About Me

I'm Jim Chiu: nuclear-engineering trained, a former software CTO, a serial builder with a successful exit, and now founder and CEO of Holonomy Systems.

AI is running into an energy wall. In late 2025, with an AI at my side, I started chasing one question: who builds the way through? It led somewhere I didn't expect. The way through is built on semiconductors.

Working backwards: AI needs abundant energy. Fusion now looks more like an engineering problem than a physics one. A laser-fusion plant has to close its own power loop, and in a diode-pumped plant that loop starts with a semiconductor, the laser diode, firing about ten times a second for decades. Today it falls short on capacity, efficiency and cost.

In hindsight, the six months that followed, from late 2025 to June 2026, were a funnel. Nine conferences across Taiwan, Europe and the US, and two labs, narrowed it one layer at a time: energy → fusion → lasers → semiconductor lasers. What came out the bottom was a problem small enough to bet on: high-power lasers made the way chips are made.

AI did the reading for every layer in days. The conferences did the grading. For an outsider, that combination was the whole method: AI to get me to the frontier, and the people at the frontier to tell me where I was wrong.

The full story, hypothesis by hypothesis, is in Finding a Problem Worth Betting On.

Today I'm building Holonomy Systems with a founding photonics engineer who ran process integration inside TSMC. We call what we're building the photon engine for civilization. It's just a humble start.

This blog is where I keep the field notes: what I see at each stop, and what it changes.