I started writing this piece on an ironical note. On the one hand, I had finally managed to hack my heart rate down to the ideal 72 beats per minute after much work aided by artificial intelligence (AI)—a theme I had written on with much gusto earlier. On the other hand, the faces of a few men assailed my thoughts, made me smile, and compelled me to wonder whether AI can ever recreate humanity and the innate intelligence embedded into our beings.
To put that into perspective, if you’re a bibliophile and have been a Mumbaikar for a few decades now, the name Mr Shanbhag will ring a bell. The genial man who owned the iconic Strand Book Store in the city, knew all of his patrons. There were no back-end algorithms to tell him I was a poor kid looking for a good read at an affordable price. There were many like me. He knew what each of us liked. And he’d guide us to what we may want. Many, like me, have fond memories of him and the store.
In much the same way, I have fond memories of Crossword, the first contemporary bookstore chain India was witness to. It emerged in the late 1980s and was unlike anything anyone in the country had ever seen—except in Hollywood flicks.
It was large, spacious, airy, and contemporary, with no one to peer over and urge you to either buy or get out, as was the norm then. You could browse its friendly aisles, cradle a few books in hand and stroll across to an in-house café and spend a few hours looking through them.
Above all, it exuded warmth and infused trust. Both these traits were best exemplified by two words: “Sriram Recommends”. If it appeared on a book, it meant the book was worth reading.
No reviews or qualifiers accompanied these words. But embedded in them was a promise by R Sriram, who co-founded the chain with his wife Anita. He has since then moved on and is now co-founder at Next Practice Retail and sits on the boards of many non-profit organisations.
So, early in the morning, as I sat down to write this dispatch, I reached out to him to ask what it was about “Sriram Recommends” that compelled a poor boy like me in his graduating days and early working years to take his word as the final authority. Why was it the only stamp of approval most people who visited the chain needed? How did he manage to institutionalise trust without reviews on the back of his name?
What was intended to be a short 10-minute phone call turned into an incredible and untold story. Neither he nor I realised that an hour had passed. And a lot of the story remains to be told. We promised to catch up again to pick up on the threads.
What emerged was that Sriram could empathise with me. Much like me, he said, he came from a middle-class background. He too had to scrounge and save up to buy a book. So, he knew what it felt like to own one. And how careful one had to be before making a choice around what to settle on. His wife Anita understood implicitly what a woman in India has to go through in crowded aisles. He described it as the “butt brush” factor.
So, the place had to be designed such that there was no pressure to buy—which is why you could take your time and browse, and if need be, at the café. Women could do it without worrying somebody would try to feel their bottoms. The wide aisles provided lecherous men with no room for cover or excuses.
And there were his personal experiences, and that of people he had seen from close quarters, built into the design. These were people who had to give up on something to buy a book—because books in those days were expensive.
Because he was a voracious and wide reader—and on the back of much sales data collected from the store, his interactions with people at the stores, and his own experiences from the past—the idea of “Sriram Recommends” was born.
But to be able to whittle down to 100 books, he’d have to read at least 500. And of those 100, he’d pick only what he really, really liked. It cut across genres—from cooking to literature to management.
“And how did the books on the aisles that bore this stamp fare?” I asked him.
“It contributed to at least 15% of our sales,” he said. The numbers were significant.
There was much more there that needs to be told and I intend to engage with Sriram over that on a later date. Our conversation had to be interrupted because he had to leave and I had popped up on him out of the blue. But I did manage to ask him what he thought of AI and its role.
Again, a long answer followed, because it is a theme he is interested in as well. But the answer was left incomplete because he had to leave. The sum and substance of the ground we covered, though, was this:
People trust people—not algorithms
1. As things are, AI on its own is not good enough.
2. That it does a pretty damn good job is obvious. But it needs to be jump-started. And that can come only from humans. Because people trust people—not algorithms.
3. By way of example, Amazon, which I had gushed of in a breathless tone in an earlier piece as an entity powered by AI, has figured it out. That is why, Sriram pointed out, much to the consternation of a lot many people, it paid $150 million to acquire Goodreads, a social media site powered by real people.
AI and machine learning can go only so far—at least as things are now. It can tell you basis what you have read or purchased in the past or nudge you to pointers in similar directions basis past behaviour. But not necessarily implore with you in a tone that only humans can to explore worlds you may not have experienced before.
4. Sriram also pointed me to the Apple Music store. Apple is an entity that understands this nuance. That is why it employs both AI and humans to curate content on Apple Music.
5. That is why people like him are unwilling to write off the innate humanity of intelligence yet and leave it for algorithms alone to decide what is in our best interests.
Like I said earlier, our conversation is incomplete. What it did though is compel me to introspect. What, for instance, was I doing the other day when I walked into a shop just around the corner where I live called Krishna Mobile Store?
My problem was, I use an iPhone. And I am used to headphones. But because I am a heavy user, wear and tear is high as well. At an Apple store, an original replacement costs around Rs 2,000-odd. The rate at which I lose or mangle them, this is an unviable price to pay.
Now, the techie in me knows what to do. The consumer in me though was confused by the choices on offer. And that AI-driven platform Amazon where I shop from confused me even further each time I zeroed in on something by throwing up more choices like “You may also like”.
In a fit of exasperation, I walked up to this nondescript store that had always existed, but I had never noticed. A young man was at the counter.
“I need a good, durable headset for my iPhone 6s Plus. What do you recommend at a good price point?” I asked.
