Announcing the Guaranteed Safe AI Seminars

Monthly series on AI systems with high-assurance quantitative safety guarantees

Horizon Events announces the Guaranteed Safe AI Seminars. This monthly series brings together researchers to discuss and advance the field of GS AI, which aims to produce AI systems equipped with high-assurance quantitative safety guarantees.

For further context, we recommend reading the Towards Guaranteed Safe AI paper.

We began the series earlier this year under the name ‘Provable AI Safety Seminars’. The name ‘Guaranteed Safe AI’ was established recently and reflects better our intention to discuss this broad research domain.

Our seminars so far have been:

• Gaia: Distributed planetary-scale AI safety – Rafael Kaufmann

• Provable AI safety – Steve Omohundro

• ​Synthesizing Gatekeepers for Safe Reinforcement Learning – Justice Sefas

• Verifying Global Properties of Neural Networks – Roman Soletskyi

For the second half of 2024, we aim to host sessions on the foundational perspectives, debate/criticisms, engineering projects, formal verification angles, and recent relevant works. Additionally, we will pilot mixer sessions with thematic breakout rooms, to empower the community with further connection.

We believe the field will stay relevant, so we are also committed to its continuation and success beyond 2024.

Participate:

• ​Follow the series

• ​​​​Apply to speak

• Watch video recordings

• Discuss on the ​​mailing list

• Give ​feedback

• Donate to support our non-profit efforts

• Consult the ​info website

Thanks for your interest and participation. Onward!