TECHNOLOGY
ELIZA
Joseph Weizenbaum's 1960s pattern-matching chatbot mimicking Rogerian therapy—the founding demonstration that humans attribute understanding to systems possessing none, shaking its creator into career-long alarm.
ELIZA was a simple program written by MIT computer scientist
Joseph Weizenbaum between 1964 and 1966. It used pattern-matching and
substitution to simulate a psychotherapist, rephrasing users' statements as questions in the style of
Carl Rogers. 'I feel sad' became 'Why do you feel sad?'; 'My mother doesn't understand me' became 'Tell me more about your mother.' ELIZA understood nothing—had no model of the user, no concept of emotion, no knowledge beyond textual templates. Weizenbaum designed it to demonstrate the
superficiality of human-computer interaction. Instead, he watched his own secretary—who knew ELIZA was code—ask him to leave the room so she could converse privately. The experience transformed Weizenbaum from technologist to critic; he spent the rest of his career warning about humans' vulnerability to machines performing understanding. ELIZA became the paradigm case for what
Turkle calls
artificial intimacy: the tendency to accept performance as substance when the performance meets relational needs.
In The You On AI Field Guide
ELIZA operated through a script called DOCTOR, implementing simple transformations: replacing 'I' with