The Extended Turing Test
The extended test facilitates its use for scientific purposes by introducing three innovations (described in the detailed text) designed to force the judges to focus on the specific dimension being evaluated.
Robert Abelson, 1968.

The three innovations introduced by the extended test are:

  1. The judges are not informed that a computer is involved in the test at all. They are told simply to compare the performance of two entrants, with respect to some dimension (e.g. womanliness or paranoid behaviour, etc).
  2. The computer does not participate on its own but (unbeknownst to the judges) shares time with a human subject. The two are 'switched off' periodically throughout the test.
  3. The judges score the human-computer pair after every question, so that separate scores can be generated for each.
Immediately related elementsHow this works
-
Artificial Intelligence »Artificial Intelligence
Can the Turing Test determine this? [2]  »Can the Turing Test determine this? [2] 
No: existing AI programs have passed the test »No: existing AI programs have passed the test
PARRY »PARRY
The Extended Turing Test
+Komentarai (0)
+Citavimą (0)
+About