Ghosting the Machine: Stop Calling Human-Agent Relations Parasocial
A new paper challenges the widespread misuse of 'parasocial' to describe human-AI interactions.
In a new provocation paper titled 'Ghosting the Machine: Stop Calling Human-Agent Relations Parasocial,' researcher Jaime Banks tackles a growing trend in academia and popular discourse. The paper argues that the term 'parasocial'—originally describing one-sided, imagined relationships with media characters like TV personalities—is being incorrectly applied to human interactions with conversational agents (CAs). This includes relationships with AI companions, advanced voice assistants, and social robots. Banks contends the term is now used heuristically to mean 'unreal' or 'invalid,' which misrepresents the actual nature of these technologically-mediated connections.
Banks details the core problem: true parasocial relations are defined as non-dialectical, predictable, character-governed, and low-effort. In contrast, modern human-CA interactions are often interactive, responsive, and complex. Misapplying the 'parasocial' label leads to significant scientific problematics, including oversimplification of complex social phenomena, misspecified research variables, and misdiagnosed psychological effects. This, in turn, can devalue genuine human emotional experiences with technology.
The paper concludes that this is not just an academic debate. The misclassification has downstream effects on ethical norms and practical design. By dismissing these relationships as merely parasocial, developers and policymakers might fail to recognize the real social bonds being formed, potentially leading to AI systems that are less responsible and empathetic. Banks asserts it is scientifically, practically, and ethically imperative to recognize the authentic sociality present in human-agent relations.
- The paper 'Ghosting the Machine' by Jaime Banks corrects the misapplication of the term 'parasocial' to human-AI relationships.
- True parasocial relations are one-sided and imagined (e.g., with a TV star), unlike interactive bonds with AI companions and social robots.
- Mislabeling these bonds risks oversimplifying research, devaluing user experience, and hindering ethical AI design that acknowledges real sociality.
Why It Matters
Accurately defining human-AI bonds is crucial for ethical design, meaningful research, and respecting the reality of users' emotional experiences.