A tiny intro before you land. Heads up, this is personal blog!

Sovereign AI is not just a government story. From Mistral to Latham & Watkins, the real question is how much of the AI stack countries and companies actually want to control.

Mapping 125 NVentures companies to explore how NVIDIA finances future compute demand, useful AI applications and the infrastructure needed to support them.

How IceCube turned a cubic kilometre of Antarctic ice into a neutrino telescope, why scientists need neutrinos, what the detector actually measures, and what comes next.

Vinci is building AI-powered simulation software that lets hardware engineers test heat, stress and other physical behaviour in seconds instead of waiting hours or days.

Singapore now has a server rack that needs to be fed. Inside are living human neurons wired into silicon chips, raising a bigger question: can biology become a new computing substrate?

One rewrites the maths. One rewrites the physics. Governments are spending billions on rival plans to keep encryption safe from quantum computers.

Robotics mega-rounds are multiplying, but the real signal is not that humanoids are suddenly everywhere. Physical AI is scaling first in structured environments—and the quieter opportunity may sit in the infrastructure and post-deployment stack.
Each row shows a real funding stage and round size. Drag the name cards into the slot you think matches. Hit SUBMIT to reveal the answers.
One pixel-perfect essay every Sunday. No noise, no boss fights.