Why does one strong human-like cue, like a voice, make an AI feel 'present' when a pile of weaker cues doesn't add up to the same thing?

This explores why certain cues, like a human voice or a human-like appearance, make people respond to an AI as if it were a social actor, while a stack of lesser cues often doesn't. The clearest finding in the collection is that quality beats quantity: one primary cue such as voice or appearance is enough to make an AI feel socially present, and several secondary cues added together still can't do the same job Do more social cues always make AI feel more present?. Social presence doesn't build up gradually as signals accumulate. It behaves more like a switch, and only certain cues can flip it.

A useful way to see why comes from a distinction the field has tended to blur. Van der Goot and Etzrodt separate two ways people respond to machines Do people respond to machines mindlessly or reflectively?. The first is automatic: a minimal cue triggers a social reaction without any conscious decision, which is the classic 'Media Equation' idea that people are polite to computers. The second is reflective: people stop and puzzle over what kind of thing the machine is, because it doesn't fit neatly into 'person' or 'object.' Read together with the cue finding, primary cues look like the ones that set off the fast, automatic route. A voice is something our social instincts react to before we think. Secondary cues seem more likely to feed the slower, deliberate route, where people weigh evidence instead of simply reacting.

Secondary-style behaviors still matter, but they work differently. Chatbots that consistently shared emotions led users to open up more in return, which follows the human norm of reciprocating vulnerability Do chatbots trigger human reciprocity norms around self-disclosure?. Interestingly, consistent sharing worked better than adaptive matching. That suggests behavioral cues gain their force from a steady pattern built over the course of a conversation, while a voice works in an instant. Primary cues create the sense that someone is there. Secondary, behavioral cues shape how the relationship goes once that sense exists.

A further point is that a primary cue like voice isn't one fixed signal. The same acoustic features that come across as extraversion in a calm interview come across as neuroticism under stress Does personality sound the same in stressful and neutral conversations?. A voice reliably triggers a social response, but which social response depends on the situation it's heard in. Designers can't treat 'adding a voice' as a neutral feature.

The collection is thin on direct head-to-head comparisons of primary and secondary cues. Beyond the one sufficiency finding, there's little on why the switch works the way it does or where it breaks down. The pieces fit together, though: primary cues get you treated as a someone almost instantly, and secondary cues decide what kind of someone over time.