As a civil engineer with a PhD in sociology who now works in artificial intelligence (AI), my varied skill set and research background are perfect for this burgeoning field. If you’ve moved across disciplines, you could also be an ideal candidate to work in AI — and to make a difference, as I believe I am doing.
My dream of working as an academic civil engineer, but with a perspective beyond engineering, came true in January 2012, when I received an e-mail from the Australian National University (ANU) in Canberra that offered me a PhD scholarship in its sociology department.
At the time, I was a civil engineer working in Mashhad, Iran, and I was passionate about finding answers to sustainability challenges — especially those around water and energy consumption and the depletion of natural resources.
In my projects, which ranged from river-basin water management to urban water planning, I had always felt it was crucial to adopt a sociologist’s perspective to see beyond an engineer’s concept of ‘common sense’ solutions. Those would call for building a new dam or desalination plant, or increasing the ‘efficiency’ of water by upgrading irrigation infrastructure or reusing wastewater.
Looking at these solutions from a sociologist’s viewpoint, however, led me to a very different set of questions, such as: ‘Efficiency for whom, and for what purpose?’ ‘What about people who live nearby, and their history and culture?’
So I had turned down offers of admission to PhD programmes in civil engineering and had started to pursue social sciences — it was there, I felt, where I could make a difference to other engineers who were also searching for new insights from the social sciences.
Receiving a PhD scholarship from a non-engineering school made me feel like I had been granted a visa from an exotic place. I was confident that I would be happier in my career from then on. I could develop a new perspective by integrating my engineering and social-science skills.
But I’d failed to realize that such a big shift comes with its own challenges. As I considered the academic job market and the question of which academic ‘tribe’ I belonged to, I began to worry.
Perhaps the selection committee would see me as a civil engineer without a PhD in engineering, or as a PhD holder in sociology without a background in the social sciences. Although many academics were increasingly talking about the permeability of disciplinary boundaries, I was not sure that those boundaries were porous enough for me to pass through.
So as I neared the end of my PhD in 2017 and started to scout for job opportunities, I looked for positions in sustainability programmes — which, I knew, strongly support and encourage transdisciplinary work. In fact, sustainability was what took me from the world of engineering and mathematical modelling into the social sciences.
I thought I had a good chance of securing a post. I had completed multiple fellowships in sustainability during my PhD, published in sustainability and interdisciplinary journals and won many scholarships from prestigious sustainability organizations in the United Kingdom and Germany.
I was shocked when I received my first rejection letters. At that stage, I had no evidence to suggest that my skills could help me to fit into any discipline besides sustainability. So, right after I completed my PhD in January 2018, I applied with little hope for a research-fellow position at 3Ai, a newly established ANU institute focused on AI.
AI wasn’t a new topic for me. I had used machine-learning techniques for projecting levels of precipitation and for rainfall-runoff modelling. I had also worked on the social implications of emerging technologies as a research fellow in 2016–17 at the Harvard Kennedy School in Cambridge, Massachusetts.
Crucial skill sets
Nevertheless, I was surprised when I secured an interview, and intrigued when I learnt why I’d made the shortlist. 3Ai was not looking at developing AI technologies with legions of software engineers. Instead, it aimed to build a new field of applied science to manage the machines. I learnt that, to accomplish such a goal, transdisciplinarity was not just ‘nice to have’ — it was crucial. I started the position in March 2018.
Until I met my colleagues, I had considered my academic journey to be unusual. I was surprised to find that everyone at the institute had an intriguing mix of disciplinary skills: computer science and human geography, nuclear physics and journalism, public policy and medical anthropology, law and nanotechnology. The realization that I was far from alone in travelling between disciplines was eye-opening for me.
My colleagues and I launched an experimental master’s programme at ANU to develop, test and iterate a curriculum for this applied science with a group of hand-selected students who had diverse backgrounds ranging from psychology and social work, security ethnography and theatre directing to teaching, policy and defence.
Working with people who have different skill sets, who are open to different ways of thinking and doing things, gives me hope that AI research can be the new destination for those interested in transdisciplinary research and education.
Why? Because current events around digital technology, along with debates about biased algorithms and ethical and regulatory challenges of autonomous systems, underscore the fact that AI is more than a technology of the future. It is impacting today’s sociopolitical climate. More importantly, current events and debates highlight the fact that AI management is more of a social and political issue than an engineering challenge.
Transdisciplinary approaches here, I think, can help us to understand and navigate this socio-technical challenge. The challenges of AI research call for scientists from many disciplines who are actively working with policymakers, non-governmental organizations and communities, among other stakeholders.
I predict a new genre of training, research and education that includes hiring scientists and others who are willing to tackle the important questions around AI.
I also expect that this demand will push institutions, and universities in particular, to transform their approach towards researching and teaching AI, expanding scientists’ capabilities beyond machine learning and software engineering. Universities are already starting to invest in this area. In October 2018, the Massachusetts Institute of Technology in Cambridge announced a US$1-billion investment in computer and AI research. Stanford University in California, and the University of Cambridge and Oxford University in the United Kingdom, among others, have also started to build their own centres for AI research.
How can you join the AI conversation? Keep researching crucial questions about the human and environmental aspects of AI-powered technologies. Keep thinking about your skill sets and how they might be used for designing a future in which the use of AI is expanded.
Your skills in whatever discipline can still be relevant in a technical conversation. Keep stretching your mind in new directions.
This is an article from the Nature Careers Community, a place for Nature readers to share their professional experiences and advice. Guest posts are encouraged. You can get in touch with the editor at firstname.lastname@example.org.