A leading media channel called to check if I would be open to a debate on tv with the TCS leadership. I am happy to debate with any IIT engineer. It is unfair to let me make mincemeat out of a barely literate non-engineer even if he calls himself chairman. But yes, a public debate is an excellent idea.
Q2: cash on books
*Indian IT firms sit on deep cash reserves and massive balance sheets built over decades.*
Actually they don’t have much cash. Deep and massive are all relative terms. Ai has redefined the frame of reference. Ai hardware companies are now bigger than the entire banking sector in the planet in terms of market cap. So I don’t see anything massive in India. Or deep. Cos like TCS are a rounding off error in the ai landscape.
*Question* If renting AI infrastructure is a margin-destroying trap, why can’t a cash-rich giant like TCS simply build its own sovereign, proprietary foundational language models from scratch? Is the barrier a lack of capital, a severe deficit of deep research-grade computer scientists, or an absolute, inescapable dependency on American silicon hardware like Nvidia chips?
The challenge is always two fold.
One is technology. The second is leaderhsip. Get this right and the money will come to you on its own seeking a gate pass to enter your gate. It will also treat you with respect.
I don’t think there is any dependency as yet. Like I was telling the so called leaders at the India AI Summit – in crude hostel language (pardon its use) – “they don’t have a pot to piss in as yet”.
But then money can create very deep moats around any castle. And then you can keep building a larger castle alongside deepening the moat. I am very hopeful of Bharat Gen – not because it will succeed or be allowed to succeed – that is serendipity – but because it will tell us what doesn’t work. It is very hard to predict a winner. It is easier to predict what won’t work.
That is true in the stock market as well. Any seasoned investor will tell you what won’t work. But it is really really hard to predict what or who the next big thing will be. We have no idea who is making a bullet with your name on it. A small garage in Dharavi may be preparing a bullet with the TCS name on it.
I think the Americans are being foolishly over confident on ai. They underestimate the power of numbers. A pack of dogs can kill a lion. Depends on how many dogs.
Nvidia is a nice engg company. Jasen is a good sales guy. His narrative that a country’s progress can be measured by the size of the chq he writes to Nvidia will go down in modern management thought as really brilliant. No one could ever do it so brilliantly. Not even bill gates with his claims on solving hunger or disease.
But Nvidia is not invincible. At least not yet. Elon musk has put his hat in the ring – first with Grok. Now with Cursor. I don’t like his weed habit and what he does when he is high. But as a dreamer he is great.
