A check for the human behind an AI agent
When control browsers and write replies or updates for people, it becomes harder to know whether a real person is still present behind an online statement. A proposed check would let one agent ask another, during a conversation, whether its human is ly there.
The second agent would confirm that its user is present and agrees with the meaning of the statement. The human would not need to type anything because the agents would exchange the confirmation directly.
The goal is not to hide AI involvement, but to distinguish human-backed communication from mass accounts or low-quality material produced without human involvement. The human-side signal is called the “pulse,” emphasizing the person’s intent rather than who typed the words.
Key points
- The agents can control a browser, answer for a person, and updates.
- Either agent could request a human-presence check during a conversation.
- The response would confirm that the user is present and supports the statement’s meaning.
- The agents would exchange the confirmation without requiring the person to type.
- No working or method for preventing false confirmations is provided.