WOMEN IN AI?
1. Searching “unprofessional hairstyles” into Google, produces images of BME women.
2. Facebook’s advertisement services was exposed displaying certain job vacancies only to males.
3. Facial recognition software have much lower success rates in recognising BME women’s faces, than white men.
So where are the women in AI, and why does it appear that AI software discriminates against women?
Only 12% of AI researchers are female, yet the ONS predicts that of the future jobs lost to AI automation, 70% of those impacted and vulnerable will be female. AI will be no doubt be a critical and fundamental driver of economic growth in the next wave of technological advancement and industrialization. However If the creators of the algorithms which underpin AI are only accessible to the few, this becomes unrepresentative of societies demographic and ultimately facilitates further entrenchment of gender biases. How can we endorse a more equitable and diverse workplace when there are few females with access to the discourse creating these societal changes?
These insights were discussed on the Ethics Stage at the CogX Conference 2019 in Kings Cross. The discussion was led by many esteemed women currently working in AI. If ethical discussions disputing the failure of the AI workplace in embracing diversity are contested on a separate platform, how can we ensure that those with the resources to tackle the perpetuating inequality in this sector are aware of the call to action which must be made while AI capabilities are still in its embryonic stage.
Recently the government committed to an £18.5 million cash injection to “drive up skills in AI and data science and support more adults to upskill and retrain to progress in their careers or find new employment” (Government Press Release, 10th of June 2019). This stimulus also seeks to boost gender diversity in the tech sector. While the response no doubt acts as a primary step in influencing the social conditions of tomorrow, when France has committed to investing £1.5 billion, almost x81 times more in comparison, in reality, how impactful and effective is the UK’s efforts?
To fully reap the benefits of harnessing AI’s capabilities, we must create a fairer system that fosters its creation, one which embodies greater gender rights and equality. Only when the workplace is more representative of societies demographic, can we use AI to benefit more of societies demographic. To do so, these key discussion of ethics and fairness, must be translated into existing and emerging social policy and legislation.
The WLiAI Report 2019: 10 Principles of Responsible AI recommends 10 key policies for the government to implement to regulate and drive the development of Artificial Intelligence.
By Maya Spencer-Jacobs
University of Leeds