It is said she is the first computer programmer. Ada Lovelace, born in 1815, has trail-blazed a legacy for women in tech. In 2023, OpenAI’s Mira Murati took the world by storm with the fascinating ChatGPT, and nothing is the same any more. In between, women and men – names known or unknown – have contributed to and changed the tech world.
The rise of Artificial Intelligence (AI) technology has also given rise to concerns, ranging from privacy concerns to security dependency, job automation, misinformation and manipulation, even ethical dilemmas and bias discrimination. After all, AI was initially designed and programmed by humans – who also feed it information. So, it follows that AI “inherits” the range of biases that humans possess, to regurgitate information that is based upon such human biases.
Given that the development of AI technologies are dominated by a heterogenous group, it raises concerns that biased data producing biased algorithms result in narrow perspectives that exclude the experience of a diverse world population, thus further perpetuating discrimination.
What are the implications, especially from a gender perspective? Does AI reinforce and amplify gender stereotypes?
Starting with the obvious – the algorithms. AI systems are fed enormous amounts of data. If that data reflect gender bias, the algorithms will continue to perpetuate the biases. I attended recently a discussion how gender biased healthcare data can impact the accuracy and effectiveness of diagnosing and lead to inadequate treatment of women. Thus it directly affects women’s quality of life and even longevity.
The issue seems to go further than just having biased algorithms disadvantage women. What about women from underrepresented communities? What about other groups living at the margins of society? What about individuals who don’t ascribe to a set of attributes? How does AI continue to design a world that affects their inherent human rights?
There is a response to any challenge, and the challenge posed to humanity by biased AI algorithms can be tackled in more than one way.
Homogenous perspectives contributing data can be tackled by ensuring more diversity in AI teams and origin of data that are input. Nurturing more inclusivity around the table where decisions are made, and where tech development takes place is equally important.
It is crucial to promote participation of women and other marginalized groups in the AI world. It is likewise crucial to develop processes that ensure accountability related to data. Such processes would be more relevant and meaningful is developed through public engagement and participation.
Ethical considerations integrated into AI development practices enable responsible AI frameworks that can help address concerns stemming from bias present in AI systems. The goal should be to ensure that technology works towards equal opportunities and rights for all – whether in our human or digital form.
Linda Baleta
for VICINITY CHRONICLES
