The thing I keep circling is getting an intention to land as someone else’s perception of it. What actually sets your audience’s posture is the texture around what gets said, and that is the part missing from AI memory and conversation right now.
We are not orienting our agents to adopt the right posture. We hand them the task and almost none of the orientation. It is the same reason that, when a person will not match your posture, you feel disconnected: they are not holding the idea the way the conversation needs them to for it to move forward. Texture is the layer that would let an agent do what a good listener does, take the right posture before the words even land.