A lively book launch explores how AI can become understandable, practical, and human‐centered for everyone, not just experts.
Mr Sairam Sundaresan
AI Engineering Leader · Author, AI for the Rest of Us — Main Speaker & Book Address
It is 1943 in Chicago. The world outside is burning — millions of lives lost and millions more at stake. Yet in a basement, two unlikely collaborators are waging a quieter war. Warren McCulloch, a neuroscientist, and Walter Pitts, a self-taught logician with no fixed address and no institutional backing, are trying to understand one thing: how does the human brain work? Their hypothesis was elegantly simple — the neuron, that biological cell that fires every time you think or feel, is not magic. It is logic. True or false, one or zero, on or off. And if you connect enough of these logic units together, you can simulate thought, reasoning, and problem solving.
That idea is now over eighty years old, and we are living inside the edifice it built. We walk through it every day — yet most of us do not know which walls are load-bearing, which corridors lead to meaning, and which lead to darkness. My book is the floor plan I wish to offer.
The most dangerous recurring problem we face is not a problem of technology. It is a problem of ego. In 1966, Marvin Minsky and Seymour Papert at MIT assigned undergraduate students the task of solving computer vision in a single summer. What could possibly go wrong? It turned out to take fifty years, billions of dollars, and countless failed experiments before a machine could reliably distinguish a cat from a dog. We are repeating that exact mistake today with AI agents — spending billions, wasting trust, and falling prey again to what is known as Moravec’s Paradox: what is easy for humans is very hard for AI, and what is easy for AI is very hard for humans.
The CEOs who purchase AI models often believe they have bought a truth machine — an oracle. What they have actually bought is a probabilistic engine that dreams, guesses, and predicts. It is brilliant at finding hidden patterns at scale. It will write 100,000 lines of compiler code and pass 99 per cent of the test suite. But ask it to carry the implicit, unwritten context of your organisation — the engineer who left in 2012 and added a comment saying ‘do not touch this line’ — and it fails. We have automated the power of reading, but not the power of understanding. We have automated the data pipeline but not the meaning pipeline. Garbage in, garbage out.
In this age, creation is free. The highest-value skill for the next decade is not creation — it is curation. Taste is knowledge. It is the ability to see what is good from bad, and to know when something works. A rising tide lifts all boats, but only the boats with sails move faster. Curation is that sail. I hope this illustrated book gives you the floor plan you need to navigate this extraordinary structure we are all living inside.
Mr Suresh Raman
Past President, MMA · Former Vice President & Regional Head, TCS Chennai
Before I went through the book, I asked myself a few honest questions: Am I a data scientist? Do I code every day? Am I an AI architect or a machine learning expert? Am I a mathematician or a statistician? Being in the IT industry for decades, I can say with confidence that I am none of those things at an expert level — which means I am squarely among the rest of us. And that is precisely why this book matters.
What struck me most in reading the 250 pages over 25 days is that instead of dry technical definitions, you get real narratives. There is literally a story about a robo-baker and stolen muffins to explain supervised learning. There are practical, real-world examples showing how we are already encountering AI in daily life — whether it is a spam filter using logistic regression, Netflix recommendations built on collaborative filtering with embeddings, or the moment Instagram tags you in a photograph through convolutional neural networks. Sairam has made things easy. A lot of these are buzzwords that intimidate most people, but he strips away the mysticism.
I also came to understand how transformers revolutionised language processing by solving a deceptively simple problem: how do you make a computer understand that in the sentence ‘the animal didn’t cross the road because it was too tired’, the word ‘it’ refers to the animal and not the road? That kind of contextual reasoning is something we take completely for granted, and yet it was the frontier problem of a generation of researchers. This book helps you appreciate what AI is, what it can do, and how we can leverage it — without requiring you to write a single line of code.
Mr Antony Prashant
Partner, Deloitte Touche Tohmatsu India Pvt Ltd — Life Sciences & Healthcare Consulting Leader, South Asia
I lead life sciences and healthcare consulting at Deloitte South Asia, and what Sairam’s mountain analogy captures perfectly is the central question my sector is wrestling with every day. You can reach the summit by helicopter or by climbing — both get you there, but only one of those journeys transforms you. AI is the helicopter: it streamlines, automates, and compresses timelines dramatically. The question is whether we use it wisely.
