We are delighted to bring you the first edition of the ‘Diversity Diaries’, a ten-part series dedicated to discussion, advice, tips, and foresight from women working in STEM fields. This week, contributors from Facebook AI, DeepMind, Affectiva, Samsung, the Alan Turing Institute and more provide their advice for women looking to pursue a career in STEM. The challenges which lay ahead for those when choosing a career in AI, Data Science and wider STEM fields are well documented, with studies suggesting that many perceive various glaring issues to be growing year on year. As an organisation that aims to promote the works of Women in STEM around the globe, we thought we would ask some of our friends in both industry and academia what their advice would be to those stepping onto the career ladder.
Georgia Gkioxari, Research Scientist, Facebook AI Research
Two things! First, it’s important for any young person starting their career in a technical field, like AI or Data Science, to realize that persistence is the biggest key to success. There will come a time where the technical challenges might be too overwhelming. Or your proposed solution to a problem is not producing the anticipated results, despite your best efforts. These are all positive things, not negative! It means that you have chosen to work on a challenging problem, you are pushing yourself beyond your comfort zone, and all you need is to persist and put more thought into it.
Second, and equally important, is to find the group of people that can support you, mentally but also technically. The best ideas come from talking through them and brainstorming with your peers. You also need mental support, people you can rely on, on your bad days but also on your best days.
Jane Wang, Senior Research Scientist, DeepMind
Find a mentor. They can be male or female, but should be someone who is willing to not only give you advice, but also be your champion in your company/field. If the first person you ask doesn’t fit the bill, ask them for suggested names and keep reaching out until you find someone. Another piece of advice I would give is to focus more on the problems you want to solve and the questions you want to ask, and less on the latest fads and tools. You can and should learn many techniques, but once you find those burning questions that really fascinate you, it will drive your entire career.
Sarah Laszlo, Senior Neuroscientist, X the moonshot factory
Believe you can learn. After that, the learning comes easily. It’s unfortunately all too common, in my experience, for women at the start of their careers to believe that something is “too hard” to learn; that they aren’t smart enough or numerate enough, or that they don’t have the right background or training to be able to learn technical skills. It’s almost never true. In my experience, it is much, much more common for a person to (falsely) think that they are unable to learn something, than for a person to falsely believe they can learn something that is in reality too difficult for them (this latter almost never happens). And, also in my experience, it’s harder to teach someone to believe that they have what it takes to learn something than it is to actually teach them that thing.
Throughout my career, I have taught and advised hundreds of students in computational areas, across the undergraduate, graduate, and postdoctoral levels. Too often, I would see women who were stars, but who didn’t believe they could do technical work or learn technical skills. Far more often, my sessions with female students were dedicated to helping them believe that they could do it — that there was no magical trick that they were missing — than to teaching them any particular technical tidbit. You CAN do it. Believe you can do it, and you can.
Jekaterina Novikova, Director of Machine Learning, Winterlight Labs
I think the most important thing for any person in any field is to do what you love. In this sense, women in AI/Data Science have a huge advantage over men – they are in this field only because they love it. It is not because of some stereotypes, not because “all girls are naturally good in Maths” or similar. Women are still a minority in AI so those who are there love what they are doing. Because of this my only advice would be – just continue doing what you love and don’t pay attention to anyone saying you are not able/capable/good enough to do something you want to do. Don’t ever give up.
Afsaneh Fazly, Director of Research, Samsung Toronto AI Lab
Trust your instinct, and have a growth mindset. It is ok to not know everything. In fact, most of us only know a little. Women tend to focus on what they don’t know more than what they do.
Alexia Jolicoieur-Martineau, AI Researcher, MILA
It’s hard to think of a single idea, but here are a few issues that I commonly see: 1) Don’t cling to ideas that failed, failure is a normal part of the process and is not actual failure on your part. Moving on quickly from an idea to another will make you more productive. 2) Don’t wait for permission before trying something new, do it first and then report the results to your manager/PI/advisor. Worst-case scenario, what you did is wrong, but at least you tried. Sometimes your manager/PI/advisor may be wrong, if you always ask for permission to try something new, you may miss opportunities. 3) If working, ask for a salary increase every year and don’t be afraid to negotiate during interviews. If paid lower than average, don’t buy the justifications made by your boss, ask for more.
Rana El Kaliouby, CEO, Affectiva
Develop strong skills to set yourself apart from other candidates. Beyond comprehending key approaches in machine learning and data science, you should look to areas of innovation in the field. For example, data synthesis is becoming incredibly important in advancing AI algorithms. If you educate yourself in innovative approaches like these, you’ll be a hot commodity in the job market and better prepared in your career. There are also a lot of concerns regarding ethics in AI. So it’s crucial to be an advocate for the ethical development and deployment of AI. And don’t just advocate, but also make sure that in all the work you do, you guard for ways that could inadvertently introduce data and algorithmic bias.
Finally, as a woman in tech, don’t be afraid to stand up for yourself. Speak up when you have ideas to contribute and don’t let anyone talk over you. Support other women in making them feel comfortable doing the same. We all need to advocate for and support one another if we want to build a strong ecosystem of women in tech.
Chanuki Seresinhe, Visiting Researcher, The Alan Turing Institute and Lead Data Scientist, Popsa
A lot of people in this industry blag about how good they are. I think this is an important thing to remember, as knowing that people exaggerate or are overconfident with how good they are means that you are probably as smart or even smarter than the people who might make you feel a bit intimidated!
Bianca Curutan, Software Engineer, Postmates
Use your network and resources. If there’s something you don’t know yet, ask questions – don’t think in terms of “I can’t” or “I don’t know how”. If there’s someone you would like to meet, contact people in your network for an introduction or reach out yourself. Similarly, be a resource for others to lean on as well.
Julia Kroll, Data & ML Engineer, Amazon
Change and uncertainty are inherent in quickly evolving fields like AI and data science. In my experience, women tend to be more risk-averse than men, and base their qualifications on what they have already accomplished rather than their potential. My advice is to recognize that many projects and technologies will be in yet-to-be-explored areas, and no one is qualified from the position of having done it before. Embrace the risk and experimentation and volunteer for new projects that help you learn and grow. That is how you can build a successful and exciting career!
In the next instalment of the Diversity Diaries, we will be discussing the ways in which our contributors think we can actively promote diversity in AI, Data Science and STEM Fields.