“Inequality and disadvantage don’t really show themselves when it comes to teaching or student grades. It’s the circumstantial stuff where we see the difference,” says Dr Robert Craven, looking back on his first year of teaching computational science and artificial intelligence (AI) to students from underrepresented backgrounds.
Computer science courses are dominated by white male students, which seems remarkable as this didn’t used to be the case. Back in the 1940s women led the way in computer programming as software development was seen by men as inferior to developing hardware solutions.
Women made up roughly 26% of computer science professionals in the 1960s according to the New York Times1. But as the tech field grew so rapidly in the mid-20th century, companies had to hire thousands of workers. Recruiters favoured young white men with an aptitude for machines, over women and people from ethnic minorities, so creating the systemic imbalance that exists in today’s tech sector workforce.
Now, a growing number of universities and tech companies are addressing this imbalance and the need for greater diversity in tech.
The department of computing at Imperial College London is running a master’s degree and is actively recruiting students from underrepresented backgrounds. Supported by Deep Mind, a not-for profit team of scientists, engineers, machine learning, who are offering financial means and additional support to bridge the gender and ethnicity gap.
“There are students on the course I’m teaching that are having to work part-time to make ends meet and that is the big difference. Our master’s degree is very intense and lasts for the whole of the academic year.
Even without taking on other responsibilities outside the course our students find it very difficult to get things done on time. We don’t want our students to add another level of pressure and stress and we need to support them to try to get through that,”
Eloise Withnell, one of three female students on the Deep Mind scholarship, has benefitted from the additional pastoral care that is provided:
“Apart from the financial support that has allowed me to fully dedicate myself to my studies, I have found the mentorship particularly reassuring.
I’ve been allocated a female mentor who’s a successful computer scientist at DeepMind who works on health data. Talking to her made me realise that computer science doesn’t need to be this daunting and scary field.”
Fairness through diversity
“My project is all about explaining the model so that we can unpick and understand how decisions are made and root out bias. I think it is a really important area within AI, especially as models are being used more and more in real-world applications.”
Craven is hopeful that greater diversity in AI education will create a more robust sector:
“We see more students coming through who are interested in the idea of algorithmic fairness. It would be great if we could take more people through our master’s and PhD programs doing explicit work on this material.
These students will feed into research and development departments of the large companies and that will have an impact on the sector as a whole.”
This article supports one of the recommendations of a new report by independent thinktank Demos.
The report, Research 4.0, Research in the Age of Automation, seeks to understand what impact AI is having on the UK’s research sector and what implications it has for its future. The report recommends that the current post-16 curriculum should be reviewed to ensure all pupils receive a grounding in basic digital, quantitative and ethical skills necessary to ensure the effective and appropriate use of AI.
For more information about the impact AI is having on the UK’s research sector, read Research 4.0, Research in the Age of Automation. This new report is delivered by independent thinktank Demos and supported by Jisc.
- 1 The Secret History of Women in Coding - https://www.nytimes.com/2019/02/13/magazine/women-coding-computer-progra...
- 2 Scientometrics volume 86: issue 3 - How to improve research quality? Examining the impacts of collaboration intensity and member diversity in collaboration networks https://akjournals.com/view/journals/11192/86/3/article-p747.xml
- 3 Harvard Business Review: What Do We Do About the Biases in AI? https://hbr.org/2019/10/what-do-we-do-about-the-biases-in-ai