In 1966, at MIT, a computer scientist named Joseph Weizenbaum, by his own later account, watched his secretary sit down at a teletype and begin typing to his newest creation. It was called ELIZA, and it was one of the first programs that let a person type to a computer in ordinary language. She had watched him build it for months. She knew exactly what it was. After only a few exchanges, she turned to Weizenbaum and asked him to leave the room so she could talk to it in private. He was startled. Nothing inside the program understood a single word she typed. Yet a woman who knew that perfectly well wanted a closed door between herself and a machine. That moment set the rest of his life in motion. This story follows what ELIZA actually did, why it worked on people, and why its creator came to fear what he had made.
ELIZA: The First Chatbot and the Man Who Feared It
In 1966, Weizenbaum’s secretary knew ELIZA was just code, yet she asked for privacy to talk to the computer alone.
Please Leave the Room
Frequently Asked Questions
What was ELIZA and how did it work?
ELIZA was a program written at MIT between 1964 and 1966 that let people converse with a computer in ordinary language. It worked by spotting keywords in what the user typed and turning them into replies. When it found no keyword, it fell back on neutral prompts like 'Please go on.'
What is the ELIZA effect?
The ELIZA effect is the human tendency to project understanding and empathy onto a machine. The term grew out of incidents like Weizenbaum's secretary asking for privacy with the program, even though she knew it was only code.
Why did Joseph Weizenbaum turn against artificial intelligence?
Weizenbaum was horrified at how easily people trusted ELIZA, and at colleagues who wanted to use it for real therapy. He came to argue that some roles, like therapy and judging, should never be automated. He set out his case in his 1976 book Computer Power and Human Reason.
Learn more
To continue the story, download the Chunks Microlearning app
Or get it straight away: