The Relational Capacity Framework: A Lens for AI-Mediated Relationality
Organizations: University of Washington, USA · Google, USA · Google Research, USA
Abstract
A growing literature examines the impact of engaging with AI chatbots on our relationality: our ability to form relationships with others, maintain them through life's changes, and exit when needed. Currently, substantial work focuses on purpose-built AI companions (e.g. Replika, Character.AI), with less work investigating socioemotional uses of the more widely-used general-purpose chatbots. In this paper, we present a 100-person investigation into how people engage general-purpose AI for socioemotional needs. Participants' AI uses varied, from practicing specific social skills to seeking additional bandwidth and information as they navigated their social contexts. To explain this heterogeneity, we provide 8 states that describe how AI expands or contracts relationality for people with different levels of skills, bandwidth, and information. We then present a unified conceptual model for the role of technology in mediating human relationality, the Relational Capacity Framework, and outline future research designing appropriate roles for AI in human connection.
Figures & tables
| Category | n |
|---|---|
| Gender | |
| Woman | 49 |
| Man | 47 |
| Non-binary | 4 |
| Age Group | |
| 18–24 | 19 |
| Domain | Skill Type | Description |
|---|---|---|
| Communication | Adjusting tone or tactics to accommodate others or self-advocate. | |
| Interpretation | Understanding others’ thoughts, actions, and emotions | |
| Judgment | Gauging whether to begin, intensify, or downshift a relationship. | |
| Being vulnerable with others | Recognizing a need for help and reaching out to others. | |
| Interpersonal | Supporting others | Recognizing when others need help, and providing it effectively. |
| Self-soothing | Calming oneself from heightened states of emotion or distress. |
| Relationality State | Capacity | How AI use expands relational capacity | How AI use contracts relational capacity |
|---|---|---|---|
| RS1 | Low skill, low bandwidth, low information | AI directs user to improve capacity by practicing skills, seeking information, or expanding their bandwidth | User relies on AI to make decisions, replacing other relationships, resources, and skill-building |
| RS2 | High information, low bandwidth, low skill | AI helps users overcome momentary states of limited capacity | AI fulfills user’s request for short-term coping in the face of limited bandwidth, but never helps build durable skills or address bandwidth constraints |
| RS3 | High bandwidth, low information, low skill | AI directs user’s energy to better appraising and addressing their situation | Lacking context, AI enables user to ruminate on available details of their situation, drawing resources away from long-term skill-building |
| RS4 | High skill, low bandwidth, low information | AI supplies the information necessary for a user to execute their skills | AI hallucinates or misinforms, and users do not have the time or information to correct it |
| RS5 | High bandwidth, high information, low skill | AI provides rehearsal environments where users can practice their skills | AI is used to practice skill in low-stakes environments, but never helps the user build the confidence to execute in higher-stakes contexts |
| RS6 | High skill, high bandwidth, low information | Users leverage AI to better understand the context necessary to apply their skills | AI gives information that is too general or not accurate to the specific context a user is trying to operate within |
Human-AI Interaction
Human-AI interaction studies how people and AI systems communicate, collaborate, make decisions, and shape one another over time.
See all 354 papers