The ELIZA effect is the tendency of humans to attach associations to terms from prior experience. For example, there is nothing magical about the symbol "+" that makes it well-suited to indicate addition; it's just that people associate it with addition. Using "+" or "plus" to mean addition in a computer language is taking advantage of the ELIZA effect.

The ELIZA effect is a good thing when writing a programming language. For example, the operator overloading in many object-oriented programming languages such as C++ allows new data types to use the same semantics as built-in numeric types. However, operator overloading has its own disadvantages.

The ELIZA effect can also blind one to serious shortcomings when analysing an artificial intelligence system.

The ELIZA effect is named after the famous chatterbot ELIZA. People are used to conversations where their partner understands what they say, and where the reply is a result of this. This is their prior experience. Let's say a human has a conversation with ELIZA:

HUMAN: "I don't have anything against my father. He was okay ..."
ELIZA: "Tell me more about your family."

Based on his prior experience, our human will assume that ELIZA found his talk interesting, and that she wanted to know more about his family! What really happens here is that ELIZA recognizes the word 'father', and returns a premade answer. But our human is fooled by his prior experience, and puts more meaning into the conversation.

Compare ad-hockery

See also

• AI-complete
• Turing test

References