Anand Kumar/AI-Generated Image
Every morning, before I open a single newspaper, I read a briefing that no editor commissioned and no publication will ever print. It is a personal investment bulletin I built myself in an AI tool by patiently telling it, over many weeks, what to watch, what to ignore and how to talk to me. For an audience of exactly one, it now goes deeper and proves more useful than anything the business press manages for its millions. It covers items the press will get to later in the week and, in two paragraphs, tells me what the morning’s filings actually mean for the portfolios I care about.
I should admit that I am not a young man given to coding for fun, and that making this tool has given me pleasure quite apart from its usefulness. If a reader of my vintage can teach a machine to read the markets for him each morning, you might consider building something of your own as well.
I begin with a hobby because it is a part of the AI story that actually touches our lives. And it is nearly the only part nobody is arguing about.
The arguments are all about money, on a scale that is hard to grasp. Just this week, Alphabet, one of the most dependable cash machines ever built, announced it would raise $80 billion in fresh equity. A company that generated $174 billion of operating cash over the past year is now diluting its shareholders to keep up. Part of the placement goes to Berkshire Hathaway, whose Google holding now exceeds its Apple holding.
That, depending on whom you believe, is either the surest sign of a coming golden age or the clearest symptom of a bubble.
IBM CEO Arvind Krishna has done the maths in public. The amounts being committed to AI data centres run into the trillions; the hardware becomes obsolete every few years and must be replaced; and the profit needed merely to service such spending exceeds the combined earnings of the entire industry. But he reminds us that the overbuilding of fibre-optic cable in the late 1990s bankrupted the firms that laid it while enriching the Amazons and Netflixes that later ran their traffic over it. The builders lost. The ones who used what the builders left behind won.
An equally confident school argues the reverse. Cheap, good-enough AI models that a customer can run on his own premises, they say, will devour the expensive state-of-the-art ones from below. The giants now pouring in $80 billion are paving roads on which they will soon be overtaken.
I do not know which camp is right. I would bet that neither do they. What I notice instead is that nobody on either side seems to be in any doubt. When DeepSeek wiped 16 per cent off NVIDIA’s market value in a single session early last year, the market briefly decided the AI boom was over. The mood barely outlasted the panic that produced it.
For the ordinary Indian saver with a monthly SIP and a horizon measured in decades, the lesson is not to back small AI over big AI, nor to buy into whoever is spending the most, nor to dump our IT-services shares because somebody on a podcast has pronounced the sector doomed. We do not know who wins. And neither do the people staking $80 billion on the answer.
The one part of this revolution I can hold in my hands is its use. That is where I would like you to spend your energy. Not on your portfolio, which should stay as dull and diversified as ever, but on the tools themselves. They cost little or nothing and improve by the week.
Start with the small problems. A tool can summarise a fund factsheet in a minute, translate a dense SEBI disclosure into plain language, or draft the questions you mean to ask a registered advisor. The tools will surprise you, and learning what they can do is a small pleasure in itself.
The revolution worth having is not the one being fought over on the spreadsheets in Silicon Valley.
It is the small, private and genuinely useful one you can build for yourself at any age.
Also read: SEBI's new rule helps the wrong people



