A camera feature can create very different privacy expectations depending on whether it waits for a deliberate action or continuously observes. Oneira Sight uses the first model. The owner chooses a moment, selects a bounded mode, and sends one frame for one short response.
Continuous sensing would change the product
An always-running camera would create ongoing capture, bystander awareness, storage, power, security, and policy questions. Those are not small implementation details. By avoiding a background feed, Sight keeps the camera event more legible and reduces the amount of visual data needed for each interaction.
Deliberate does not mean consequence-free
One frame can still contain a face, private screen, address, key, badge, document, or sensitive location. The owner-triggered design helps create a boundary but does not make every subject appropriate. Consent, venue rules, and careful framing remain necessary before the button is pressed.
The distinction should remain visible in the interface
A useful camera indicator, mode label, and short disclosure help the owner understand the request. Product updates should not silently change a deliberate-capture feature into passive monitoring. If the processing model changes materially, the policy and consent flow should change before the new behavior is used.
The short version
- Sight sends one frame after an owner action, not continuous video.
- A deliberate frame can still contain sensitive information.
- A material processing change should require an updated disclosure and consent flow.