The Jackson School’s Blue Center for Global Strategic Assessment hosted a lecture February 11 on “States’ Use of Data,” as part of its Co-Lectures in the Theory and Practice of Statecraft series. The event featured Aleh Tsyvinski, Arthur M. Okun Professor of Economics and Global Affairs at Yale, and Nick Zamiska, head of corporate affairs and legal counsel at Palantir Technologies.

Tsyvinski opened the session with a discussion on how forecasting abilities can be improved through precise probabilistic estimates, rather than simple yes-or-no predictions, and by continuously updating one’s assessments in light of new information. He posed the question of whether the rapid rise of big data and AI might lead to a genuine revolution in forecasting.

Zamiska stressed that while the output and performance of large language models are garnering the AI headlines, “the unglamorous task” of integrating disparate datasets remains a substantial challenge. Drawing on Palantir’s work ranging from logistics systems to intelligence reports, he highlighted how unexpected correlations — such as moon phases influencing roadside bomb placements — can emerge when diverse data sources are effectively merged.

He added that, despite advances in large language models, U.S. intelligence agencies are increasingly turning to smaller, proprietary AI models to keep sensitive data secure in-house. Noting the recent release of Chinese AI platform DeepSeek, he argued that the models themselves may become commoditized over time, while the data-integration layer and AI-driven applications could prove most valuable.

Both speakers discussed the role of “the wisdom of crowds” in forecasting and innovation. Zamiska pointed to occasions where crowd wisdom has faltered, such as major automakers overlooking electric vehicles until Tesla’s emergence. Tsyvinski observed that people often underestimate “fat tail” events, which are more probable than commonly assumed, and expressed interest in “small models” specialized for particular domains.

A recurring theme was how to attract top engineering and scientific minds to public service. Zamiska noted that, historically, bold government initiatives like the Manhattan Project successfully rallied elite talent, but he argued that today’s public-sector salaries often cannot compete with the private sector, challenging innovation. Tsyvinski emphasized that collaboration across academia, government, and industry is essential to achieve breakthroughs.

Concluding the discussion, the speakers examined the growing influence of tech firms in public affairs. Zamiska raised the question of “importing the best of Silicon Valley” while still ensuring democratic oversight. Tsyvinski noted that “top-down” societies can address predefined tasks effectively, but bottom-up approaches and public-private partnerships are critical for fostering long-term innovation.

The Blue Center’s lecture series pairs practitioners from government and industry with Yale professors to deliver joint, hour-long lectures on subjects of mutual expertise. Practitioners discuss events or policy initiatives in which they participated, and academics cover related insights or results generated by their research.

For details on other events in the series, subscribe to the Blue Center’s mailing list.