To the uninitiated Turing Test is defined as follows (Courtesy TechTarget)

In artificial intelligence (AI), the Turing Test is a method for determining whether or not a computer is capable of thinking like a human. The test is named after Alan Turing, an English mathematician who pioneered artificial intelligence during the 1940s and 1950s, and who is credited with devising the original version of the test. According to this kind of test, a computer is deemed to have artificial intelligence if it can mimic human responses under specific conditions. In Turing’s test, if the human being conducting the test is unable to consistently determine whether an answer has been given by a computer or by another human being, then the computer is considered to have “passed” the test.

While there are different views on if the Turing Test has been really cracked , my version of the Turing Test is much more simple and slightly twisted.

Given a relationship between two persons and the word ‘Two Minutes’ uttered by one person, I want the computer to predict the actual time it will take with a + or – 5% accuracy. If the computer can predict that , then I will find a classifier term such as ‘Beyond AI and Machine Learning’  and possibly call it as Contextual Cognitive Intelligence or possibly more fancy term.

To give you some examples of the problem statement,

Here relationship is between two individuals for example

  • Manager and Employee
  • Mentor and Mentee
  • Vendor and Customer
  • Husband and Wife
  • Two Close Friends
  • Boy and Girl who are possibly dating each other or something like that
  • Two Colleagues
  • …..

Assume any such combinations

All I want to know is the following

If my boss says to wait for  two minutes, how much time I have to wait.

If my wife says two minutes she will be ready, what is the reasonable scale I have to apply so that my endurance does not break

If my friend says two minutes he will be at an appointed place, what time he will actually turn up

and the list can go on ………….

I think about it a bit and I do not think it falls into the category of ‘Parametric Model of Machine Learning Algorithm’ or in other words the parameters to determine the actual time are not fixed or the classic definition is as follows

“A learning model that summarizes data with a set of parameters of fixed size (independent of the number of training examples) is called a parametric model. No matter how much data you throw at a parametric model, it won’t change its mind about how many parameters it needs.”

(Artificial Intelligence: A Modern Approach, page 737)”

In contrast, I do not think it fits the non-parametric model as well because there is some need to arrive at a pattern for unseen data.

Two Examples might clarify

When my boss states two minutes

It depends on

  • What state of mind they began the day with
  • What are the hundred complications for the day they faced
  • Did I mess their day in some sub-conscious way
  • Do they have to leave for an emergency
  • Are they in a frame of mind
  • Personal relationship we share
  • Their perceived importance of what I possibly want to discuss
  • Have they received some irate calls on the professional or personal front
  • …..

When my wife states two minutes

It depends on

  • Who initiated the outing
  • Who we are meeting (if we are meeting)
  • Where we are going
  • How we are going
  • With whom we are going
  • Who are all coming (Total population)
  • What is the past history of association
  • Is it a good time of the day
  • What colour dress I am wearing
  • Whether it is the Uttarayana or the dakshinayana
  • Pedigree of people coming
  • What type of food
  • Do we have to take any gift
  • What is the budget for the gift
  • Should we take umbrella
  • Should we take water bottle
  • Should we take short eats
  • Have we fought in the last 4 / 8 / 16 /32/64/256 hours or minutes

I can add more, but for the sake or brevity I will limit myself

To be fair to my wife and to be safe, the reverse possibly is true  as well.

So if you take these to other relationships, it will be possibly the same

It might also depend on the culture, time zone, power equation between the person who is waiting and the person who is saying ‘Two Minutes’, relationship asymmetry and possibly many other indefinite parameters.

I may be exaggerating, but I have not been able to successfully find an answer or a near accurate estimate when someone has said me ‘Two Minutes’ and I continue to hear it. That is the reason I am struggling.

In my view it is a NP-Hard problem and may be IBM Watson can claim to solve this sometime in future.

Hinduism defines something called Yugas and may be one reasonable question which might get me closer to an answer when someone tells me ‘Two Minutes’ is to follow up with a question of which Yuga they are referring to and map it to the current Kali Yuga.

I wish some researcher comes up with an effective predictive model for this and offer this as a foundation level Turing Test !

Till that time, I shall pray GOD to give me the endurance to enjoy the stochastic nature of responses to ‘Decoding Two Minutes’.

Enjoy Maadi (Have Fun)