Hope For Inclusive AI
Artificial Intelligence is revolutionising our lives, reshaping our workplaces, improving efficiency and altering our decision-making processes. While I'm admittedly scared of this technological wave, turning away from it is just not an option. Instead, I'll take a deep breath, acknowledge my fears and hopefully get a sense of comfort embracing the change.
One of my concerns is AI’s impact on equality and diversity. If the algorithms rely on historical data to make predictions and decisions, and if this data reflects gender stereotypes or inequality, we will perpetuate and amplify the existing biases.
Thinking of my day-to-day life, if I am using a recruitment platform powered by AI that’s trained on historical data that favours male candidates, I might unknowingly be exacerbating gender discrimination. This scenario happened with Amazon in 2019 when they had to shut down their hiring tool after discovering it favoured applicants using words like “executed”, which was more prevalent on male’s CVs.
Gender bias issues were also found in search tools. Carnegie Mellon University revealed that Google’s online advertising system was displaying high-paying positions more frequently to males than to females.
When AI technologies are primarily developed by white-male-dominated teams, the resulting products and services tend to neglect the needs and perspectives of women and other marginalised groups. No breaking news here; similar disparities have happened in various fields, including medical and pharmaceutical research. In March 2024, the Australian Department of Health and Age Care released “#EndGenderBias Survey Summary Report,” detailing the experiences of women regarding barriers and biases in the system. But despite past experiences, these biases persist within the health sector when using AI. For instance, computer diagnosis tools have demonstrated lower accuracy rates for black patients. Shouldn’t we know better by now?
Moreover, unequal access to technology deepens existing disparities, ranging from gender gaps to wealth inequalities between countries, communities, and individuals. Are we advancing into an era of Tech Colonialism? How long will we continue perpetuating the gap between the haves and the have-nots?
Tech companies driving AI innovation are driven by profit, not social responsibility. While some may take proactive steps to mitigate bias in AI algorithms, such as conducting audits, employing diverse development teams, and ensuring transparency, it is crucial for governments to establish and strengthen legislation in this domain. Global organisations can play a pivotal role by advocating for these improvements and catalysing discussions around control mechanisms, effective AI governance, and frameworks guiding its development so we can progress toward a more equitable, inclusive and fair society. We cannot keep fostering marginalisation.
Instead of headlines like “AI will affect 40% of jobs and probably worsen inequality, says IMF head”, “8 Times AI Bias Caused Real-World Harm”, “UNESCO finds ‘pervasive’ gender bias in generative AI tools” and “Ageism, sexism, classism and more: 7 examples of bias in AI-generated images”, I am looking forward to reading more news such as “3 Key Ways Artificial Intelligence Helps Organisations Drive DE&I”, “3 Ways ChatGPT Can Be Used To Advance Workplace DEI”, “20 Ways Technology Can Further Diversity And Inclusion Efforts”, “How cognitive diversity in AI can help close the disability inclusion gap” and “AI For Equity: Bridging Racial Gaps With Artificial Intelligence”. There is hope.
In the meantime, I will employ AI under supervision, doing my best to leverage its potential responsibly with empathy, fairness, and a keen awareness of biases.
