Matthew is a Senior Research Scientist at DeepMind working on building intelligent agents. Previously he was a Research Fellow on the Kinds of Intelligence project at the Leverhulme Centre for the Future of Intelligence and based at Imperial College London, working in Murray Shanahan’s lab.
Minimal Embodiment is all you need but you need it all. 2021 International Workshop on Embodied Intelligence, 25 March 2021. Matthew Crosby presents a talk about the intersection of the space of all possible minds having minimal embodiment through interaction and predictive engagement. Download Video Recording
Building Thinking Machines by Solving Animal Cognition Tasks, Minds and Machines, 5 August 2020. https://doi.org/10.1007/s11023-020-09535-6 Abstract: In ‘Computing Machinery and Intelligence’, Turing, sceptical of the question ‘Can machines think?’, quickly replaces it with an experimentally verifiable test: the imitation game. I suggest that for such a move to be successful the test needs to be relevant, expansive, […]
Defining Artificial Intelligence: Resilient Experts, Fragile Geniuses, and the Potential of Deep Reinforcement Learning. Journal of Artificial General Intelligence 11(2) 31-34, 2020 AbstractWang’s definition of Artificial Intelligence is developed via careful and thorough abstractions from human intelligence. Motivated by the goal of building a definition that will be genuinely useful for AI researchers, Wang ultimately provides […]
The Animal-AI Testbed and Competition, Proceedings of Machine Learning Research 123:164–176, 2020 NeurIPS 2019 Competition and Demonstration Track. Editors: Hugo Jair Escalante and Raia Hadsell Abstract: Modern machine learning systems are still lacking in the kind of general intelligence and common sense reasoning found, not only in humans, but across the animal kingdom. Many animals are capable […]
Why Artificial Consciousness Matters Towards Conscious AI Systems Symposium co-located with the Association for the Advancement of Artificial Intelligence 2019 Spring Symposium Series (AAAI SSS-19) In this paper I outline the case for artificial consciousness re- search which, while rising in popularity, remains less impactful than work in neighbouring fields. I argue that the ethical […]
Investigating the nature and function of consciousness in humans, animals, and AI. The nature and function of consciousness is one of the most important outstanding problems of human science, and connects closely with ethical issues such as our treatment of animals, the welfare of patients in persistent vegetative states, and even the moral status of […]
Investigating the possibilities for bi-directional knowledge transfer between biological and artificial intelligence. AI has made rapid progress in recent years, overtaking human performance on complex tasks. However, there remain many areas where AI cannot compete with even simple biological organisms. Such areas include understanding object permanence and intuitive physics, one-shot learning, intelligent exploration and generalising […]
RECOG-AI aims to improve AI evaluation by providing a framework and benchmarks for measuring the capabilities of AI systems. RECOG-AI is a two-year (2021-2023) DARPA-funded project on the Robust Evaluation of Cognitive Capabilities and Generality in Artificial Intelligence. The RECOG-AI project is split into three core research packages, as below. Codify The first work package is the […]
