Tracing the evolution from pattern recognition to intersection discovery—and asking whether AI can take the final leap. Human knowledge forms a multidimensional lattice; ideas emerge where concepts converge. AI explores intersections at unprecedented scale, but can it create entirely new conceptual dimensions? That capability might mark the arrival of true intelligence—or the question itself might be wrong.
The internet was built for humans. Then machines started talking to each other in the background. Now AI is becoming the primary consumer of web content, and websites are adapting to serve it. Cloudflare just made that adaptation trivial.
A class-action lawsuit claims Meta can access WhatsApp messages. I don’t know if it’s true. But the lawsuit gave me a reason to revisit a technical capability I’ve been writing about.
The internet is quietly being hollowed out. As AI answers our questions directly, we stop visiting the websites that created the knowledge in the first place. What happens when the infrastructure of human knowledge loses its economic foundation?
Companies can already mine your non-encrypted data on their servers. It’s your end-to-end encrypted data that was supposed to be untouchable. On-device ML first changed that, turning your own phone into the profiling engine that encryption was meant to prevent. On-device LLMs now take that to another level.
This dish is inspired by a Brussels sprouts plate I kept ordering at RT Rotisserie in San Francisco. Deeply roasted, aggressively garlicky, bright with lemon, and finished with crisp edges that almost shatter when you bite into them. I wanted to recreate that feeling at home without cheese or cream, letting heat, acidity, and umami do the heavy lifting instead.
Why do we bend our bodies to furniture instead of bending furniture to our bodies? A simple proposal: fix the table height, free the chair, and let two adjustable planes end the tyranny of one-size-fits-none seating.
A practical guide to building scalable map interfaces that render thousands of polygons without overwhelming your backend. Learn progressive rendering, smart caching strategies, and production techniques used by Google Maps, Uber, and Airbnb—from quick wins to architectural patterns like vector tiles and spatial indexing.
AI learned to write from us. If my natural voice sometimes sounds like a language model, that’s their problem, not mine.
A quiet Mysore evening at a café window, an old man with his kulfi, and the fleeting glance we shared — proof that the smallest moments can linger the longest.