The 1960s Chatbot That Explains Why You Tell ChatGPT Your Secrets

Key points

• ELIZA, created by MIT professor Joseph Weizenbaum in the 1960s, is widely considered the first chatbot, a computer program that holds a back-and-forth text conversation.
• The original source code was missing from public record for decades until researchers found it in the MIT Archives.
• Weizenbaum warned in his 1976 book that people were forming emotional attachments to ELIZA despite it having no real understanding.
• The "ELIZA effect," the human habit of treating responsive computer programs as more intelligent than they are, applies directly to modern AI tools like ChatGPT.
• ELIZA was deliberately named after Eliza Doolittle from Pygmalion, a character who learns to perform a social identity without truly becoming it.

Sixty years ago, an MIT professor built a simple computer program that pretended to be a therapist. People told it things they'd never told another person. Weizenbaum, its creator, was horrified.

That program was ELIZA, the great-grandparent of every chatbot you've used, and its story, first reported by Wired AI, has just got a lot more complete.

A new book called Inventing ELIZA has recovered the actual source code from the MIT Archives, offering the first close reading of the original instructions. For sixty years, accounts circulated without anyone seeing what was underneath. It's a bit like writing the history of the printing press without ever seeing one.

The book also scrutinises ELIZA's most famous exchange, reprinted in textbooks everywhere. A woman types: "Men are all alike." ELIZA replies: "IN WHAT WAY." She mentions her boyfriend, then depression. ELIZA says it's sorry to hear that. The exchange is simple, almost laughably so. The authors ask who this woman actually was, and whether Weizenbaum wrote her himself. Nobody's been sure.

What isn't in doubt is the effect the program had on real users. Weizenbaum watched people pour genuine feelings into a script built, by his own description, to conceal its lack of understanding. It stored almost nothing and mostly reflected questions back. Yet people treated it like a confidant.

Why do people keep opening up to machines?

Because a little interactivity goes a long way. Sociologist Sherry Turkle defines the "ELIZA effect" as our tendency to treat responsive computer programs as smarter than they actually are, projecting our own complexity onto an undeserving object. Cognitive scientist Douglas Hofstadter puts it differently: people read far more understanding than is warranted into words strung together by computers.

Here's the uncomfortable part. That description fits a 1960s parlour trick built from a few hundred lines of code, and it also fits a 2024 large language model, the type of deep-learning system that powers ChatGPT and Claude. The technology has shifted enormously. Human psychology hasn't budged much at all.

Weizenbaum named his program after Eliza Doolittle, the Pygmalion character who learns to perform upper-class speech without truly changing who she is. He said he chose the name because the program "could be taught to speak increasingly well" while it was "never quite clear whether or not it became smarter." That tension between performance and genuine understanding sits at the heart of every AI assistant sold to consumers today.

We first covered ELIZA's cultural footprint on 14 July 2026, when writer Lauren Oyler documented what it actually feels like to build a relationship with an AI. The new book gives that feeling a sixty-year backstory.

My read: the ELIZA story is being used to argue that nothing has fundamentally changed since 1966, which flatters neither the technology nor its critics. What has changed is scale and intimacy. ELIZA had a handful of users. ChatGPT has hundreds of millions, and it fits in their pocket at 2am.

If you use ChatGPT for personal advice or venting you'd normally save for a friend, check your AI tool's privacy policy first. Most free tiers use your conversations to train future models unless you opt out in settings.