Self-awareness has a blind spot: the parts of your behaviour you cannot observe because you are too close to them. Human Mirror is designed to surface those patterns — with evidence, and without pretending to know you better than you know yourself.
Most people can describe how they intend to communicate. Fewer can describe how they actually come across under time pressure, disagreement, or uncertainty. That gap is not a character flaw — it is simply hard to watch yourself while you are talking.
A conversational observer can register response latency, how you handle a vague question, whether you expand or deflect, how you signal disagreement, and how you recover after being corrected. None of these are diagnoses. They are interaction signals.
The value is not a number. It is being shown a specific, evidenced pattern you can then confirm, reject, or refine. That is why every dimension in your Human Model carries evidence, a confidence score, and an alternative explanation.
Q: Is this the same as a personality test?
A: No. A personality test asks you to rate yourself and returns a type. Human Mirror observes a conversation and returns behavioural inferences with confidence and evidence, which you can correct.
Q: Could the AI be wrong about me?
A: Yes, and the product assumes it. Every inference carries a confidence score and at least one alternative explanation, and you can mark any trait as wrong — which regenerates the model.
Q: Does it analyse my face to detect emotions?
A: With camera consent it derives coarse interaction signals such as attention and speaking turns. It does not diagnose emotional or mental states, and it does not infer race, religion, politics, health, or sexuality.