White House "Science: A New Golden Age" Summit Launches a Suite of SI-for-Science Initiatives — $2.4B Industry Compute for Genesis Mission, $1.8B DOE–NIH Biohub Data Partnership, NSF Office of Metascience
October 8–9, 2026. At the White House "Science: A New Golden Age" summit, OSTP Director Michael Kratsios announced what he called "the most comprehensive and far-reaching suite of science initiatives by America's partnership of government, the private sector, and universities since the Second World War bound them together in the first place." The through-line is superintelligence-for-science ("SI for science"): frontier compute, autonomous labs, and AI-ready data as federal science infrastructure.
The headline numbers announced at or just before the summit:
- Genesis Mission compute: "Genesis Mission industry partners are committing $2.4 billion in compute, frontier SI for science models, and other tools for researchers." (The Genesis Mission itself was expanded in July from DOE to 15+ federal agencies with $5B committed to national science/technology challenges.)
- Autonomous experiments: NSF and DOE are launching "a $100 million investment in developing the instrumentation needed for autonomous, SI science, and a nationally integrated research system in which anyone can send their experiments to programmable cloud labs."
- Biomedical data: "yesterday, DOE and NIH announced a $1.8 billion dollar partnership with Biohub to develop the SI-ready data needed to build predictive models for biology and medicine."
- Data federation: DOE will federate 190 key datasets across 14 federal agencies through its American Science and Security Platform.
- X-Labs Consortium: $310 million of private contributions to support X-Labs teams focused on "physical SI and quantum technologies" (NSF's X-Labs — $1.5B over 10 years, announced in May — also added three new topic areas on Oct 9).
- Regional compute: a Southeast Regional AI Computing Consortium (14 institutions, 10 states), a North Carolina hub (UNC, Duke, NC State, Clemson, RTI), and $1 billion from Georgia Tech and Emory for frontier compute and SI tools.
Metascience and replication — the self-audit layer. NSF is launching an Office of Metascience (first experiments: "Golden Tickets" and distributed review); NASA is starting a double-blind peer-review experiment in research-merit review; DOE is launching its own Golden Tickets program; and NSF is partnering with UKRI's Metascience Unit toward "an international network of metascience units." On replication, Kratsios highlighted that "the NIH is focused on addressing the replication crisis, launching R3PEATS, a $174 million reproducibility initiative, as well as 'linked discoveries,' a tool to determine whether or not research has been replicated or invalidated at scale."
Talent and friction-reduction. NSF is renewing the Young Scholars Program (with National Math Stars and Jane Street), launching a $50M NextGen Workforce Program (integrated 6-year Bachelor-to-PhD), and — with NIH — converting more than half of conventional grant application forms to a "just in time" approach. NIH announced Generation Discovery, bringing 100 early-career scientists into its intramural program (Kratsios cited a 71-year median age for NIH intramural researchers, and a drop from 22% to 1.3% in the share of first-time NIH grantees under 35 between the 1980s and 2025). DOE is launching $100M Genesis Mission Fellowships (4-year PhDs with a year in national labs/industry).
International framing. The Kyoto Vision for a Golden Age of Science, endorsed by 17 countries last week, extends the agenda diplomatically. Other stated goals: a fault-tolerant quantum computer "by the end of this term," commercial fusion activity by the mid-2030s, an orbital nuclear reactor in 2028, and first moonbase elements by 2030.
What to make of it. The summit pairs large announced sums with structural experiments (metascience offices, blind review, just-in-time paperwork) that aim at the actual frictions researchers report. The open question — typical for summit announcements — is which of these commitments are appropriated funds versus industry pledges and repackaged existing programs, and whether the R3PEATS replication push and the DOE–NIH Biohub data partnership get operational detail.