Gender Diversity in AI Growth

RECENTLY, I attended a talk on the rise of artificial intelligence (AI) organised by a company whose board I sit on. This session made me think about how AI’s growth will impact women in Malaysia.

One speaker shared insights on the growing number of data centres in Malaysia (49 at last count) and predicted this trend will continue due to our increasing reliance on technology. With AI-related stocks like NVIDIA soaring (its stock price increased by 194% year to date), AI is becoming a utility.

However, there’s not enough discussion in Malaysia about how AI might affect society, especially women.

AI presents both opportunities and challenges for women, particularly in patriarchal societies. The quality of AI output is only as good as its input. Here are some key issues:

Gender bias in AI algorithms: AI can inherit biases from its training data, reinforcing existing gender stereotypes. This can lead to discrimination in hiring and lending for example. AI is at best the input it receives. Under-representation in AI research and development: Women are under-represented in AI and technological fields, affecting the design and implementation of AI systems. For example, AI on social media often assumes women are interested in beauty products, while financial topics are targeted at men. Access to AI education and training: Unequal access to STEM education limits women’s participation in AI. Increasing access to STEM education for girls and women is crucial. In a world where the economy is growing more digital, continuous education and lifelong learning needs to be encouraged, perhaps the current tax incentive can be enhanced and future skilling of Malaysians can be provided for free or for little costs as how Singapore has done so for its citizens and PRs.

Gender pay gap: Women in AI and technology in general face wage disparities. In Malaysia, women earn 30% less than men. Closing the pay gap and promoting equal opportunities are essential to ensure that AI is inclusive and benefits all.

AI-generated content and vulnerability: Women already face complex social expectations. The risk of AI-generated explicit content adds to their vulnerability, potentially hindering their public and professional participation.

Social media’s psychological impact: has been shown to be more negative on young girls and women than on their male counterparts.

Ethical considerations: Ethical issues arise as AI evolves. Women’s rights, privacy and safety must be central to AI development and deployment.

Addressing Challenges

Promoting gender diversity in AI research, creating inclusive policies and ensuring AI benefits all society members is essential. However, with women under-represented in AI, tech and the labour force, who will regulate these ventures?

Currently, it does not appear that there’s strong political will to increase female labour force participation, even though it would benefit the economy.

Strategies

• Diverse teams: Ensure diverse teams develop and evaluate AI systems. Gender expertise should be welcomed. Diversity, equity and inclusion efforts in Malaysia should go beyond number of women in boardrooms to create a more inclusive society and more inclusive strategy and decision-making.
• AI and technology literacy training: Provide AI literacy training for gender experts and the public. - - Lifelong learning should be encouraged so decision-makers are equipped to make informed decisions.
• Data collection: Collect diverse and representative data and make it accessible to all. Address gender data gaps by including under-represented groups. In Malaysia, there’s insufficient public data on women investors, making it difficult to understand their challenges.
• Fairness metrics: Use fairness metrics during product development. Evaluate performance across different gender groups to ensure inclusivity.
• Bias-aware algorithms: Develop algorithms that actively mitigate bias using techniques like re-weighting, adversarial training, and fairness-aware learning.
• Collaboration: Collaboration between technology, gender studies and public policy experts is essential for creating accountable and unbiased AI systems that benefit women and prepare Malaysia for the future.

We know we want to get out of the middle-income trap but will we utilise our full capabilities and work towards a more inclusive workforce and deal with the rising challenges that technology brings? Something to ponder on.