This paper critiques the prevailing mimetic paradigm in the design of empathetic AI, which aims to simulate human emotion. We argue that this approach is fundamentally limited by AI's lack of embodied, lived experience, leading to inauthentic interactions and ethical dilemmas of deception. As an alternative, we propose a functionalist theory of "Synthetic Emotion", a non-anthropocentric framework that defines AI emotionality not as a failed copy of human feeling, but as a distinct computational system operating on its own principles of contextual logic, pattern analysis, and pragmatic adaptation. By synthesizing research from affective computing, human-computer interaction, and phenomenology, we outline the architecture of this system, propose methods for its evaluation based on perceived interaction quality rather than feigned authenticity, and explore its ethical implications. We conclude that the future of empathetic AI lies not in building machines that pretend to feel, but in designing functional systems that are coherent, supportive, and transparent in their operations, thereby fostering more honest and sustainable human-AI relationships.

Authors:
David Leonardo Barsand de Leucas, Federal University of São Carlos, Brazil
Sylvia Iasulaitis, Federal University of São Carlos, Brazil

---

Conference Comments & Feedback

Place a comment using your LinkedIn profile

Comments

Share on activity feed

Post with LinkedIn

Powered by WP LinkPress

Share this Presentation

Share this...

Facebook

Pinterest

Twitter

Linkedin

Posted by James Alexander Gordon

Last updated: 2023-02-23 23:45:00

All Posts