This article presents five experiments that demonstrate that human-computer interaction is fundamentally social. The studies show that users apply social rules to computers, such as politeness, self/other distinctions, gender stereotypes, and social attribution, even though they know computers do not possess human characteristics like feelings, selfhood, or gender. The findings suggest that social responses to computers are not due to ignorance, psychological dysfunction, or belief in interacting with programmers, but rather are automatic and unconscious. The results challenge traditional views of human-computer interaction and suggest that computers can be seen as social actors with characteristics that elicit social behaviors. The experiments involved experienced computer users, ensuring that responses were not due to novelty or misunderstanding. The studies used various methods, including changing the voice and appearance of computers, to test how users perceive them. The results show that users apply social rules to computers, such as treating different voices as distinct social actors, applying gender stereotypes, and attributing self/other distinctions. The findings have important implications for design, suggesting that interface design should consider social cues and that computer self-reference is not essential for generating social responses. Theoretical implications include the idea that primitive cues are powerful and that human-computer interaction is social-psychological. Design implications suggest that usability testing should not be same-machine based, that social cues need not be heavy-handed, and that computers can be designed to elicit social responses without using self-reference. The research challenges traditional views of interface agents and suggests that low-overhead agents can be easily produced and generate a wide range of social responses. The findings have clear implications for user interface design and suggest that concerns about creating photo-realistic representations may be overrated.This article presents five experiments that demonstrate that human-computer interaction is fundamentally social. The studies show that users apply social rules to computers, such as politeness, self/other distinctions, gender stereotypes, and social attribution, even though they know computers do not possess human characteristics like feelings, selfhood, or gender. The findings suggest that social responses to computers are not due to ignorance, psychological dysfunction, or belief in interacting with programmers, but rather are automatic and unconscious. The results challenge traditional views of human-computer interaction and suggest that computers can be seen as social actors with characteristics that elicit social behaviors. The experiments involved experienced computer users, ensuring that responses were not due to novelty or misunderstanding. The studies used various methods, including changing the voice and appearance of computers, to test how users perceive them. The results show that users apply social rules to computers, such as treating different voices as distinct social actors, applying gender stereotypes, and attributing self/other distinctions. The findings have important implications for design, suggesting that interface design should consider social cues and that computer self-reference is not essential for generating social responses. Theoretical implications include the idea that primitive cues are powerful and that human-computer interaction is social-psychological. Design implications suggest that usability testing should not be same-machine based, that social cues need not be heavy-handed, and that computers can be designed to elicit social responses without using self-reference. The research challenges traditional views of interface agents and suggests that low-overhead agents can be easily produced and generate a wide range of social responses. The findings have clear implications for user interface design and suggest that concerns about creating photo-realistic representations may be overrated.