The reactor is not the bottleneck. The plant is.

The next wave of nuclear power will not be won on reactor design. It will be won on sequencing. Blue Energy's first US construction permit application, filed with the NRC for a site at the Port of Victoria, Texas, makes that concrete: the load-bearing variable in AI-anchored nuclear is the project-finance structure, not the small modular reactor.

The company-stated evidence is twofold. A natural-gas bridge, not an SMR, is what Blue Energy says gets electrons to the grid in 36 months. And the reactors themselves, per the company, are under 10% of plant cost. The remaining 90% is construction and regulatory friction at the rest of the plant, which the bridge restructures. Build a conventional gas plant first, anchor offtake to Crusoe's 1,600-acre AI factory campus, and reuse the balance-of-plant when the SMRs come online.

Wire coverage will frame this as "AI meets nuclear." The useful reframe is that AI demand is the collateral, not the innovation. The innovation is a sequenced balance sheet, and Blue Energy's filing is the first time that structure has gone in front of the US regulator. If the gas bridge delivers commercial output on the 2028 company-stated timeline, the sequencing thesis becomes a template for any AI-anchored nuclear site. If it slips, the bridge collapses on its own evidence.