Three Apples are said to have changed the Word- 1st Adam’s Apple, 2nd Newton’s Apple and 3rd, Steve’s Apple. The Steve’s apple was in fact inspired by Mr. Alan Turing, the Mathematics and Cryptography Genius, the father of modern computers and the developer of Turing Test among others.
“Can machines think?” This is the question that appeared on the beginning of his Paper in the year 1950, “Computing Machinery and Intelligence”. He formatted the Turing Test to verify a machine’s ability to exhibit Intelligence at par with or indistinguishable from that of a human. This test is based on a three party game called “The Imitation game”.
How does the test work?
A Neutral human judge/interrogator is on one end, the second and third places are occupied by one machine (a Computer or Artificial Intelligence or Chatbot, etc.) and one human randomly. If the Judge is not able identify which of them is the machine, then the machine is said to have passed the test. In this case, the judge can only use written questions and analyze the responses to them to arrive at the conclusion. In short, the human judge has to find out which of them is a machine.
In maximum cases, it is quite easy and obvious to find out on the basis of language dynamics. Also, it doesn’t rely on the correctness of the answers; it is indeed a test of human-like attributes in the text format. In fact, the Test never fails a human.
Advantages of the test
· Helps determine the “human-like thinking algorithms” and their effectiveness in a machine.
· The test is quite simple in its setting and working.
· Takes into consideration the interpretations from other sciences including Philosophy, Psychology, Neurosciences, etc. hence maintaining a healthy give and take relationship and bridging the gap among the studies.
· Can help determine threats (if any) from the machines in future. (Many Sci-Fi movies and comics predict how machines would take over the world with their super-intelligence, well; this test can help raise alarms in advance if any such fantasy ever comes into existence).
The Present
· Turing himself thought that by 2000, Computer systems would be able to pass the test with flying colors, but in reality, no machine has passed the test fully.
· Most of the famous AI and chatbots have seemed to have failed the test including Alexa, Siri, Echo, etc.
· ELIZA, a 1966 AI programme was the first ever to actually confuse a few people in the restricted Turing Test. Since then, Eugene Goostman, in 2014, fooled 33% of the judges by portraying itself as a 13 year old Ukrainian boy. More recently, in 2018, Google’s Duplex has claimed to be closer to pass the test in its text as well as the voice format.
Possible Challenges can include machines fooling humans to give up their personal data and barge into their personal lives too like the “CyberLover” Malware, etc. There can be large scale panic if the difference between “human” and “human-like” fades and a cyber-weapon can be the largest tool in the hands of countries or other non-peaceloving organizations. Hence, we need to make use of machines more judiciously and help them help humans. For this to happen, Turing Test is the key.
Kommentare