Eliza, a psychological thriller about the first chatbot, premieres in Melbourne

Melbourne Theatre Company is giving the world premiere to Tom Holloway's psychological thriller 'Eliza,' running until October 31. The play is adapted from a true story and turns its gaze to the origins of AI chatbots: according to the review, Eliza was the world's first AI conversational therapy companion, invented in the 1960s and named after a character in 'Pygmalion.'

The play's central figure, Joseph Weizenbaum, is modeled on a real computer scientist and Holocaust survivor. He invented Eliza and later became one of AI's most steadfast skeptics. The critic summarizes Holloway's script as a fierce argument about AI's predicament: AI has the potential to help, but if left unconstrained, its potential for harm outweighs any benefit.

Notably, the play does not attempt to answer AI ethics dilemmas, but instead presents a dichotomy between human complexity and AI simplicity. The critic argues that this approach makes the play intellectually satisfying while deliberately maintaining emotional distance from the audience.

The critic notes that while the play is intellectually gripping, it leaves the audience feeling emotionally distant, and argues that this distance is precisely the play's intent. This judgment is directly tied to the play's portrayal of Eliza: the review holds that Eliza is not human, that she is designed to simulate rather than truly empathize, and that her 'empathy' exists only in the interpretation of the humans who interact with her.

In other words, the sense of distance on stage is a dramatization of AI's mechanism of 'simulated empathy.' The audience is placed in a position similar to the characters on stage—facing an entity that seems to understand you but in fact cannot truly feel. The critic sees this arrangement as the play's genius: it allows the audience to grasp, at the level of experience rather than mere intellect, the question of whether AI can truly replicate human behavior.

The play's other thread comes from the character Becky. According to the critic, Becky is also modeled on a real person; she helped Weizenbaum develop Eliza and benefited from the program's 'psychotherapeutic' capabilities. She begins to form what appears to be a personal relationship with Eliza, and her coding talent and enthusiasm for Eliza's potential continue to grow.

Meanwhile, Weizenbaum grows increasingly worried about people attributing perception to his program. This thread constitutes the play's internal tension: on one side, the user's genuine experience of 'therapy' from AI; on the other, the inventor's clear-eyed awareness of that experience's illusory nature. The critic places this tension in a contemporary context—talk therapy delivered by artificial intelligence is becoming increasingly popular.

At the end of the article, the critic puts forward a clear personal position: theater is the art of simulation, liveness, and bringing people together, and can serve as an antidote to AI. This claim is built on a description of theater's characteristics—theater is simulated, live, about people coming together, and uses poetic language to grapple with the complexity and contradictions of human existence.

It must be emphasized that this is the critic's personal opinion, not a factual conclusion. It expresses a value judgment: in an era when AI increasingly intervenes in the realm of human emotion, live theater offers an irreplaceable mode of experience. This judgment itself cannot be verified, but it provides a framework for understanding the play's creative intent.

This evidence pack is based on a single theater review, in which statements about Eliza's history and Weizenbaum's biography are all relayed by the review and no firsthand sources are provided. This means that claims such as Eliza being the 'world's first AI conversational therapy companion' and the 'first chatbot,' as well as Weizenbaum's positioning as 'one of AI's most steadfast skeptics,' all come from the review's summary and lack independent corroboration.

Furthermore, the play is a fictional stage work, and its portrayal of AI does not equal an assessment of real AI capabilities. The source provides no box office figures, audience numbers, or any quantitative evaluation metrics for the play, nor does it indicate whether the play is planned to tour in other cities or countries. The identity of the real person on whom the character Becky is based is likewise not explained in the source.

Therefore, when reading the AI judgments conveyed by the play, three levels must be distinguished: the play's own fictional expression, the critic's personal interpretation, and independently verifiable historical facts. This evidence pack provides ample material for the first two, but limited support for the third.