Research

My principle scientific passion is understanding cognition, particularly as it relates to explaining human culture, but also natural intelligence more broadly.  My main methodology for doing this is designing intelligent systems to model and test scientific theories.  We build theories of intelligence into working AI models.  Modelling allows us to learn more than we could using unassisted human reasoning about whether a theory is reasonable and what its implications are.  Once we understand a theory's implications and predictions, we can compare these to data empirical scientists collect from the natural system we are trying to explain.  Increasingly, we are collecting our own data too.

Most (but not all) of my research has focused on the unintentional and non- or proto-linguistic aspects of human intelligence, and how intelligence evolves more broadly.  The more we understand both the universals of cognition and computation, and the variation we see across species (particularly but not exclusively in animals), the better we will understand the context of specifically human behaviour, including human culture.  From 2000–2007 I worked primarily on understanding non-human primate behaviour.  Since 2008 my group has been more focused on characteristics of human cognition such as consciousness, artificial intelligence "feelings" (emotions), language, religion, and especially cultural variation in cooperation.  This doesn't mean we've lost interest in comparative cognition; my group now studies evolutionary and learning dynamics in many contexts, from public goods games in humans to instruction sharing in microbia, from evolvability and epistasis in gene regulatory networks to computer game strategies. We have looked at cognition and information sharing in species from macaques to tortoises to ravens to Mongolian asses.

Designing AI models of natural intelligence isn't as easy as it should be.  My research therefore has always included a great deal of work on systems AI (intelligence by design) including my work in action selection and the development methodology I initially developed at MIT,  Behavior Oriented Design (BOD).  We apply this work into a variety of domains besides science, including cognitive robotics, computer game characters, and intelligent environments / "smart homes". 

Given I work on both bettering AI and understanding human society, I feel obliged to also work on Robot and AI Ethics. This work there didn't initially seem like research but rather just public understanding of science, for example informing discussions on robot rights (tldr: anthropomorphising AI is the flip side of dehumanising humans.) However, it has become clear that humans have a lot of trouble understanding AI, partly because AI is built without sufficient concern for accountability, partly because humans over-identify with the computational aspects of our intelligence and thus inject a pschological block to AI transparency, and perhaps mostly because we just don't understand ourselves, including our ethics.

Consequently my group now also does empirical research in AI ethics. Much of our work concerns systems engineering and devops – making the use of AI more transparent and accountable. We also do work on HCI/HRI on what makes AI comprehensible to ordinary users, and we do both experimental and theoretical psychology on understanding what leads people to overidentify with AI.

Rob Wortham Holly Wilson Andreas Theodorou and Joanna Bryson shot by Alex Rotsidis

Research Group and Students

Research is done by researchers.  Many of the people I've written papers with have been affiliated with Artificial models of natural Intelligence.  These include my PhD and other dissertation students.   I have also been founder and "research leader" for Bath's Artificial Intelligence Group. "Research leader" is very much a Frederick Brooks' type of leadership-as-service role,  and has now been taken over by my awesome colleague Özgür Şimşek.

If you would like to do research with me, and

Other and Older Research Projects, Funding

I have been involved in promoting European Cognitive Systems research and education.  Some time ago I used to occasionally get around to maintaining this research-oriented list of Related Web Sites.  Even my really old Code is on line.  All of my code from published projects is available either there or from the Amoni Software Page.

Research projects and labs previous to Bath

Bath Projects and Funding Acknowledgement

Please see the AmonI Web pages for descriptions of projects and links to code.

J J Bryson
Last updated February 2019