“The Philips SHE1405WT/94T for Rs 550,” he answered without blinking an eye. “Six months’ replacement warranty hai. Market mein best quality aaj ke taarik pe.”
Instinctively, I pulled out my phone and fired up the Amazon app to see if the product was available. Sure, it was. At Rs 312. And it was on Flipkart and Snapdeal as well, marginally more expensive than on Amazon, but significantly lower than the Rs 550 this bloke was demanding. All three platforms offered warranties and free delivery by the next day.
“Give me a good reason why I ought to buy it from you for Rs 550,” I told him. He looked unfazed and offered me the same spiel of how he cannot afford to compete with online vendors.
I didn’t blink. Business is business.
He caved in. “I’ll give it to you for Rs 400. At least that way I’ll get to keep Rs 20 in profits. And I’m just around the corner if you need to replace it. Off the counter, no questions asked. You won’t have to go online and wait a day.” I took the offer. In theory, I ought not to have paid the premium of Rs 80 or so.
But I looked at it this way:
1. I asked him for advice and he gave me his unvarnished opinion. It sounded genuine at a point when I was confused.
2. He allowed me try an opened pair to check if I liked the sound and fit.
3. When confronted with a competitor he couldn’t outbid, he made me the best offer he could and made no bones about it.
4. More pertinently, he proactively asked me if there was anything else I was looking out for. And that if I was, he would scour the markets when he goes out next and get me the best possible price he can so I don’t have to expend time researching what I need to buy. He is on top of his game. I could see it. I save on time, which, to me, is as precious, if not more precious, than money.
5. And then there is a bond that was forged. AI isn’t designed to match that yet.
It’s much the same thing with my neighbourhood kirana store, run by a gentleman called Kanjilal. He knows every family in my apartment block. He knows what everybody needs and stocks just that—except meat and eggs. His faith does not permit him that liberty. But that’s okay.
The store is manned by him. And in his absence by his tightly knit family, who double up as delivery boys and don’t utter so much as a peep when I call on the intercom for a loaf of bread on credit. No minimum charges to deliver. And on credit.
In the earlier article I mentioned before, I had referred to Amazon’s patent on anticipatory shipping. And that it is where the future lies. Be that as it may, to get the transaction done, I have to sign up using a credit card or some such payment mechanism.
While the algorithms can be programmed to make life as easy for me and offer as much value for money, it isn’t designed yet to understand there are months when I’m broke. And that I may need a line of credit for two or three months.
But Kanjilal knows. He understands my family’s spending patterns and if there is a crisis on, gossip wafts to his store as well. He doesn’t state it in as many words. Unless pushed to the wall, he does not stop the line of credit.
He understands my credit scores better than the algorithms that power credit rating agencies
Because he knows I will not default. He understands my credit scores better than the algorithms that power credit rating agencies like CIBIL or those that lie under the hood of online retail platforms.
When I pass by for a short walk, he knows I will pick a half-litre bottle of packaged drinking water from his refrigerator. I don’t need to inform him that I have taken it. He knows I will. Behavioural economics, anyone?
The pro version of Day Cost, an app I use, urges me to input every purchase I make. And it tallies with every input of Kanjilal when he shares a monthly bill made on a handwritten note that he sends across to my wife.
I don’t know how he computes the math in his head. I tried my damndest best to get him to talk, but the Kutchi businessman refuses to engage in a conversation and fobs me off.
Anticipatory behaviour? Anticipatory line of credit? Anticipatory shipment of stock? He’s got it all figured out. And no, he doesn’t offer any discount code or cashback. Take it or leave it. But the damn store works. And it works well.
I don’t know what his margins are or how much he earns. I know nothing about him. But he knows everything about me.
Is this what they’d call a “Human Centred Approach” in management textbooks?
Just to add some more perspective, a day after the earlier dispatch was published, a Twitter handle @bokbokwoosh I hadn’t noticed until then—nor had Twitter’s AI engines suggest I look it up—sent me a message suggesting that I may enjoy reading a book called Embodied Cognition by Lawrence Shapiro.
As an introduction to the theme to get me started off, he offered me a link that I may start out with at Stanford University. It doesn’t get more personalised than that. How was I to know or Twitter’s AI to figure out what this man does?
For that matter, all I shared with you in the earlier article is that Girish Nathan is a senior scientist at Microsoft and works on AI. What if I added that he is trained in Carnatic music and has a badass sense of humour I can never hope to match?
Sure, AI can pin down where he lives and what he may buy next. But I wonder how will it extrapolate what kind of thoughts may a recommendation from Sriram trigger in Girish and whether it can make an offer to counter Kanjilal’s line of credit to somebody working at the cutting edge of AI?
I disengage from technology every once a while. Else I now run the risk of getting into a rabbit hole
I don’t know the answers. What I do know now on the back of engagements like these is that I disengage from technology every once a while. Else I now run the risk of getting into a rabbit hole.
Which is why, I had planned to head out for an overnight trip with my buddies from school after I sent this dispatch out. We exchanged notes as we always, and cracked the same dirty jokes as usual.
And we agreed, as we always did, the saddest day in our lives was when the prettiest woman the good lord created—one of our teachers, whom all of us harboured a crush on—blushed and announced to us she had agreed to get married. The only sound that followed was that of our hearts breaking.
Perhaps AI could have predicted the inevitability of it all. Mr Shanbhag and Sriram could have empathised and offered us pointers to a book or poem to console our hearts. Kanjilal may have harrumphed and offered us a free cola.
For now, though, I’ve got to pack my bag and head out with my buddies.
(This is a slightly modified version of an articles first published in Livemint.)