The core risk is not whether AI needs more optical bandwidth; it does. The real risk is whether POET can convert design-ins, partner deals and
qualification work into repeatable, profitable shipments before concentration, pricing pressure and operational mistakes erode its share of the economics. Recent capital solved the near-term funding problem, but it did not solve proof of demand quality or manufacturing consistency.