Uncertainty, Stated

Every Kronos design point carries the uncertainty around it; a number without a range is only half a result, and we do not report half-results as whole ones.

A number is not a result without its uncertainty

A design point like Q_sci 3.424 is a central estimate. On its own it hides how sensitive that value is to assumptions that are not yet pinned down by hardware. Kronos reports uncertainty — the range a value can take as inputs vary within their plausible bounds — so a reader sees the confidence, not just the estimate.

Where our uncertainty comes from

• Physics inputs not yet measured on a Kronos machine — confinement, breeding, transport
• Model form — different valid models give somewhat different answers
• Engineering tolerances in fields, geometry, and materials
• The gap, unquantifiable today, between any model and the built machine

Honest confidence

We attach explicit confidence to forward-looking claims rather than implying certainty. Some of our own confidence figures on the harder milestones are deliberately modest — around the level you would expect for physics that hardware has not yet settled. Reporting a modest confidence is more honest than reporting none.

The point of quantifying uncertainty is not to hedge everything into meaninglessness. It is to say precisely how much weight a number can bear, so no one builds on it more heavily than the evidence allows.