Skip to main content

Christina Lu

Personal photo - Christina Lu

Christina Lu

Doctoral Student



evolutionary algorithms, AI-AI communication, online learning, multi-modal models, ecological models, embodied cognition, philosophy of computation


Christina Lu is an AI researcher and technologist based in London, via Cupertino and Shanghai. Her work combines rigorous technical research with speculative philosophical analysis and illustrative world-building to instantiate a more viable future with artificial intelligence. She is a doctoral student in computer science at Oxford, prototyping techniques for machine learning models inspired by the generative frictions of evolution. She is also an affiliate researcher at Antikythera, a thinktank for the philosophy of computation, hosted by the Berggruen Institute—where she was a researcher in the inaugural 2023 studio.

Previously, she was a software engineer at DeepMind, building systems for ML experimentation at scale and publishing sociotechnical research. Her work on the Whole Earth Codec, a proposal for a multi-modal foundation model trained on planetary-scale ecological sensing, and Vivarium, a vaporware platform for training AI in embodied cognition through modular "toy world" simulations, has been supported by Antikythera, Serpentine Arts Technologies, RadicalxChange, and PACT Zollverein. She is a frequent co-conspirator at Trust, where she mentors projects for Trust Quests.

She has given talks at Google People + AI Research (PAIR), Central Saint Martins, and Design Academy Eindhoven, and appeared on podcasts such as The Good Robot and Medialab Matadero's Synthetic Minds. Christina's research at Oxford is funded by the Scholarship for Women in Computer Science from the department and the James B. Reynolds Scholarship for Foreign Study from Dartmouth. She has a BA in Computer Science and Studio Art from Dartmouth.

Selected Publications

View AllManage publications