The Eliza Effect
Updated 2026-08-08
INTRODUCTION
English translation pending.
CORE DEFINITION
Named after ELIZA, Joseph Weizenbaum's chatbot, which used simple pattern matching to mimic a Rogerian therapist. Weizenbaum was disturbed to find that users, including his own secretary, confided in the program and asked him to leave the room, despite knowing it was a program. The effect describes the human readiness to read comprehension and intention into any system that produces conversationally appropriate output. It remains a central concern in human-computer interaction and in discussions of artificial intelligence alignment and overtrust.
SCAFFOLDING EFFECT
Reduce cognitive load
- Anthropomorphism check: ask whether you are responding to the system's output or to an intention you have projected onto it. - Capability boundary: separate fluent language from actual comprehension, and never let fluency imply reliability. - Safety rule: never let a decision depend on the impression that the system understands what is at stake.
Anchor fast decisions
Humans run a fast automatic intention detector that treats conversational signals as evidence of a mind behind them. Fluent and contextually appropriate language is exactly the kind of signal that triggers this attribution, so it fires whether or not any understanding is present. Once a mind has been attributed, social norms such as politeness, trust, and deference follow automatically, which is why the effect survives explicit knowledge that the interlocutor is a program.
MINIMUM ACTION
In progress 0/1Practice this model in one real situation:
account_treeGenealogyexpand_more
menu_bookReferencesexpand_more
Source support: Explicit
- en.wikipedia.orghttps://en.wikipedia.org/wiki/ELIZA_effectverified
PRIVATE NOTES · Only visible to you
SAVED Q&A
ENTRY Q&A · Private saving available
Ask with a clear boundary
thinkingmodels answers from published entry context only.
Your question is sent to thinkingmodels. The answer uses public entry context only.
RELATED MODELS