Presenter
Nora Ammann
Nora’s work seeks to catalyse R&D that can help make transformative AI go well. She is currently a Technical Specialist at the UK’s Advanced Research & Invention Agency where she helps manage a £59m-backed programme which prototypes provable guarantees for safety-critical applications of AI. She has worked on hardware-enabled mechanisms for AI security & verification, co-authored a paper on systemic AI risk coining the term ‘Gradual Disempowerment’, and co-founded an organisation fostering interdisciplinary AI safety research. Nora has an interdisciplinary background spanning complex systems, political theory, philosophy of science, and AI.
Abstract:
As intelligence is becoming increasingly abundant, it will reshape the world in many ways—among others, it will lay bare new vulnerabilities to the complex systems and processes that support modern human life, or exacerbate existing ones. What does it take for our civilization to be resilient in the age of intelligence? We’ll examine key “attack vectors” through which intelligent systems could introduce fragility, and consider what technological and institutional innovations–including AI-enabled ones–we need to differentially accelerate to ensure our ability to prevent, mitigate, learn from, and adapt to adverse events. Resilience does not only promise safety—it is a necessary ingredient for sustained, open-ended progress.