In diagnostics, the typical cycle is painfully long. A patient walks in, sees a general physician, is sent for blood tests, then MRI, then perhaps a PET scan, and eventually a biopsy — six to eight weeks can pass before a clear diagnosis emerges. AI models trained on thousands of scans carry the combined knowledge of more than a hundred physicians. They can correlate data across multiple imaging types simultaneously and flag a potential issue far earlier than a single clinician reviewing results sequentially. The question this raises for me is: is a physician with AI more valuable than a physician without AI? And what happens to the emotional dimension — someone still has to sit with the patient, explain the diagnosis, and hold their hand through what follows?
I tend not to call it artificial intelligence. I call it assisted intelligence, or augmented intelligence. That framing matters because it orients us correctly. AI should be doing the heavy analytical lifting — pattern recognition at scale, early flagging, reducing diagnostic lag — while the human physician retains accountability, exercises clinical judgment, and provides the emotional presence that no model can replicate. For any high-stakes decision in medicine, law, or finance, you must keep a human in the loop. Not as a checkbox, but because the model has been trained on data, and if that data carries historical bias or poor annotation, the output is not just wrong — it is dangerously confidently wrong. Augmented intelligence, done right, gives us the best of both.
Q&A
Mr Suresh Raman
You describe the book as analogy-driven with zero coding and real-world examples. As someone who sits squarely among ‘the rest of us’, I found the robo-baker story explaining supervised learning delightful. But my question is this: India has an extraordinary ability to leap technology generations — we went from a passbook airline ticket to DigiYatra facial recognition in two decades. What role do you see AI playing in shaping our country’s economy, and what gives you confidence that we are ready?
Sairam Sundaresan:
You are absolutely right that India is exceedingly good at adopting new things and skipping generations. AI is no different — except that with AI there is no user manual. There is no single prescribed right way or wrong way, and the guidelines are being rewritten every time a new model is released. But what we do have as a culture is courage. We are not hesitant to try things, and that willingness to experiment and update one’s priors is the single most important trait for AI adoption.
Think about what AI does for a young person with a brilliant idea. Previously, you needed funding, a pitch deck, a team. Now you have a tireless Swiss Army knife next to you that is very good at most things you are probably not good at. Instead of building a pitch deck, you can build a working prototype on day one, put it in front of users, collect feedback, and iterate. Imagine the number of solo entrepreneurs who can emerge without needing a traditional four-year degree or months of fundraising. In a country like India, with so many bright minds trying out ideas in parallel — that is the economic impact I feel AI can have. The only thing stopping you is your own inertia.
Mr Antony Prashant
I lead life sciences and healthcare at Deloitte. In our sector, an AI model trained on thousands of MRI scans can detect critical disease months before a physician can — yet there is also a human being in that consulting room whose life is at stake, who needs not just a diagnosis but empathy. Is the future a physician with AI, or does the emotional dimension mean there are places where AI simply must not replace the human? I tend to call it assisted intelligence rather than artificial intelligence — does that framing resonate with you?
Sairam Sundaresan
Your framing resonates completely, and I would extend it further — augmented intelligence captures it even better. What AI does in diagnostics is remarkable: it analyses scans at a scale that represents the combined knowledge of more than a hundred physicians, and it can preemptively flag something before it becomes visible to a single pair of eyes. That shortcut in the diagnostic sequence — from months of sequential tests down to an early alert — can be the difference between a life saved and a life lost.
But here is the thing: would you truly believe a diagnosis from a machine, delivered cold and without context? Or would you believe the same finding from a physician who has seen hundreds of thousands of patients, who can explain it to you not just logically but with the warmth and clarity that a frightening moment demands? For any decision where the stakes are high — medicine, law, finance, education — you must keep a human in the loop. Not because AI is incapable, but because the model has been trained on data, and if that data was poorly annotated or carries historical bias, you are trusting something that is entirely false. The future I would love to live in is one where AI is precise and tireless in the operating theatre and in the clinic, while the physician carries the emotional intelligence, the accountability, and the human touch that no model can replicate. That is augmented intelligence at its best.
