On July 17, the Pentagon awarded $800 million to OpenAI, Anthropic, Google, and xAI to develop advanced AI systems, including agentic models. What are the implications for global stability, AI governance, and AI safety?
Jason Yung traces the arc of AI’s integration into U.S. defense policy and explores what might lie ahead under Trump 2.0.
Many of our recent events have covered plausible scenarios where the development of highly advanced AI leads to very bad outcomes for humans.
At this event, Mario Gibney will discuss the various actions that people can take to decrease the risks and increase the odds of human flourishing in the coming years and decades.
Savannah Harlan explores how the rise of synthetic data is redrawing the boundaries of power at three levels: between governments and the public, between ordinary people and corporations, and between humanity and artificial intelligence itself.
We'll finish by examining whether synthetic data is likely to increase x-risk as we shift more of the work of alignment onto the models themselves.
LLMs are pre-trained on a large fraction of the internet. As a result, they can regurgitate private, copyrighted, and potentially hazardous information, causing deployment and safety challenges.
Lev McKinney will guide us through machine unlearning in LLMs—how models retain facts, methods for identifying influential training data, and techniques for suppressing predictions. Finally, we'll assess current research and its effectiveness for policy and safety concerns.
Pushing to develop transformative AI as fast as possible risks catastrophe—a concern often met with the objection that “if we don’t do it, China will.”
This “but China” objection often doesn’t give much consideration to diffusion: the speed and way in which AI is integrated into the military and economy. Wim Howson Creutzberg will explore how diffusion might change the stakes of racing towards transformative AI.
David Atanasov will speak about AI models' potential to increase human capabilities in biological scientific tasks and how this might increase certain biological risks.
We'll discuss what is being done about this and future plans for mitigating these risks.
Released just days after its American counterpart, China's new "Global AI Governance Action Plan" puts forth their take on advancing global AI development and governance.
Ajay Nandalall, who recently took us through the American plan, will explore how the two plans compare, and what changes these documents may drive in the global landscape of AI policy, risk, and safety.
Thursday, August 7th, 6:00pm - 9:00pm
Several labs are talking about building superintelligence, but can it be made safe with the tools that we have? What if we had weakly superhuman AGI to help us?
In this talk, Giles Edkins will explore some of the projects, ideas, and counterarguments in the quest to use AI to align AI.
Registration Instructions (IMPORTANT)
This is a paid event with limited tickets - you must RSVP on Luma to secure your spot.
If you can't make it in person, feel free to join the live stream at 6:30 pm, via this link.
Description
At this event, we'll take look at the recently announced U.S. AI Action Plan; a seemingly significant step in America's race to achieve global dominance in artificial intelligence.
Ajay Nandalall will walk us through the plan and help explore what it may mean for America, potential impacts to the global AI landscape, and the future of AI Safety.
Timeline
6:00 to 6:30 - Food & Networking
6:30 to 7:30 - Main Presentation & Questions
7:30 to 8:00 - Discussion
PLEASE RSVP ON LUMA!
Description
As AI approaches human-level capabilities, Mario Gibney explores "The Intelligence Curse"—how automation could make human labor obsolete and break the social contract between powerful actors and regular people. We'll explore possible strategies for building AI that augments rather than replaces humans, ensuring people remain relevant in the intelligence age.
Event Schedule
6:00 to 6:30 - Food & Networking
6:30 to 7:30 - Main Presentation & Questions
7:30 to 8:00 - Discussion
If you can't make it in person, feel free to join the live stream at 6:30 pm, via this link.