The White House wants a new 'Endless Frontier,' this time for AI and against China

Michael Kratsios, the White House science chief, wants to rebuild American science on the bones of its 1945 self. In an Economist By Invitation essay published August 13, the Office of Science and Technology Policy director invokes Vannevar Bush's "Science: The Endless Frontier" to argue for redirecting the National Science Foundation toward artificial intelligence and strategic competition with China.

The 1945 document that gave Bush's framework its name proposed a clean division. Basic research, the unfettered pursuit of new knowledge, would be federally funded but institutionally insulated from military and industrial mission work. Mission-driven research would live elsewhere, mainly inside the Defense Department, the Atomic Energy Commission, and the National Institutes of Health. The National Science Foundation, the post-war vehicle for that blueprint, became the institutional embodiment of that split: curiosity-driven first, strategically useful second, with the strategic usefulness left to emerge on its own schedule.

Kratsios wants to flip the priority without abandoning the language. His July 2026 OSTP report, "Science: A New Golden Age," leans on Bush's prestige to argue that the United States now needs a research enterprise organized around the demands of AI and the geopolitical contest with Beijing. The continuity with Bush is the rhetorical move. The inversion is the policy substance. Basic research is reframed as the servant of national strategy rather than its freestanding purpose.

This is where the structural critique lands hardest. A commentary in The Conversation on August 10 argued that the plan ignores the very structures that made the United States a scientific superpower. The piece's central claim is that the post-war architecture, including the insulation of basic research from mission-driven redirection, is what produced eight decades of breakthroughs from quantum information to mRNA vaccines. Redirecting the NSF toward AI and China competition, the argument runs, would dismantle the arrangement that made those breakthroughs possible.

The operational picture emerging from coverage of the report matches Kratsios's stated direction. According to Unite.AI's reporting on the July OSTP document, the White House is steering science funding toward AI and individual investigators, a shift away from the large multi-institution centers and interdisciplinary consortia that have defined NSF spending for the past two decades. The pattern matches the thesis in Kratsios's essay: a leaner, more directed, more strategically focused enterprise.

Community reaction has been skeptical. A Hacker News discussion of the essay questioned whether the current NSF setup is the obstacle the op-ed describes, and what "harnessing AI" operationally means in peer review, in grant structure, and in the day-to-day work of working scientists. The skepticism is not partisan. It is structural. The argument the current OSTP is making, that the post-war settlement is now an obstacle, is one a critic at the National Institutes of Health or a department chair at a research university might also make, in the opposite direction.

The falsifier is concrete. If the redirection produces a step change in the pace of foundational AI advances that can be traced to NSF-funded individual investigators, the architecture critique loses force. If it instead produces a flatter output curve and a more directed but less surprising science, Kratsios's case will be read as another well-intentioned restructuring of an institution that did not need restructuring. The Bush comparison is the bet: the original "Endless Frontier" was vindicated by seventy-five years of work that nobody in 1945 could have specified in advance. The new one will be measured against the same standard.