IDHAR UDHAR KI BAAT 122 - CYA to AI Brig PS Gothra (Retd)

  “I can’t find my passport photo,” I said, after wasting a good ten minutes digging through my phone to upload it on the Artificial Intelligence (AI) Summit registration app.

“Just take a selfie and upload it,” my son said casually.

And just like that, the registration was done.

I made a mental note: Remember to take off your blinkers while thinking. After all, you’re visiting an AI summit.

The first stall I visited claimed it could automatically detect and resolve fraud in banking transactions.

“So even then, is it possible AI might miss something?” I asked in my best sceptical tone.

The young chap looked at me as if I had just asked whether the sun rises in the west, but he patiently began explaining the safeguards. By then, my brain had quietly switched to standby mode.

This had happened before—ten years ago, when a junior officer tried to explain that automation had made physical registers unnecessary. I was in full CYA (Cover Your Ass) mode then.

“But how will the book inspection be carried out?” I had asked solemnly.

Seeing the depth of my professional myopia, the officer had wisely abandoned the argument. I often wonder—maybe he’s senior enough now to have finally abolished those archaic registers. A small pang of guilt hit me. My CYA syndrome may have delayed technological progress more than enemy action ever could.

I moved on to the next stall—AI in cricket.

It got me thinking: why not in the fauj?

After scratching my head a bit, ideas started flowing. AI could analyse enemy pilots’ manoeuvring patterns from peace time data and help predict or counter them in combat. It could study movement patterns on the border, identify indicators of infiltration, or even convert night-vision inputs into clearer daylight images.

For a moment, I felt like an “AI Baba”.

And then I felt worse—if I hadn’t been so busy covering my backside in service, maybe I’d have contributed something useful myself.

Next stop: the DRDO stall.

Frankly, it was underwhelming. Only four things were on display.

One was a system to test AI systems for security.

Second: something called “Adding Value to Information Enabling Battlespace Dominance”—which I couldn’t understand, and there was no one around to explain it.

Third: a Mandarin-to-Hindi translation tool—actually quite useful, if it works.

Fourth: face recognition. It felt like asking for apples and being offered oranges.

That was it. Maybe the rest of DRDO’s miracles are too secret to display. Or maybe that’s enough to justify the pay grades.

While I was standing there, a young fellow said, with a faint trace of sarcasm, “Sir, it’s nice to see a veteran showing interest in AI.”

I wondered how he figured I was a veteran. Probably my chitti daadi gave me away.

“What’s wrong with that?” I asked, slightly irritated.

“I worked on an automation project,” he said, “and I feel people of your generation are the biggest impediment to progress in the Army.”

That stung. “How?” I asked.

“First, you don’t want to shed baggage.”

“What baggage?”

“You cling to formats, minor staff duties, abbreviations. Do you realise your data becomes useless because it’s riddled with abbreviations? I hear some forward-looking commanders have finally dropped the abbreviations.”

“What else?” I asked, cautiously.

“Your habit of keeping data close to the chest. AI systems grow on data and usage.”

I realised that we have a tendency to over classify the data. I myself used to keep even the information regarding the enemy under lock and key rather than disseminating it to the people who need it. Maybe this obsession came from our old Official Secrets Act mindset—so much self-imposed restriction that we end up jamming our own wheels of progress. I once knew an Indian working on a highly secret project abroad. He couldn’t divulge any information because he faced regular lie detector tests. Why can’t we do something similar?

“And then,” he continued, “there’s the inability to clearly define the problem and the solution. Plus, shifting goals.”

That hit home. Deep down, I too suffer from the dil maange more syndrome. I often couldn’t define the problem because I was too busy firefighting trivia.

I stayed quiet. I had a feeling the worst was yet to come.

“Some people do things only for optics, not for improvement. So half-baked solutions get fielded just to score brownie points with senior officers inspecting or about to move on posting.”

“I don’t agree,” I said, stiffening. “Vendors delay things and use this as an excuse. And what’s wrong with optics?”

“That,” he said calmly, “the Kalkotias will explain.”

And with that, he walked away from the stall.

I thanked my stars for retiring at  good times. With the advent of AI, some day the performance of the individuals will be assessed by the AI analysing all the outputs. That day the CYA types like me will become teen numbri from nau numbri.

Note :- Will it not be prudent to analyse why the Prospective Planning Directorates and the think tanks failed to  make us a leader in AI. It will be worth going through the old LTPP and TPCR documents.