Pink noise in dyads' semantic similarity trajectories predicts conversation enjoyment.

Pink-noise organization in dyads’ semantic similarity trajectories is present and its strength predicts higher conversation enjoyment and connection, with effects linked to practice.

• Why it matters: Linear semantic entrainment models fail to explain why perfectly aligned discussion can become boring, motivating a multiscale account of how similarity evolves over time.
• What they did: The study analyzed interlocutors’ conavigated semantic-space trajectories to detect pink noise, then tested whether its strength related to conversation outcomes and whether practice increased it.
• The result: Pink noise was found in these trajectories; stronger pink-noise signals predicted enjoyment and connection, grew with practice, and matched patterns in a large language model’s conversation-quality representation.

AI summary of the abstract · may contain errors · not medical advice · read the paper.