Shubhang Borkar was not about to be defeated by a crowded temple and an ocean of shoes. Making short work of a hunt for his clogs that might have been otherwise exasperating, the Goa resident made it a case study in everyday AI. In a matter of minutes, with the help of ChatGPT’s image tools, he had his pair back.
An everyday mess meets AI improvisation
Those who have stood in line at a busy shrine are familiar with the scenario. You put your footwear down at the door and come back to a tangle of things that all seem the same. Borkar did not rely on memory; he put his phone to use. He took to the racks with ChatGPT, snapping away until the program flagged something that looked like a match.
It was a simple approach but he was focused on it. One after another he would film an angle and put the still up for the chatbot to see. The AI would make its comparisons and nudge him in the right direction. What was clutter was now a dataset, and one pair of clogs was easy to spot.
Why it matters
Do not mistake this for high tech. There was no custom model or special scanner involved, just a smartphone and some patience. That is the whole point. It is a demonstration of how a general tool can make a routine problem solvable in public spaces where there is little order to be found.
Temple-goers and festival crowds will understand. Rather than second-guessing oneself while sifting through the racks, one can leave the visual legwork to a system that does not get tired.
The quick hack he used
Borkar’s video shows him first recording long rows of footwear to illustrate how indistinguishable many of them are. Then he has ChatGPT open and uploads a picture of his clogs for reference along with shots of the shoes before him.
He lays out the situation to the chatbot: he is at the Mahakal Temple, his clogs are missing somewhere in the jumble outside, so it should scan for them. As the search went on, he fed it more photos from various parts of the rack.
After running the numbers, ChatGPT zeroed in on a particular pair. The camera is on him as he goes to verify it. A hit-or-miss affair is made into a guided search.
The video that started it
Four days ago he put a clip of it on Instagram under the heading ‘Used ChatGPT to find my lost clogs’. He calls himself a ‘hacker on weekdays (cybersec), explorer on weekends’ and presents the exercise as a way to see what the AI can do beyond the screen.
There is a playfulness to it, yet the process could be followed by anyone. He gives the AI context and steady updates and lets it whittle down the possibilities.
What people can learn from this
In a way the episode is a cheat sheet for employing AI vision in a crowd. The logic holds up anywhere you have a jumble of similar items, be it an event entrance or a school shoe room. Some practical points:
– Get some wide shots in first, then go in close
– Have a good reference image of the item in question
– Tell it where you are and what is gone
– Put in several angles to aid the matching
A small success with larger hints
This was no lab demo. It was a real world test of how image analysis can handle micro problems. The fact that he made a match is a reminder that AI is of use if you keep the feedback loop tight and provide the proper context.
You can see the frustration give way to curiosity in the video. By making a problem of the shoe pile, Borkar put an end to the chaos and had a tale to tell.
The bigger picture at public places
It is a fitting example of how AI can be put to work in India’s more active venues. With the right prompts and some iteration, disorder can be put in its place without calling on staff or any extra hardware.
At the Mahakal Temple the clogs were recovered. But what stays with you is the method: a pocketable means of turning confusion into a result.
