Towards a Comprehensive Intelligence Test (TCIT): Reconsidering the Turing Test for the 21st Century Symposium

 
 

2010 marks the 60th anniversary of the publication of Turing’s paper in which he outlined the Turing Test for machine intelligence. In his paper, Alan Turing suggested that the philosophical possibility of genuine machine thought should be replaced by a simple behaviour based process in which a human interrogator converses blindly with two entities: a machine and another human. Although the precise nature of Turing's test has been debated, the standard interpretation is that if, after five minutes interaction, the interrogator can not reliably tell which respondent is the human and which the machine then the machine can truly said to be a 'thinking machine'. Across the intervening years this test has become synonymous as 'the benchmark' for Artificial Intelligence in popular culture, however new advances in cognitive sciences and consciousness studies suggest it may be useful to revisit this test. During a recent debate at AISB09 [SSAISB AGM, Wed. 8th April 2009 at Heriot-Watt University, Edinburgh], it became clear that, whilst the Turing Test has served well in driving the research agenda in AI forward, it should no longer serve as a true test of intelligence.


The aim of this symposium is to reconsider the Turing Test in the light of current advances in Artificial Intelligence, cognitive systems, and other competitions that provide insights into different types of intelligence, with the goal of outlining a new test - or suite of tests - that may more usefully be employed to evaluate 'machine intelligence' at the dawn of the 21st century.

 

About the Symposium