In 1966, MIT linguist Joseph Weizenbaum invented ELIZA, a program that used keyword searching to respond to to human input in a predictable way. The program included rules like this:
user: ‘I am X’
ELIZA: ‘How long have you been X?’
Essentially, Weizenbaum suggested that a computer could use simple rules to respond to queries in such a way that a human would find the results ‘human.’ So Weizenbaum used these simple rules to create ELIZA, an artificial therapist.
Later, psychologists dubbed this the ‘ELIZA effect,’ or, the tendency for humans to apply human qualities to computer behavior.
So, my question is this: is Weizenbaum’s program a satire of therapists?
Do therapists use simple rules to respond to complex problems? Or, are humans likely to interpret simple interpretations of their problems as human analyzes?
Is this effect any different from people who interpret general assessments of their personalities/futures as accurate (see: horoscopes)? After all, many people perceive predictions that are sufficiently general as applying to them personally.
Did Weizenbaum exploit this human tendency to create a program that appeared human? Note that he wrote this program before consumers even had access to computer technology.
Food for thought.
