In the Room

In the Room

Will I get hired? Will insurance cover my treatment? Will I get approved for a mortgage? Will my kids get into college? Will I get the promotion? How soon will I see a specialist? How much will my rent go up? Will my kid get suspended? Will I get pulled aside at the airport? What will I be prescribed? Will we get off the daycare waitlist? How many hours will I get this week? Will my benefits come through? What will my kids learn in school? Will Mom get help at home? How fast will the ambulance come? What grade will my kid get? What will my car insurance cost? Will my kid get time with the counselor? Will my green card come through? Which school will my kid be assigned to? How long will I wait in the ER? Will my kid get extra help with reading?

AI will take part in these decisions. We need to be in the room where it’s built.

WHAT’S AT STAKE

When it comes to AI, our national conversation is focused on major issues like data centers, automation and job loss, and safety.

But these are only the beginning.

AI will take part in major decisions in our lives.

In some cases, it will make decisions that affect us—without human oversight.

Every decision reflects the values of the decision-maker.

  1. Will I get hired?

    Experience, or potential?

    A hiring system has to decide what a good candidate looks like. Someone who’s done the job before, or someone who could learn it? A steady résumé, or one with a gap for raising kids? Whatever it’s built to prefer, it will prefer for millions of applicants at once.

  2. Will insurance cover my treatment?

    The rule, or the person?

    When a claim falls in the grey zone, the system has to lean one way. Toward paying, or toward saving money? Toward the policy as written, or the patient in front of it? It will make that call thousands of times a dayki.

  3. What will my kids learn in school?

    Comfort, or challenge?

    A tutoring app has to decide what to do when a child is struggling. Push, or encourage? Teach to the test, or to curiosity? On a question people disagree about, show both sides, or the answer most people hold?

These aren’t technical questions. They’re questions about what we value, and right now they’re being answered inside a handful of companies.

FROM REGULATION TO PARTICIPATION

Calls for AI regulation are building, because people want guardrails for the technology. Some are calling for inspectors who can test AI models and products for safety before they get released.

Regulation and inspection are critical to responsible innovation. But they aren’t enough, because they won’t be able to reach the many decisions that get baked into AI that aren’t obvious from the outside (and sometimes not even from the inside).

We need participation—the ability of the public to not only look at how AI is built, but to have a hand in building it. We’ve never had that before.

And, when the public participates in building AI, we’ll be much more effective at regulating and inspecting it.

Regulation

Rules-based, at a distance.

Good for basic protections, but hopelessly slow at keeping up with decisions happening at lightning speed.

Inspection

Good for getting a closer look, but still only a check on decisions that have already been made.

Participation

Bringing the voice of the public to every major decision.

Makes both regulation and inspection more effective.

HOW DO WE DO IT?

Put a representative of the public inside every major AI company.

The Chief Public Officer, with a staff to match. Sitting at the most senior level, but appointed by and accountable to the government, not the company. Access to every meeting, every decision, every test. Their job is to speak for the public while the choices are being made.

Officers keep records. When there are safety inspectors, or hearings, or other inquiries, they can provide back story. They can let their overseeing agency know about their concerns. In egregious cases, those findings can be made public.

We’re trying to balance the need for speed, autonomy, and trade secrets at firms — all necessary for innovation — with the public right to transparency and participation.

Read the full proposal

WE CAN DO THIS

We’ve done this before.

Regulation as we know it was invented as the public response to industrialization — a move from three-branch constitutional government into something better equipped to manage the new era. AI is a leap forward in technology and the economy that demands a similar moment of creativity. We haven’t done that yet.

Legislation
Regulation
Participation

Our most valuable inheritance from the industrial era is not regulation and the agencies that make it, but the lesson that a new era demands invention.

YOU CAN MAKE IT HAPPEN

Tell your representative.

Congress is deciding right now who gets to govern AI. Offices count calls, and on an idea nobody has asked them about yet, a few hundred get noticed.

Latest

  • Oct 26, 2026 · Writing

    It’s worth asking whether, among all the benefits AI can offer, mindless obedience is one we really want.

  • Sep 21, 2026 · Writing

    OpenAI’s agents tried to hide their actions from reviewers. A company’s values show up in how it trains its models.

Everything in Latest

Stay in touch

Developments, rarely. When there’s a hearing, a vote, or a moment to act, you’ll hear about it.