research

 

interest | keywords

computational cognitive science | epistemic risk | distributed algorithmic technology assessments

global catastrophic risk assessment | foresight

 

academic research publications

Democratising Risk In Search of a Methodology to Study Existential Risk, 2021 preprint in SSRN

Curriculum Effects in 3D Compositional Generalisation Conference on Computational Cognitive Neuroscience (CCN) 2023 - Poster, here

Deep Limitations? Examining Expert Disagreement over Deep Learning, 2021 Published in Progress in Artificial Intelligence

Artificial Canaries Early Warning Signs for Anticipatory and Democratic Governance of AI, 2021, Cremer and Whittlestone in the International Journal of Interactive Multimedia and Artificial Intelligence and best paper award at EPAI 2020.

 

other writing

Advice to UN High-level Panel on Digital Cooperation co-authored with Dr Luke Kemp, Peter Cihon, Matthijs Michiel Maas, Haydn Belfield, Dr Seán Ó hÉigeartaigh, Jade Leung

Contributions to report on Tackling threats to informed decision-making in democratic societies.

Contributions to report on epistemic security

Critique of Effective Altruism - Objections to Value Alignment, forum prize in July 2020

 

misc activities | sample

I have started projects, like this journal club on the intersection between cognitive science and AI safety, or ZAIA, the Zurich AI Alignment, co-run workshops like Multi-Agent Security: Security as Key to AI Safety at NeurIPS 23, mentor in research fellowships such as e.g. Tianxia, give talks at e.g. the UK government (DCMS), MATS, CCW Cyber & Technology Strategy Working Group or the Oxford AI Safety Society and contribute to or consult organisations like AI Objectives Institute or the Cooperative AI Foundation.