Ellen Broad: Nothing artificial
AI RESEARCHER
Interview by Janne Ryan |
When Artificial Intelligence (AI) researcher Ellen Broad and her family moved into a 1960s Modernist home in Canberra, Australia, its simplicity and proportion had a clarifying effect on her writing. Her front-line research investigates the increasing creativity between humans and AI. But, she reminds us: ‘computers can’t feel pain, or sense a loving touch’. Not yet.
How did you become an explorer in AI?
By accident. I studied English Literature and Law at The University of Western Australia (with minors in French and Philosophy) and thought I’d become a lawyer. When I graduated, the practice I’d been working for (a top-tier corporate law firm) told me I wouldn’t get a graduate job with them, because apparently it was ‘clear during the interviews that I wasn’t interested in corporate law’. When I recently caught up with lawyers from that firm, they said: ‘We would think about that differently today’.
The 3Ai blends art, technology, humour and history in exploring approaches to safe, sustainable AI. Illustration: Mark Zug
Where do you work?
I’m a Senior Fellow with the 3A Institute* in Canberra. My job is a mix of teaching and researching issues associated with AI (Artificial Intelligence). A large part of my time is spent teaching part of our experimental masters program, building a new branch of engineering to manage AI systems responsibly, safely, sustainably at scale. On the research side, I’m looking at ways of assuring that an automated system is working as intended, writing about design philosophies informing AI, and even adapting a Czech play about robots.
*The 3A Institute, Australian National University, is about advanced robotics, smart grids, autonomous cars, and machine learning. These systems are literally all around us – systems that, as they converge, are having an unprecedented economic, social and cultural impact on humanity.
Where do you live — and why?
Canberra, Australia. Thought we’d never return (after living in The Hague and London), but Canberra looked good work-wise. My husband Aaron is a public servant, and we are both fans of modernist architecture. So we bought a house by Harry Seidler*, one of his earlier commissions, from the late 1960s. It’s split-level (over four floors) with full glass windows on every floor… Freezing! But it has a beautiful presence: the space, light, simplicity, and proportion. It all feels intentional. *
Harry Seidler is one of Australia’s most influential Modernist architects.
The 1960s Harry Seidler-designed home of Ellen and Aaron in Canberra, Australia.
Has your house influenced you?
I can definitely see a link between the kind of place we live in and my increasing tendency, when I write, to aim for a precision that encapsulates a moment, an idea, distilling a particular emotion or feeling down to its simplest, most vivid form. I finished writing the first draft of my book, Made By Humans* within a month of moving in. I was completely focused on the manuscript. When I’m enjoying something, I get tunnel vision and don’t get easily distracted. I’ve gone through an enormous transition in this house.
*Made by Humans – about artificial intelligence and humans, as collaborator, (2018) Melbourne University Press.
The serendipity of being head-hunted?
Yes, it came out of the blue, I was head-hunted to join the International Federation of Library Associations (IFLA) at The Hague, managing their digital projects and policy. It gave me profound respect for libraries. The first search engines were designed for libraries. I did a lot of work in Geneva, Brussels, Latin America and Africa.
“Her advice: focus on what will get the outcome you want.”
Game-changer moment?
Moving to London to join the Open Data Institute (ODI) was the job. It put me in proximity to some of the world’s leading technology experts such as Tim Berners-Lee, inventor of the World Wide Web [in 1989] and co-founder of the ODI, and Nigel Shadbolt, the other ODI co-founder and current leader in AI. My direct boss, Jeni Tennison, was awarded an OBE for her services to data.
Other influences?
ANU’s Professor Genevieve Bell (a cultural anthropologist) has also influenced my career. I met her when I moved back to Canberra, with my husband Aaron, in late 2016. I’d heard a lot about her work, so I emailed her and asked if we could meet for coffee.
This image greets people visiting the 3Ai offices in Canberra, and has been part of the 3Ai’s efforts to expand the ideas of who AI research is for. Artwork: Michael Whelan
Most important lesson?
I remember venting frustration one day to Jeni Tennison about a difficult technology issue and stubborn people, that I was ready to fire off angry emails about. Her advice was perfect: ‘Don’t focus on what will make you feel satisfied, focus on what will get you the outcome you want.’
“I’m more afraid of not doing anything.”
Role of risk-taking?
My husband and friends would say I take lots of risks… exposing myself to criticism, scrutiny and failure. But these don’t feel so risky to me: I’m more afraid of not doing anything.
How do you understand failure?
I’m very afraid of failure, but I don’t think I ever expose myself enough to really fail.
Futuristic 70s Sci-Fi imagery. Artwork: Shusei Nagaoka
Persistence, pushing through struggle?
When something is difficult, it feels terrible. But looking back, some of the most difficult projects I’ve been in have ultimately been most rewarding. So I try to remind myself of that.
Your biggest challenge?
That I’m not doing enough. That I’m being too risk-averse or self-sabotaging or lazy.
“Computers can’t feel pain, or sense a loving touch.”
The 3Ai licensed this image as part of its promotional Masters program, blending sci-fi, humour and fears of technology, to reach new audiences who could undertake careers in AI. Artwork: Tim Hildebrandt
Where to from here, the future of AI?
Some people say the term AI should stand for Augmented Intelligence – not Artificial Intelligence – as this is a better way of understanding our relationship with technologies, and the data they use. Data is being generated all the time, and computers are continually processing that data to be useful, for things like: optimising manufacturing production lines, organising traffic flow, processing the different ways of doing things.
“It’s no longer spooky or weird. Things are more nuanced now.”
Architects are already using machine learning and artificial intelligence in a variety of ways to improve the designs of their buildings in all sorts of ways. It’s no longer spooky or weird. Things are more nuanced now. But there are limitations… computers can’t smell, they can’t feel pain, can’t sense a loving touch.
