Artificial Intelligence - A Friend or Foe, YOU Decide!

Imagine you decide to buy a new car.

The first thing, people generally tend to look out for, is to gather information about the features, the design, and the mileage (being most important to many out there).

Going by the numbers, the car is set to give you an average of 25 kms/litre. And, the reserve tank has 4 litres capacity.

Now, how would you know for sure the distance you could travel on a full tank and/or on a reserve tank?

Did the company test all the possible scenarios before determining the average?

Or, did they just tested the car under limited conditions and ascertained it to perform that much?

AI can get me the conditions (internal - like speed, oil density, etc and external - like weather, road condition, etc) and all the required data sets, but a human has to intervene to know for sure how the results are impacted, to make a sense out of all the information available.

Subsequently, think about an investment firm which is trying to test out a new algorithm on their stimulation, replicating the market conditions they have been working with.

You can pre-fill the tests with the values to be tested, but it would need a human intervention to determine how the results are progressing or when the change the conditions on the fly.

It’s like having a huge dataset in front of you, but without knowing what to look for, it’s all trash.

Machine Learning or Artificial Intelligence is all about teaching a machine to do things for us, which would help us reduce the burden of doing things time and again. Just like the interns (in the kitchen) do the basic preparation before a subject expert, like a chef, would walk in and make a hell of a dish for you (dazzling and delicious).

We (should) use AI or automation to get done with the things which are similar or repetitive or mechanical. However, it would be irrational to rely completely on machines.

Be it a human or a machine, if not contained within defined boundaries or not controlled by set of rules, has a tendency to run rouge – as it happened with the bots which automatically started interacting with each other, at Facebook!

“By far, the greatest danger of Artificial Intelligence is that people conclude too early that they understand it.” ~ Eliezer Yudkowsky

To me, quality and AI should go hand in hand (with endless list).

1. AI can help convert my actions on screen or my speech in to meaningful steps or test cases; however, the QA/QE has to decide if the steps make a relevant journey and what data to be used during that flow.
2. AI can help gather information about which module is most impacted by changes in the system, or which module has been failing over and over again; however, the QA/QE has to decode the environmental changes which are impacting the system.
3. AI can help me execute; however, the QA/QE needs to be the brains behind it – that’s where the creativity of the human mind comes into picture.

Most importantly, one should understand that a machine can (and should) never fully take over human participation.

“Some people call this artificial intelligence, but the reality is this technology will enhance us. So instead of artificial intelligence, I think we'll augment our intelligence.” ~ Ginni Rometty