This is an opinion piece. The views or opinions expressed in this blog are those of the guest blogger and do not necessarily reflect those of C Minds.
Author: Analisa Ruiz, Projects Intern at C Minds and Student of Economics, Urban Studies, African and African Diaspora Studies at The University of Texas at Austin
In an effort to improve diversity, Amazon developed an Artificial Intelligence (AI) algorithm that was meant to select the most qualified candidates for various jobs. However, in the end it only perpetuated the biased standard that has historically oppressed women. It was not the algorithm itself that was biased, but the data it had been fed. Simply put, garbage in, garbage out. Because Amazon’s hiring practices had favored men in the past, the algorithm learned to repeat this pattern, using gender indicators on resumes to discount qualified female candidates. Amazon has clarified that this algorithm was only ever used in trials, and has since been abandoned for a number of different reasons.
(This is not the first time AI has reflected human bias, for more information see ProPublica’s article on how machines replicated racial discrimination.)
AI learns through data, and up until now the data in employment is largely dominated by men. In 2017, women in made up only 36.5% of Mexico’s labor force. Imbalances in female participation are even more extreme in STEM environments. For example, studies estimate that women only make of 10-14% of the machine learning discipline. Diversity reports at Google, Facebook, and Uber, find that women only account for about 20% of their engineers. Luckily, awareness is growing, and companies are becoming more proactive about preventing discrimination in their work environments.
Artificial Intelligence (AI) has its place in all of this.
Some recruiters are using AI to analyze job descriptions and identifies words or phrases that impact which gender applies for jobs. AI discovered words like ‘supportive,’ or ‘committed’ often attracted more women in the application process, whereas jobs described as ‘competitive’ or ‘dominate’ resulted in more male applicants. In a case study with Johnson and Johnson, an additional 90,000 women applied to J&J after they changing their job descriptions using augmented writing technology. This technology uses large quantities of data from companies to reveal meaningful patterns in language. By recognizing these patterns in job descriptions, Johnson and Johnson was able to produce a stronger, more inclusive recruitment process.
AI has also been used to engage potential employees and encourage them in the process. An internal report by Hewlett-Packard revealed that women only applied for a promotion if they met 100% of the qualifications, while men in HP would apply if they met at least 60% of the qualifications. With AI chat boxes, AI can interact with candidates and even suggest jobs in which they can excel. Results from this technology have shown an increase in women applying for jobs they originally thought they were not qualified for, increasing the number of women hired in these respective positions.
In addition to all these measures AI can take to increase gender equality in the workplace, technology is also now being encouraged among girls from a young age. Nonprofits like Girls Who Code and Women in AI educate girls and women about technology and empower them to pursue their passions. In Latin America, Laboratoria provides a six month boot camp for young women who show a propensity for technology. Over 80% of women who graduated from this program in 2017 are now working in the tech industry. On average, the income of graduates in this program increased by 300%. These incredible results from Laboratoria are just some of the examples of what can happen when we support women in fulfilling their potential.
C Minds recognizes that while AI introduces a lot of new possibilities, it is nothing more than a tool. As Andrea Escobedo Lastiri, in charge of Government Affairs, Foreign and Public Policy at IBM, so wonderfully put at an event a couple of weeks ago, ‘It will be us humans who decide whether or not to be more inclusive, not machines.’ C Minds has made the conscious choice to make the Fourth Industrial Revolution an inclusive one, and we will continue to have conversations like this and ensure that the technology we use reflects the values we hold.
If you enjoyed this blog post, consider reading this opinion piece about everyday safety for in Mexico City’s public transportation system or this interview with Sonal Shah (Director and founder of the White House Office for Civic Participation and Innovation under former President Barack Obama) where she describes her view on using technology for social impact.