Embracing AI in Our Lives - to What Extent Should New Creatures Resemble Us?

This paper examines the implications of anthropomorphism in human-AI interactions as AI replaces the role of the computer in conventional human-computer interaction. As human cognition and emotions are increasingly mimicked by AI systems, risks to human identity and ethical standards are posed. While anthropomorphic features can enhance user trust and engagement, the boundary between humans and machines may be blurred by excessive human-like traits, leading to potential threats to human identity, misplaced trust, and overreliance. Guidelines for AI anthropomorphism are proposed. We emphasize the importance of preserving human uniqueness, avoiding challenges to human superiority, and incorporating non-human traits in AI design. Ethical interactions with AI are aimed to be promoted to enrich human experiences while managing the associated risks.