Lennart Heim

Featured in: Favicon

foreignpolicy.com Favicon

csis.org

rand.org

Articles

The Rise of DeepSeek: What the Headlines Miss

2 months ago | rand.org | Lennart Heim

Recent coverage of DeepSeek's AI models has focused heavily on their impressive benchmark performance and efficiency gains. While these achievements deserve recognition and carry policy implications (more below), the story of compute access, export controls, and AI development is more complex than many reports suggest. Here are some key points that deserve more attention: Real export restrictions on AI chips only started in October 2023, making claims about their ineffectiveness premature.
Lawfare Daily: Janet Egan and Lennart Heim on the AI Diffusion Rule

Jan 17, 2025 | lawfaremedia.org | Lennart Heim |Janet Egan |Jen Patja

Published by The Lawfare Institute in Cooperation With Janet Egan, Senior Fellow at the Center for a New American Security (CNAS) and Lennart Heim, an AI researcher at RAND, join Kevin Frazier, a Tarbell Fellow at Lawfare, to analyze the interim final rule on AI diffusion announced by the Bureau of Industry and Security on January 13, 2025. This fourth-quarter effort by the Biden Administration to shape AI policy may have major ramifications on the global race for AI dominance.
Understanding the Artificial Intelligence Diffusion Framework

Jan 14, 2025 | rand.org | Lennart Heim

Availability: Web-Only Year: 2025 Pages: 39DOI: https://doi.org/10.7249/PEA3776-1 Document Number: PE-A3776-1 Heim, Lennart, Understanding the Artificial Intelligence Diffusion Framework: Can Export Controls Create a U.S.-Led Global Artificial Intelligence Ecosystem? RAND Corporation, PE-A3776-1, January 2025.
(Training) Compute Thresholds - Features and Functions in AI Governance

Apr 6, 2024 | blog.heim.xyz | Lennart Heim

Post-training enhancements can indeed significantly improve model performance. Regulators should be aware that AI capabilities might be substantially improved through post-training enhancements. We observe that, with current methods, capabilities can be increased to an equivalent of 5x to 20x increase in training compute through post-training enhancements.
To Govern AI, We Must Govern Compute

Mar 28, 2024 | lawfaremedia.org | Markus Anderljung |Lennart Heim |Haydn Belfield

Published by The Lawfare Institute in Cooperation With Computing power—compute, for short—is a key driver of artificial intelligence (AI) progress. Over the past 13 years, the amount of compute used to train leading AI systems has increased by a factor of 350 million. This has enabled the major AI advances that have recently gained global attention. However, compute is important not only for the progress of AI but also for its governance. Governments have taken notice.
Chinese Firms Are Evading U.S. Semiconductor Sanctions

Jun 21, 2023 | foreignpolicy.com | Tim Fist |Lennart Heim |Jordan Schneider

This past fall, the Biden administration moved decisively to cut off China’s supply of powerful chips, targeting the cutting-edge semiconductors used for supercomputing and artificial intelligence. However, recent reporting points to blacklisted Chinese entities—including China’s top nuclear weapons lab—gaining access to restricted chips through a combination of smuggling and renting through the cloud.
Video and Transcript of Presentation on Introduction to Compute Governance

May 17, 2023 | blog.heim.xyz | Lennart Heim

In February 2023, I gave an introduction to Compute Governance as part of a series of talks I gave at think tanks. All of the ideas are somewhat evolving, particularly what counts as fundamental properties. However, I think I'm gesturing in the right direction here. The recording is available here. In this talk, I present the idea of using computational resources (short compute) as a node for AI governance.