrong thing. This is a pedagogical problem wearing a technology costume.
- Teach production explicitly. Students do not become producers by proximity. Someone has to teach interrogation, sourcing, revision, and the judgment about when not to use the tool at all.
- Put the adults through it first. A teacher who has never used the tool on their own work cannot supervise a student using it on theirs. This is the 201 and 301 problem, and it is the real constraint.
- Keep the guardrails at the system level. Data handling, age-appropriate defaults, vendor terms, and audit trails belong in procurement and infrastructure, not in a rule that a fifteen-year-old has to follow on their own honor.
The question underneath
Every generation of school technology policy answers one question, usually without asking it out loud: who is trusted to create with this, and who is only trusted to receive it. The answer has been remarkably consistent, and it has mapped onto income and race with a precision that should embarrass us.
We are answering it again right now, faster than we answered it before, and the decisions are being made in board meetings with two weeks of notice.
Our students should leave us as producers of this technology, not only consumers of it. The through-line of this work, and the reason the capability levels matter more than the tools.
LEVER Framework © Dr. Marie Martin of Alexandria's Design.