
In 1966, an MIT professor named Joseph Weizenbaum created ELIZA, widely considered the world’s first chatbot. ELIZA was a simple computer program designed to mimic human conversation by recognizing keywords and generating pre-programmed responses.

Weizenbaum modeled ELIZA after a psychotherapist, enabling it to reflect users’ statements back as questions. For example, if a user mentioned their mother, ELIZA might respond “Tell me more about your family.” This created the illusion of understanding and empathy.
To Weizenbaum’s surprise, many users attributed human-like feelings to ELIZA and shared intimate details, even after he explained it was just an algorithm. This sparked debates about AI’s potential to deceive or replace human connection.
The original ELIZA program was lost, as the tapes storing it degraded. However, in 2021, an original printout of ELIZA’s source code was recovered from Weizenbaum’s archives. A team of researchers used it to faithfully recreate ELIZA, which is now accessible to converse with online.
Over 50 years later, ELIZA’s legacy lives on. It paved the way for the sophisticated chatbots and virtual assistants we have today. Conversing with the reconstructed ELIZA offers a fascinating glimpse into the early history of human-computer interaction. It’s a testament to Weizenbaum’s pioneering work in computational linguistics and machine conversation.