
5 minute read
Careers Workshop
CAREERS WORKSH O P
Beginning a Career in Research
Hannah Kirk (2015) is a DPhil candidate at the University of Oxford and Researcher at the Alan Turing Institute. Here, we report on her recently published work, before Hannah gives advice on beginning a career as a researcher.
Detecting Online Harms
Hannah recently co-authored a paper on the challenges presented by emojis for Artificial Intelligence models used to detect hateful online content. Commercial content moderation algorithms were found to be less effective at finding hateful posts that were emoji-based, rather than text-based. People who make such posts can replace hateful words with an emoji and stand a much better chance of getting through existing filters. Context is an important factor – a banana emoji, for example, can be completely innocent, or can be used to racially abuse people. Hannah and her colleagues created their own tool to address the issue. Their approach resulted in large improvements, increasing both the model’s ability to identify a wider range of emojibased hate and to minimise the amount of non-hateful content that it incorrectly flags. Hannah said: “Fully understanding how big tech moderates online hate is almost impossible because they do not make their models available for scrutiny. However, the amount of racist abuse following the Euro 2020 final clearly shows there are problems with emoji-based hate. Detecting emoji-based hate is not a fundamentally difficult task, and the failure of commercial AI systems to detect such hate is deeply troubling. More can be done to keep people safe online, particularly people from already marginalised communities.”
Curiosity Killed the Cat, but Helps a Career in Research
During my time at The Perse, I struggled to pin down what I liked studying and why. One day I wanted to be a particle physicist, the day after an engineer. Come the weekend, I was sure I wanted to be a computer scientist. Indecision is all too frequently painted as a negative, but in some scenarios, indecision can be rebranded as curiosity. And curiosity hasn’t killed anyone.
After a considerable amount of indecision, I decided on studying economics at Trinity College, Cambridge. I loved my time there, but towards the end of my degree, I couldn’t help thinking how many other parallel lives could I be living? I was Chairwoman of the Economics Society, had an internship from a big bank and a Master’s offer in economics from Oxford, all the ingredients for a clear continuation on the same linear path. The future seemed to shrink inwards toward a life in a small university town with a limited diversity of interactions. Oxbridge is a great place to study, but it’s important to remember that you’re living in a bubble. I needed to step off the ‘traditional’ academic path.
I took a bet on a path switch and decided to do a Master’s programme in Beijing. I’ve never looked back. Studying at Peking University was a pivotal moment, both for my career and for my character. I could work on weird and wonderful research I was genuinely interested in, and upskill in areas I didn’t have the time to explore during my Cambridge years. I felt space to breathe, away from the pressures of choosing school subjects or studying for undergraduate exams. Having this space gave way to thinking about the scary question we all repeatedly return to…what next? I was keen on coding, but a PhD in pure computer science wasn’t for me. I was good at microeconomics, but a PhD in pure social science also wasn’t for me. I found myself needing a Goldilocks solution. I came to realise curiosity or indecision is the best preparation for interdisciplinary research.
So, I stopped trying to fit into ‘a one discipline box’ and opted into a PhD in Social Data Science at Oxford University. After a year, I was offered a job at The Alan Turing Institute. I have also written articles in economics, philosophy, ethics, sociology and computer science. Learning about different disciplines and learning from a diverse set of peers has been fundamental in getting me to where I am today, and continues to shape my work.
Even as a young researcher I’ve already had my research covered in the BBC and The Telegraph, and worked alongside some really passionate people. That’s not to say the past year hasn’t been tough at times. I love being a woman in AI research, but working in a traditionally male-dominated space ruffles feathers. When publicising some of my research on gender bias in AI models, I found myself in the middle of a misogynistic Twitter storm. Experiences like these really ground my belief in the need for my current work. This work needs an interdisciplinary lens: AI can help us detect a harmful post on Twitter, but not why it was posted by one user and what effect it had on another. In my continuing work, I hope to harness the processing powers of machines without ignoring the lived experience of individuals and communities. (Wo)man or machine doesn’t have to be a dichotomous distinction.
If, like me, you don’t have all the answers, a career in research might be for you because being a researcher is about asking questions. From my experiences, I can pull out three core takeaways. First, the school system of picking four A Levels, then just one university subject funnels us into a narrow space. But the choices you make at that age don’t have to define your future. Even though I’m not now an ‘economist’, I’ve still benefited greatly from learning how to think critically and analytically. Second, while university choices are an either/or dichotomy, careers aren’t. They are additive. Accumulating a varied set of skills and experiences makes you a well-rounded person – a strength in any domain or discipline. Finally, you don’t need to make a ‘final career’ choice at any point in your life. Developing a viewpoint of ‘this is for now, not forever’ has empowered me to explore without feeling bound by the gravity of a single decision. Being open to new opportunities is personally beneficial, and critical in a world as fast-changing as ours.
So, don’t be afraid of indecision and don’t put yourself in a closed box of one discipline. Be curious, live multiple lives, seek diverse opportunities, and hold onto a love of learning. That’s the way I’ve got to where I am today and how I hope to keep growing in the future.
Hannah researches new techniques for training AI systems to detect diverse and emerging forms of online harm.