Biography
John is programme co-director of the Kinds of Intelligence. He is also a senior research fellow on the RECoG-AI project within Kinds of Intelligence.
Here John's work focuses on developing robust evaluation frameworks for AI systems in order to properly understand these systems’ capabilities and limitations. John also is a Senior Research Associate on the Solid Foundations project at the Centre for the Study of Existential Risk, where his work investigates the risks arising from ubiquitous, general-purpose AI systems.
John has a background in Computer Science, having completed his PhD at the University of York, as well as holding a Master’s degree from Oriel College, Oxford.
Resources:
Rethink reporting of evaluation results in AI
Harms from Increasingly Agentic Algorithmic Systems
How General-Purpose Is a Language Model? Usefulness and Safety with Human Prompters in the Wild
Inferring Capabilities from Task Performance with Bayesian Triangulation – (Preprint)
Back to people