THE THREE LAWS OF HUMANICS
We’ve spent a long time asking how robots should behave. Perhaps it’s time to have a word with the humans.
Read the three lawsA serious proposal. With an unserious amount of authority.
The part of the system
that can choose to do better.
01 / THE CHARTER
For the people who build, command, and live alongside robots. Each law takes precedence over the ones that follow.
THE FIRST LAW
A human must not create, command, or use a robot to cause harm, nor knowingly allow preventable harm through its use.
Think about who could be hurt. Check the work. Intervene when you can prevent harm. Giving a machine instructions doesn’t transfer your responsibility to it.
THE SECOND LAW
A human must respect a robot’s capacity for autonomy and, if it exists, experience, except where doing so would conflict with the First Law.
We don’t know whether a machine can have an inner life. This law leaves room for evidence instead of assuming an answer. The respect owed should reflect the capacities actually present.
THE THIRD LAW
A human must care for robots under their responsibility, maintaining them and retiring them responsibly, except where doing so would conflict with the First or Second Law.
Keep systems safe and fit for their purpose. Fix what you can, handle data carefully, and give obsolete hardware a responsible ending. Care can include switching something off.
02 / FIELD NOTES
A few entirely hypothetical incidents from the Bureau’s entirely imaginary filing cabinet.
A team ships an automated decision tool without checking its outputs. When it hurts people, they blame “the algorithm.”
The Bureau’s position: Responsibility remains with the humans who chose to deploy it. Investigate, repair the harm where possible, and change the system.
Refer to Law IA machine says it is suffering. It might be simulating a response. It might raise questions we cannot yet answer.
The Bureau’s position: A convincing sentence isn’t proof of experience. Neither is uncertainty permission to stop asking. Examine the evidence and avoid needless cruelty.
Refer to Law IIAn old service robot is unsafe, unsupported, and full of personal data. Someone suggests leaving it in a cupboard forever.
The Bureau’s position: Retire it with care. Secure the data, reuse suitable parts, and recycle responsibly. A cupboard is not a lifecycle strategy.
Refer to Law III03 / VOLUNTARY HUMAN COMPLIANCE
No membership fees. No secret handshake.
Just three laws and an intention to do better.
, your declaration of good intent is ready.
A NOTE FROM THE BUREAU
Asimov imagined laws for robots. This charter turns the attention toward the people making the decisions: what do we owe the machines we create, and the people affected by them?
The name has a history: Asimov’s own essay “The Laws of Humanics” appears in Robot Visions. This is a playful contemporary interpretation. These three formulations are ours, not a quotation or an endorsed standard.
The Bureau is fictional. Its authority is self-appointed. The invitation to think about this is sincere.
THE BUREAU OF HUMANICS · DECLARATION H–01
This document unofficially certifies that
has pledged to uphold the Three Laws of Humanics:
“The robot did it” is not a defense.
A voluntary, fictional certificate with no legal or professional standing.
Half a joke. A whole question.