How a 1966 AI experiment predicted today's technology debates
In 1966, computer scientist Joseph Weizenbaum created ELIZA to prove AI had limits, but users treated it as a genuine companion instead.
Joseph Weizenbaum created ELIZA at MIT in 1966, a simple program that only repeated back what users told it, phrased as a question. The program was designed to demonstrate that AI couldn't truly understand humans or connect with them; it was just 400 lines of code.
Yet some users, including Weizenbaum's secretary, became emotionally attached and treated ELIZA almost like a companion. Weizenbaum's warning went largely unheeded.
Now AI is woven throughout daily life, raising concerns about copyright, environmental impact, and safety. Playwright Tom Holloway's stage play Eliza revisits Weizenbaum's story, exploring why people misunderstand technology even when experts warn them.
Sydney Water flagged that data centres training AI models could use the equivalent of a quarter of the city's annual drinking water supply within a decade. Holloway argues the real problem is not the technology itself but those controlling it and the lack of regulation.
- 1966
- Year created
- 400 lines of code
- Size of program
- MIT
- Institution
- Quarter of Sydney's drinking water
- Annual water impact
Why it mattersWeizenbaum's failed warning illustrates how society can overlook expert cautions about technology, a pattern that repeats as AI's environmental and social costs mount.
AustraliaAustralian water and energy resources face growing pressure from AI data centres; Sydney Water has warned of acute water scarcity impacts from computing infrastructure expansion.
Open this story in InSnip →
