Should We Use Language Models In High-Stakes Decision-Making Scenarios? @CISACatStanford

Uploaded April 2024 | Updated September 2026, 3 days ago

Speaker: Max Lamparth

The emergence of generative language models, such as the one powering ChatGPT, has sparked widespread interest due to its potential implications for the future of work and society at large. The drive to automate decision-making is reaching high-stakes applications like military applications and mental health care, where non-zero error rates lead to individual failures with dire consequences and the potential to cause wide-scale harm. Thus, it is time to scrutinize and ask whether we should use language models in high-stakes decision-making scenarios. In this talk, I will dissect the question by studying how human decision-making differs from language models, if language models add their own dynamics to conflict situations, and whether they can recognize emergency/high-stakes user queries.

Should We Use Language Models In High-Stakes Decision-Making Scenarios?

Recent Advances in the Development of an Alternative Global Rare Earth Supply Chain

$How Much is Enough to Kill a Nation? Great Power Nuclear Deterrence in a New Era of Countervalue Speaker: Jaganath Sankaran A fundamental premise of the nuclear revolution theory (TNR) is the claim of assured destruction—the ability of a state to retaliate with a nuclear second-strike that leads to the destruction of the adversary’s sociopolitical-economic-industrial infrastructure, denying it the ability to survive as a viable modern nation-state. However, as we enter an era of renewed strategic great power competition, emerging technological advances have reanimated questions about the continued relevance of TNR. Can a state employing emerging technologies significantly undermine the assured destruction capabilities of its adversary? Using insights and techniques from Self-Organized Criticality theory, Dr. Sankaran analytically reexamines and models the requirements for assured destruction. He demonstrates that the networked structure of critical infrastructures continues to make advanced industrial states extremely vulnerable to assured destruction—at a fraction of Cold War arsenal requirements. Dr. Sankaran argues that advanced industrial nation-states remain vulnerable to assured destruction retaliatory strikes.$

The Technological Revolution in International Humanitarian Law

The New Geopolitics of Energy | Agnia Grigas

Under the Nuclear Shadow: China’s Information-Age Weapons in International Security

Low-Yield Nuclear Weapons and Crisis Dynamics

Conference pt. 2 | The Future of Decision Making

Security Through Cooperation: Space, Nuclear Weapons, and US-Russia Relations after the Cold War

From Local to Transnational: When Armed Groups Target Across Borders