Proposed by famed scientist and mathematician Alan Turing, this test aims to provide a clear cut definition of Intelligence, specifically for machines and/or computer programs.
The test is as following: on one side are 2 respondents, one machine and the other human. On the other side is a human acting as an interrogator. The machine pass the test if the interrogator can not tell the difference between the two.
This test require high level of NLP, Knowledge representation, Reasoning, and ML. Which is beginning to be achievable for modern AI programs like ChatGPT by OpenAI
One of my favorite quote about the Turing test is: "Any AI smart enough to pass a Turing test is smart enough to know to fail it."
Top comments (0)