India's regulators have decided who answers when artificial intelligence fails. But its financial system still lacks a reliable picture of what those failures could cost, and how far a single fault might spread. A Calcutta boiler and a Connecticut insurer show what to build next.
It's been less than a week after preorders opened but Temple has divided Indians. One group, mostly doctors and people who consider themselves rational, are demanding proof. Its creator Deepinder Goyal, and a community of tech founders, is asking for faith. Both are right.
There are two movies on Netflix and Jio Hotstar that aim to explain AI to people outside tech circles. One of them is trash. The other one, while unsure of its ending, is a good watch for anyone trying to figure out this beast called Artificial Intelligence.
Humans are imperfect, so it goes without saying that what we create is also imperfect. In the case of software, this imperfection means security bugs that can be exploited by AI. This month we got a glimpse of it with many calling it Bloody October. Worse, it's only the beginning and most people have no idea.
Elon Musk is lecturing India on what we should do or not do to improve telecom services in India. He might have a point or two. But he is also using logic in the way manipulators do — to gaslight the other party, which in this case happens to be an entire country. India must ignore him.
Sam Altman wants the world to suffer a few bad things for the collective good that AI promises to humanity. He is right.
The score was almost certainly true; the mistake was treating a measurement of a model as a verdict on the system built around it.
iPhones are expensive in 2026. Just like everything else. So, stop the bull, bear the cost, and get on with your life. Have an egg instead of having one on your face. Though that too is up 20 per cent in prices just like the latest iPhone 18 Pro.
When an AI agent makes a wrong decision, it can have a much bigger impact when compared to a wrong response by a chatbot.
If you feel that the world is increasingly getting more expensive, and chaotic, you are not hallucinating. It is getting more expensive, and is set to get more expensive once effects of El Nino and other inflationary forces, including RAM prices, kick in.
There is a cliche often repeated in India: data is new oil. In 2026 we are learning just how wrong it is. It's not data that is the new scarce, and yet essential, commodity. Instead, it is RAM and, as Zoho boss Sridhar Vembu has noted, it is doing to the world what oil has done for years.
In 2026, phones are expensive thanks to AI companies that have scooped up all the available memory and chip capacity, leaving everyone else fighting a costly price war. But surprisingly, by 2026 standards, the new Google Pixel 11 phones seem rather well-priced.
The possibility of MDR, aka a small fee on UPI transactions, has been raised again by the government. It is also without any logic because logic dictates that UPI transactions must remain for all, from users to merchants.
For far too long Meta and its boss Mark Zuckerberg have used the word algorithm to avoid taking responsibility for their own platforms like Instagram and Facebook. They are doing it again after Facebook deleted a video of Indian Prime Minister Narender Modi.
CJP protesters might have "won" but a close look at the saga makes it clear that neither Gen Z nor the Indian government, as shown by the Bitchat episode, are any good at using tech in a politically-sensitive environment.
The world is fast turning into a space divided between those who have access to top AI and those who don't, similar to how some nations have 5th-generation fighter planes and most don't. The US has Mythos, China now has Kimi, and that could be a headache for India.
Machines can replicate the procedural layer of reading, writing and arithmetic with astonishing efficiency, but the imaginative core that gave these capacities civilizational significance remains deeply human.
In earlier technological eras, words explained the system. Now words have become the system. That should change who sits at the table when machines are built.
Data protection and capitalism have long been perceived as opposing forces. Yet the framing is reductive, and perhaps deliberately so.
The very sophistication that made modern institutions powerful is now generating their brittleness. Every framework adds stakeholders and dependencies.
