Before we begin, Ordinary Analysis is totally free, and I’m happy to share it with you. If you want to receive every new post in your mailbox, you can subscribe by clicking on this cool button below.
Picture this: you step into a cozy Darshini Cafe in Bengaluru, known for its South Indian vegetarian food and filter coffee. But today, you have an unusual request: “Do you serve Black Tea?” The caretaker nods, promising your order. It has been 15 mins since your order, there is no sign of the caretaker or the black tea. You’re getting impatient. Suddenly, the caretaker and the chef approach you, politely asking, “Sir, what is Black tea?”
This anecdote was shared on one of the episodes of The Seen and The Unseen podcast. One of the guests (Narendra Shenoy) recounted the incident, while another guest (Krish Ashok) added an interesting observation here: “This will happen in most of the Indian states. Except maybe in Kerala. What the rest of India does, Kerala will never do.” I found this remark rather pertinent.
I visited Kerala, God’s Own Country, in October 2023, and the experience was delightful. There were stark differences compared to other states. For instance, street vendors here respect your first “no.” Unlike elsewhere, they won’t pester you repeatedly. Even at temples, there’s no persistent offer of guided tours—a common occurrence in other parts of India. In Kerala, simplicity prevails: visit the counter, buy a ticket if you want a guide, and that’s it. If you don’t want to do that, then chill!
Krish Ashok had shared another Kerala tale. At a beach shack in Kerala, a customer eagerly requested cocktails, only to discover they weren’t part of the menu. But the customer had wanted it badly so he offered instructions on mixing one. The owner, uninterested, replied, “We have rum and vodka—take your pick.”
With my very little experience in the Data Science domain (~3 years), I’ve noticed a similar trend where companies follow the non-Kerala way while marketing their products. Companies and marketers often tout “AI-enabled” products where there is no AI. Or the traditional AI.
In my previous company, I had a meeting with one of the consultants they had hired. They were sharing one “AI enabled” product for Aviation. The demo of the product looked quite good. It seemed to me a valuable tool. But I had wanted to understand the exact AI algo behind it. It seemed like some mathematical modelling with simulation techniques.
After continuous pestering, they explained that it was an ‘AI-enabled simulation.’ However, I’m quite certain it was just a simulation. Don’t get me wrong—I’m not trying to belittle the product or the technique. It was good. But the ‘AI-enabled’ label seemed like a marketing gimmick, much like many other products.
In recent times, I’ve noticed a rapid shift. Conversations now revolve around not just solving problems, but specifically how to tackle them using LLMs (large language models). This Reddit thread encapsulates my feelings.
The real question is: What qualifies as ‘AI-enabled’ now? When we break it down, most Machine Learning algorithms, including LLMs, rely on linear algebra and basic statistics. Maybe someone, who has seen more Data Science Summers, can answer this question more eloquently.
Thank you for reading Ordinary Analysis, a blog-newsletter where I write about different topics that intrigue me. If you enjoy this post and want to see more, you can subscribe by clicking on this cool button below. It’s free, easy and rewarding.
Alternatively, you can also share it with someone you think will like it.