Just before the controversy surrounding his termination and subsequent reinstatement as the CEO of OpenAI, Sam Altman remarked in a recent interview, “We are progressing towards the finest world ever.” This statement prompted me to ponder, “Whose world was actually progressing towards being the best?” as an expert focusing on gender equality in the realm of information.
Upon closer examination, it became apparent that there is a lack of female representation in the world being shaped by the Altman team. Amid the turmoil surrounding Altman’s ousting, an analysis I conducted unveiled intriguing findings. For instance, a substantial majority of the 702 out of 750 employees who endorsed the petition for his reinstatement were men, mirroring a gender imbalance akin to that observed in McKinsey’s report on The State of AI in 2022.
In the aftermath of Altman’s departure, the newly constituted board of directors at OpenAI now exclusively comprises individuals of Caucasian descent, further exacerbated by a prevailing male dominance in executive positions. Where can one access the viewpoints of female AI experts and leaders regarding this pivotal Silicon Valley narrative?
I have harbored concerns about the involvement of women in shaping our AI-driven future and the discourse surrounding relational AI for some time now. Through data analysis and consultations with experts, it has become evident that women are significantly underrepresented in the realm of AI, whether as developers, information curators, or AI specialists.
Generative AI (GAI) processes massive datasets of text, images, and videos, historically skewed towards male-centric content, thereby perpetuating a narrative predominantly influenced by men. This inherent gender bias, compounded by the systemic disparities women encounter in society today, shapes the narrative surrounding GAI’s risks, limitations, and opportunities.
According to AKAS’ analysis of the GDELT Project’s global online news database, individuals are quoted 3.7 times more frequently than women in discussions related to AI in English-speaking nations this year. Merely 4% of news coverage on scientific advancements, technological breakthroughs, and funding discoveries revolves around women, as per the latest findings from the International Media Monitoring Project.
In a survey conducted by AKAS in April, it was revealed that only 18% and 23% of tech news editors in the UK and the US, respectively, were women. Men are three to five times more likely than women to dictate the narrative of a tech story.
The immediate repercussions of a narrative predominantly reflecting male perspectives pose significant risks. While long-term challenges posed by AI to society have been acknowledged, the critical absence of women and the neglect of their perspectives, needs, and experiences with AI demand immediate intervention.
Data from the Pew Research Center for 2022 indicates that US women exhibit 8% to 16% more apprehension than men regarding various AI advancements, ranging from medical diagnostics to task automation. Given their marginalization in the AI sector and limited representation in media discourse, women’s concerns are at risk of being overlooked in future developments.
Leslie McIntosh, Vice Chair of Research Integrity at Digital Science, aptly notes that if your viewpoint is never reported, you remain excluded from the narrative. GAI, rooted in traditional texts, is shaping our future, leaving significant gaps where children’s voices should be heard.
Nicholas Diakopoulos, a Communication Studies professor at Northwestern University in Chicago, underscores the importance of rectifying disparities in the depiction of race, gender, and various occupations in conceptual AI models. Careless utilization of such models by the media risks perpetuating biases ingrained in the training data.
Laura Ellis, Head of Technology Forecasting at the BBC, highlights the opacity surrounding the datasets used to train AI models, complicating the identification of prevailing biases in AI-generated content. The absence of critical inquiries into this matter is concerning.
Inquiries arise regarding the absence of prominent female figures in artificial intelligence. Lynette Mukami, Cultural, Search, and Analytics Director at Kenya’s Nation Media Group, asserts that discussions on AI often prioritize financial gain and efficiency over societal benefits, underscoring the significance of diverse voices in shaping AI discourse.
Prof. Maite Taboada of Simon Fraser University in Vancouver, leading the Gender Gap Tracker, expresses concerns about Sam Altman’s dominant presence in AI discourse following the launch of ChatGPT. This dominance, particularly in legislative matters, raises apprehensions about the marginalization of alternative viewpoints.
The overwhelming focus on Altman, as evidenced by AKAS’ GDELT study of 2023 international online media coverage, underscores his pervasive influence in AI discussions. Despite his return being hailed as revolutionary, questions linger about the voices critical of his rapid market-driven approach to GAI at the expense of human safety.
Efforts must be made to ensure that voices on the fringes of the AI industry, such as Helen Toner and Tasha McCauley (both removed from OpenAI’s board), are not disregarded. While the efficacy of “guard rails” (coding aimed at rectifying data biases) remains contentious, experts unanimously agree that AI has the potential to address diversity gaps.
Lars Damgaard Nielsen of Mediacatch emphasizes that what is measured can be managed, advocating for AI to monitor and combat racial and ethnic biases on the internet. By measuring women’s involvement in the discourse, AI can play a pivotal role in mitigating gender bias, underscoring the imperative to incorporate diverse perspectives in one of the most significant narratives of this era.
