Thursday, January 15, 2026

The Uncertainty of AI

 The Uncertainty of AI

And the Danger of using it as a 
Research Tool 




The other day, just out of curiosity, when I had a few spare minutes, I googled my father’s name to see what information would emerge. I put in his full name and designation: Lieutenant General Chandra Shekhar.

(I initially searched for General Chandra Shekhar, and it showed the biography of Shri Chandra Shekhar, former Prime Minister of India!) 

Inevitably, the AI Overview popped up on top.


This is what it had to say. 

Generally all correct. 


But there was a catch.

The photograph accompanying my father’s description and career — was not of my father’s!


When I shared the image on our family group, my father, naturally, spotted the error immediately.

“This is somebody else’s photograph,” he said. “What can we do about this?” 

“Well, nothing that I know of,”, I answered. 



And that is the problem.

The fact that the AI Overview is what you see first when you search for something on the internet — and the fact that this first view that surfaces is generally a mix of some correct and some incorrect information. There is of course, the inevitable disclaimer, if you scroll till the end: “AI responses may include mistakes”. To sort out which parts are mistakes, we need to be equipped with some preliminary information before we search — and certainly before we focus on the AI Overview.


This is why, I advise everyone, especially my students when they are researching on something, to certainly not depend on AI. 


Who knows what strange concoction of fact and fiction it will throw up?