What Can People Do About AI?

Most people do not get to decide what a frontier lab trains next. That does not mean they have no role in how AI enters their lives.

The useful question is not, “Can I personally stop AI?” It is, “Where are decisions about AI being made around me, and what would it take for those decisions to be open to question?” AI becomes real in a school, workplace, public agency, community organization, or family long before it becomes a civilization-scale abstraction.

1. Who is making the decision?

Every AI system arrives through a decision: a district buys a product, a workplace changes a policy, an app turns on a feature, or a public agency adopts a vendor. Start there.

Ask who made the decision, what problem they were trying to solve, and who was included before it was made. “The technology is here” is not an answer. It can hide choices about money, staffing, risk, and whose work will change.

2. What can the system actually reach?

An AI system is more than a chat box. It may be connected to student records, documents, email, cloud storage, cameras, calendars, accounts, or other people’s work. Its real power comes from those connections.

Ask what information the system can see, what actions it can take, and what access it retains after a task is complete. The answer should be specific enough that a person can understand the consequences of saying yes.

3. What is the evidence that this use is worth it?

Promises of efficiency, personalization, safety, or innovation are not evidence on their own. A meaningful case for using an AI system should explain what it improves, for whom, compared with what alternative, and at what cost.

Ask what the evidence shows, what has not been tested, and whether the people who will live with the system recognize the claimed benefit. A vendor demonstration is not the same thing as an accountable evaluation.

4. Who can challenge the system—and what happens when it fails?

The final question is about power. Can a student, family, employee, or resident ask how a decision was made? Can they correct an error, opt out, appeal a result, or get a person involved? Can the organization pause or remove the tool if the harms outweigh the benefits?

If no one can answer those questions, then the system is asking for trust without offering accountability. That is not a technical gap. It is a governance choice.

Start close to home

These questions are useful because they make an overwhelming topic concrete. They apply when a school considers an AI tutor, when a workplace connects an assistant to internal files, or when a government agency introduces an automated decision system.

They will not resolve every disagreement about AI. They can make the disagreement visible, locate responsibility, and give people a way to ask for better conditions before a system becomes too ordinary to question.

Who Controls the Screen — schools decide which technologies belong in classrooms and who has control over them.
The Story About the Story — public concern about frontier AI requires more than choosing between fear and dismissal.