<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Nimrat Brar</title><description>Nimrat Brar, MBA candidate at Kellogg. Engineer by training, with work across India, Australia and Canada.</description><link>https://nimratsinghbrar.com/</link><item><title>Napkin Maths for AI at Scale</title><link>https://nimratsinghbrar.com/writing/napkin-maths-for-ai-at-scale/</link><guid isPermaLink="true">https://nimratsinghbrar.com/writing/napkin-maths-for-ai-at-scale/</guid><description>AI builders will spend about $780bn in 2026, 2.4% of US GDP. Through 2028 the pace of the build-out is set by power and electricians, which take years to add; capital becomes the limit only if AI revenue grows less than about 2.4 times a year.</description><pubDate>Mon, 16 Nov 2026 00:00:00 GMT</pubDate></item><item><title>From One Token to a Gigawatt</title><link>https://nimratsinghbrar.com/writing/from-one-token-to-a-gigawatt/</link><guid isPermaLink="true">https://nimratsinghbrar.com/writing/from-one-token-to-a-gigawatt/</guid><description>If most jobs used AI every day, the world would need hundreds of gigawatts of new data centres. Chips and capital could keep pace within a few years; grid connections are the slowest part of the build-out.</description><pubDate>Thu, 29 Oct 2026 00:00:00 GMT</pubDate></item><item><title>What Energy, Capital and Compute Say About When AGI Arrives</title><link>https://nimratsinghbrar.com/writing/follow-the-money-find-the-calendar/</link><guid isPermaLink="true">https://nimratsinghbrar.com/writing/follow-the-money-find-the-calendar/</guid><description>Electricity, capital and compute are pouring into AI at a pace with no precedent, yet a model’s release date predicts its ability better than any of them: ten times the compute buys about four months. Around 2030, electricity and then money start to limit the biggest training runs.</description><pubDate>Sun, 11 Oct 2026 00:00:00 GMT</pubDate></item></channel></rss>