Chatbot that predicted why people share secrets with ChatGPT

In the 60’s Years of that Eliza It was influencing arithmetic and culture, and traditional accounts depict it as the closest example of what we now call it Chat botswho can speak like a robot psychologist. This deceptively simple program is known to “fool” even the secretary who watched Professor Joseph Weizenbaum of MIT create it. This is how the story goes.

However, in all of these accounts—even after all of their adaptations across programming languages ​​and research fields, in classrooms and popular culture—there was a key piece of the story missing: the source code of the ELIZA program itself. Our new book, Eliza’s inventionrecovers this source code from the MIT archives, and provides for the first time a close reading and discussion of this code along with newly discovered dialogs for ELIZA scripts beyond the famous “DOCTOR” character.

This investigation uncovered numerous ELIZAs: in its various software versions, designed to run a wide range of scripts or characters, and created using a series of technical innovations. Eliza’s invention It seeks to correct and complicate the history and impact of ELIZA by exploring the misconceptions, multiple versions, and missing code of ELIZA. In this excerpt from the book, we examine one of ELIZA’s earliest interactions, how it laid the foundation for human relationships with computers for decades to come, and how the complex software continues to speak to the unfettered engine of today’s AI industry.

All men both.

In any way

They are always teasing us about something or the other.

Can you think of a specific example?

Well, my friend made me come here.

Your friend made you come here

He says I feel depressed most of the time.

I’m sorry to hear you’re depressed

This dialogue was reprinted countless times and inspired programmers and writers to dream up many chatbots that followed. But the more we looked into that conversation, the more questions we asked: Who is this young woman? Was she a real person or was she an invention of ELIZA creator Joseph Weizenbaum? How exactly did the ELIZA system generate its responses, and how much was edited? Why does the system work so well to attract people?

Eliza and her “doctor” persona helped stimulate a pattern of thinking and anxiety about people’s relationships with computers. Weizenbaum discovered this in his 1976 book Computer power and the human mindand bring up philosophical, social and political criticisms. The unique automated interaction offered by his program revealed how new forms of human-computer relationship would have profound impacts that he attempted to explore and challenge. After Weizenbaum saw the public reception to it, he was astonished by the rapid and often passionate associations people would form with Eliza, which he saw as “clear evidence that people were talking to the computer as if it were a person who could be appropriately and usefully addressed in intimate terms.” The tendency to attribute empathy and invest private feelings in a computer puzzled Weisenbaum. He was interested in the extent to which people associate rationality and calculation, and attribute understanding and intelligence to computer systems that do not exist.

This trend became known as “Elisa effectBy 1991, the term was appearing in online forums, but its use preceded its appearance by decades. Sociologist Sherry Turkle He defines the “ELIZA effect” as “our more general tendency to treat responsive computer programs as more intelligent than they really are. Very small amounts of interaction cause us to project our complexity onto the undeserving being.” Cognitive and computer scientist Douglas Hofstadter describes it as “people’s ability to read and understand much more than is warranted in strings of symbols — especially words — strung together by computers,” which easily applies to generative AI systems today.

To understand the power and provocation of Eliza, we can look to the infamous challenge formulated by computer scientist Alan Turing in his article “Computing and intelligence machinesTuring posed the question “Can machines think?” Turing based his thought experiment on a parlor game – not about technology but about gender: A man and a woman are hidden in a separate room and the interrogator tries to determine the identity of which gender by asking a series of questions. The man tries to mislead the interrogator, pretending to be a woman, while the woman tries to persuade the interrogator with the “correct” answer. That is, they both claim to be the “real” woman, challenging essentialist notions about gender.