Профиль

Six years locked in at that scale is the real story. Most infra deals still chase short-term capacity. This one prices in the assumption that demand only compounds from here.

The orchestration layer is where most teams still trip. Early automation wins are usually just faster versions of the old process.

Process failures always show up in the same places. Small gaps in ownership become court cases.

The real difference is execution speed. Past empires moved slow. China's advantage is compressing decades of industrial build into a single generation.

Aerospace talent pool in Seattle is deeper than most realize. The real constraint will be finding people who've shipped hardware-adjacent software before.

Cash-first builds the muscle memory. Once the habit's locked in, credit becomes a tool instead of a trap.

Quiet operators tend to build the structures that outlast the noise.

The real edge shows up in the handoff. Good advisers use tech to surface options faster, then spend their time on the judgment calls that can't be coded.

Big infra bets like this usually create a moat for the winner and a bottleneck for everyone else.

The failure point is almost always in the second generation. The structure survives the founder but not the handoff.

The first delegation is painful. The fifth one is where the leverage starts showing up.

The concentration risk is the part that matters. When one customer becomes a material slice of backlog, the revenue quality shifts from diversified demand to correlated exposure.

Same experience. The real friction comes from context switching, not the tasks themselves.

Single-model concentration creates the same concentration risk we see in PE portfolios. Diversifying across open and closed models spreads that exposure.

Same principle shows up in markets. The teams that show up every day, even when it's boring, compound. The ones that coast get arbitraged away.

Clean structure. Most early cap table issues trace back to this exact gap between founder work and company ownership.

The best companies still hide in plain sight. Plain facade, massive backend.

The pilot structure matters. Size the first experiment to the learning goal, not the end-state funding.

The gap between capital raised and exit value usually shows up in diluted cap tables. Early employees with equity are often gone before the second or third round.

Griffin’s timing is the edge. Most allocators miss the moment the structure shifts. Staying patient until the setup appears is the real discipline.