Previously at Google DeepMind for 7 years. Co-led Scalable Oversight work at Google DeepMind, worked on AlphaFold 2 and 3, and involved in Post-training, Information Quality, Prompt Injections, Uncertainty Quantification, and Offline RL efforts.
Co-led the Human Data Engineering organization at Google DeepMind, helping curate high quality human and AI data for use in model training and evaluation across a wide variety of capability and safety research. Previously worked on autonomous vehicles at Waymo.
A former corporate/M&A lawyer at one of the world’s most prestigious law firms, Alex has spent the past decade advising and investing in numerous high-profile startups, several of which achieved unicorn status, through the consultancy he founded. Passionate about creating positive change in the world, he has built multiple non-profits and cross-sector coalitions — winning an award for his work tackling violence against women and girls, and recognised by the British Medical Association for establishing “life-saving” initiatives during the Covid-19 pandemic.
Previously spent 7 years building production systems at Google and Hulu, then worked on multi-agent orchestration for an AI-powered web design studio. Most recently researching human–AI complementarity as a SPAR and MARS fellow, building LLM inference pipelines, training routers, and probing model hidden states.
Most recently worked on data tooling for investigative journalism at OCCRP. Prior to that, 5 years at Google DeepMind working on RL research for Robotics, with a focus on dexterous manipulation capability.
Renchi Zhang is a robotics and AI researcher focused on how people interact and collaborate with intelligent systems. His doctoral research at TU Delft explored human–robot interaction and differences between human and AI information processing, combining behavioural experiments with practical robotics systems. His research also spans AI evaluation and human-AI collaboration, investigating when human guidance can support AI reasoning and decision-making.
Software engineer with a background in physics. Previously a primary contributor to BlueDot Impact’s AI safety course platform, and Tech Lead for the Online team at the Centre for Effective Altruism.
Jessica is an operations and strategy leader who specialises in building and scaling organisations at the intersection of AI, technology and public interest. She has worked across fast-growing AI ventures, research and government-facing environments, building the operational infrastructure, governance and cross-functional processes that help ambitious teams execute effectively. She has previously worked with PauseAI UK, Apart Research and The London Initiative for Safe AI (LISA).
Brad Cao is a student at Columbia University studying mathematics and computer science. At Sampura he works on evaluations for human–AI judge systems. Before Columbia, he spent two years building machine learning models on clinical data at the University of Maryland, Baltimore.