Joanna J. Bryson
Lecturer in Computer System
University of Bath
Native Language: English
Scientific Field of Enquiry/Research: Artificial Intelligence & Behavioural Science
Date: 18 October 2003

http://www.cs.bath.ac.uk/~jjb

01. How would you describe the project/s you are currently working on to a non-scientist?
My work has two different aspects -- science and engineering. The science I do is to try to build working models of natural intelligence to test whether those models are actually viable. Models represent theories. Working models have to be very complete, it's much harder to make unstated assumptions when you build something that runs.
In particular, I build models on computers (sometimes in virtual reality) of non-human primate intelligence. I have two sets of experiments going on right now, one on how monkeys learn new tasks, and the other testing some theories about how colonies of monkeys are politically structured, and whether this is dependent on how many conflict resolution strategies different species have evolved.
My engineering research is about how to make it easier to build these AI models. Some of my students are using this to work on building game characters or do computer animation.

02. Where, as a scientist, do you see the creativity in your work and/or life?
Being a scientist is entirely about being creative -- when you are trying to understand the world you have to be very creative, though within certain constraints (like any media.) And like almost any kind of art, there are some tedious repetitive bits. But coming up with theories is creative, trying to understand the papers you read / other people's ideas is creative, writing computer code is creative, building models is creative, writing papers is creative.
I also enjoy music (esp. `alternative', folk and 20th century eastern european classical) and fine art (esp. modern & contemporary art.) I also try to live in beautiful places, because that keeps me much happier.

03. When you embark on research, do you start with a defined goal or does the goal emerge through the process?
You start with one goal. Usually several other goals emerge in the process. Getting anything done requires focussing on one longer than I like to -- it takes real self control. But I always try to record the other goals & they sometimes get met later.

04. With regard to your research, how would you describe the processes and outcomes in terms of the concepts 'natural' and 'artificial'?
I use artificial intelligence to understand natural intelligence.

05. Where do you gain inspiration and support for your work?
From nature & reading other people's ideas or hearing them speak. The whole primate social behavior work came from my picking up a second-hand book called "chimpanzee politics", then I wrote the author & he told me about the current open research questions. The primate task learning came from hearing about a result that really surprised me because it was too neat, I couldn't believe it, so I replicated it then understood it better. The engineering came from reading about a cool way to build AI, then trying to replicate it and realizing it wasn't entirely right. I spent my PhD trying to improve that, and did at least enough to graduate!
I used to get support from my friends & colleagues, now I get it from my (and other people's) students. Anyone who is enthusiastic & interested. My partner was the most interested & most interesting -- that's why we're together!

06. How are possible applications from your research determined?
That's creative too -- I just notice problems that I can solve & try to solve them. I think most computer programmers are good problem solvers. But you do have to be interdisciplinary enough to notice and really understand a problem to be able to do that.

07. Do you share a language with other scientists (across cultures)? If so please describe how this language works.
No. I speak fluent American English, so I am very poor at speaking the international scientific language of broken English. I tend to be too idiomatic & talk to fast.
Seriously, colleagues who speak often about ideas adopt words to make that process go faster -- and start understanding the words differently as they come to understand the science better. You can't help this -- it's exactly how language works and what language is for. But to talk to other scientists in other fields you have to read their papers until you understand *their* language so you can communicate to them. Sometimes errors in understanding lead to unnecessary arguments or ignorance, other times they lead to new innovation.

08. Does your work encompass or involve a possible benefit (tangible or intangible) to society? If so, how would you describe this benefit to non-scientists?
Absolutely. People often don't understand why they do what they do, or why their friends do what they do. Misunderstanding leads to all kinds of problems.
Some people don't think animal intelligence has anything to do with human intelligence, but if you have worked with these animals (or even read enough about them) you realize that huge numbers of human emotional motivations & relationships are shared with primates, though of course language makes other things very different.
On the engineering side, obviously being able to build AI is fun, useful & a possible means of creative expression. Again, some people think AI is a scary thing, but it's no more dangerous than any other technology, which is not to belittle the danger. It's to say that AI is like any other technology which needs to be understood & regulated as well as exploited.

09. Would you describe your future world view as pessimistic or optimistic? Why?
Agnostic. Depends. I guess I am fundamentally optimistic but sometimes intellectually pessimistic. In a way, these terms are meaningless; what matters is working for the best outcome you can. Like Jane Goodall, she has to respect the scientists who say the environment's already been destroyed & is just taking a while to die, but she's still dedicated the rest of her life to trying to save the habitat of chimpanzees in the wild. I think we all have to do that, we all have to work for the best outcome and stay politically engaged.

10. Do you believe you share this future world view with other scientists (in your field) or would you describe this view as personal?
Scientists are just like anyone else -- there are liberals, conservatives, and apolitical people. I suppose that scientists are a little more likely to share my viewpoints than just someone random who isn't a scientist because we share an education to some extent, but I wouldn't say it was very significant.