Meet the Creator of the World's First Chatbot ELIZA, Who Warned that AI can Never Replace Humans

Long before ChatGPT, Gemini, Claude and virtual assistants a simple computer program created in an MIT laboratory captivated the world. It was built in the mid 1960’s called as ELIZA. ELIZA is recognised as the world’s first chatbot. 

ELIZA was created by Joseph Weizenbaum. He was a pioneer in computer science. He also warned people how easily people were fooled by his invention. He spent the rest of his life campaigning against the overuse of artificial intelligence.

Who Was Joseph Weizenbaum?

Joseph Weizenbaum was born in Berlin in 1923. He fled from Nazi Germany with his family in 1936 and settled in the United States. He studied mathematics and served in the U.S. Army. Later he took an interest in the emerging field of computer science. 

Weizenbaum joined the faculty at the Massachusetts Institute of Technology (MIT) in the 1960s. MIT was one of the world's premier institutions for computer research. 

It was during his time at MIT that he created ELIZA.  It was an experiment designed to explore how humans and machines communicate using natural language.

The World’s First Chatbot: ELIZA

Weizenbaum developed ELIZA on an IBM 7094 mainframe between 1964 and 1966. 

The program was named after Eliza Doolittle. The working-class character in George Bernard Shaw’s play Pygmalion who was taught to speak with upper-class eloquence.

Weizenbaum loaded ELIZA with a script called DOCTOR to demonstrate the superficial nature of machine communication. This script mimicked a Rogerian psychotherapist. A style of therapy that largely reflects a patient's own words to them in the form of open ended questions.

What was the ELIZA Effect?

Weizenbaum's objective was to make ELIZA show how limited machine language process in reality and instantly recognise the program was simply rearranging their words without understanding what they meant.

Instead, the opposite happened. People formed deep emotional connections with the chatbot. They revealed secrets, sought advice and treated a computer program as if it possessed genuine empathy and consciousness.

The turning point for Weizenbaum was when his secretary tested ELIZA after just a few typed exchanges. She turned to him and asked ‘Would you mind leaving the room, please? She wanted to talk to the machine in private. 

That realisation shocked him. The tendency of human nature to project real thoughts, feeling and understanding onto software. It is known as the ELIZA effect in computer science. 

Weizenbaum warning against AI replacing humans

Weizenbaum changed his views after seeing how easily people could be tricked by primitive code. He transitioned from an AI developer into one of the technology's most vocal critics.

In his book, ‘Computer Power and Human Reason: From Judgment to Calculation’. He gave fundamental principles for modern AI ethics.  

Weizenbaum argued that computers can perform calculations such as processing logic, data, and probability. However, they can never exercise judgment. True judgment requires human experience, empathy, morals and emotions. These qualities code cannot replicate by AI

He argued that certain roles such as therapy, teaching and judging legal cases or caring for the sick. All these require genuine human connection. Using AI as a human substitute in care in these fields degrades what it means to be human.

Weizenbaum warned that society would become lazy. Handing critical decision making over to algorithms because machines appear objective, fast and emotionally detached.