Hassan Taher Discusses Embracing Diversity in AI

Shortly after its launch in December of 2023, Googles' new Gemini AI provided a perfect example of the challenges that so often go hand in hand with integrating diversity, equity, and inclusion (DEI) practices into artificial intelligence (AI) processes. When asked to generate images of America's founding fathers, Gemini produced people of color who only wore the clothing and hairstyles of those revered men. When asked to generate images of the Catholic pope, it, once again, got the vestments and hat correct, but failed to reflect the accurate skin color.

In short, Gemini seems unusually reluctant to create images of white people. In fact, it even confused the cultural heritage of Larry Page and Sergey Brin - the founders of the company that gave birth to it. While Page and Brin are both white men, Gemini depicted them as Asian.

But challenges like those epitomized by the Gemini debacle are simply bumps in the road that must be overcome. DEI is here to stay, for very good reasons, in nearly every segment and aspect of human society. And its place in the world of AI is both an inevitability and a necessity.

Diving into the heart of artificial intelligence, it's clear that diversity and inclusion are more than just buzzwords," says AI thought leader Hassan Taher.

They're the cornerstones of creating AI that's fair, ethical and truly beneficial for everyone.

A prominent writer, public speaker, and consultant, Taher has published countless articles on AI and related subjects. He is also the author of multiple books including The Rise of Intelligent Machines, The Future of Work in an AI-Powered World, and AI and Ethics: Navigating the Moral Maze. In addition to his regular invited presentations at global AI conferences and other important industry events, he offers advisement services as the founder and head of Taher AI Solutions. This firm allows him to work directly with government agencies and business leaders in fields that include finance, healthcare, and manufacturing.

In all his endeavors, Hassan Taher stresses the ethical and responsible deployment of AI resources and completion of AI projects. For him, DEI initiatives are integral to AI ethics. Beyond speaking and writing widely on the topic, he established a college scholarship fund for underprivileged students who are pursuing computer science degrees. The scholarship places a particular emphasis on supporting candidates from diverse cultural backgrounds.

Taher writes,

Despite its potential to revolutionize every aspect of our lives, the AI field currently suffers from pronounced gender and racial disparities"

Statistics reveal that women and people of color are significantly underrepresented in AI research and development roles."

Many of his fellow AI experts have echoed Hassan Taher's sentiments when it comes to DEI. As quoted in Time's online publication Charter, Princeton associate professor of computer science Dr. Olga Russakovsky lists correlation with socioeconomic groups and access" as well as lack of both visibility" and role models" as some of the root causes of the diversity challenges facing AI. She also places blame on implicit bias in hiring and recruiting" as well as issues surrounding the fear of AI.

Unfortunately, troubling DEI issues are contributing to public fear of AI in many ways, and examples of AI-related DEI problems abound. A 2022 study presented at the ACM Conference on Fairness, Accountability, and Transparency determined that AI-powered robots were more likely to associate African American men with criminal activity and women with homemaking. As reported by CNN, the New York City Department of Health and Mental Hygiene determined that its AI algorithms produced harmful outcomes for people of color. Health department officials released a statement explaining that these algorithms have been shown to overestimate the health status of people of color, which can lead to significant delays in vital medical treatment.

After it made headlines, AI experts rushed to offer their opinions of the recent Google Gemini debacle. In fact, Google CEO Sundar Pichai himself called Gemini's bias completely unacceptable" and further admitted we got it wrong."

In his analysis of Gemini's bias, Hassan Taher sees a stark reminder of how biases in AI can lead to public backlash and damage to a company's reputation." Tech companies that fail to properly account for diversity when developing their AI will be highly likely to regret it.

[Genini's] biased image generation sparked widespread criticism, prompting Google to acknowledge it's shortcomings and commit to making amends", writes Taher

This incident underscores the importance of diversity and inclusion in AI development, highlighting the need for comprehensive strategies to prevent similar issues in the future."

In Hassan Taher's estimation, AI's DEI problems begin with the underrepresentation of women and minorities in AI research and development (R&D). The logic is simple. R&D teams with a lack of diversity are far more likely to develop AI systems with biased algorithms and general DEI challenges.

Systems developed without a diverse set of perspectives are more likely to encode and amplify biases, leading to outcomes that discriminate against marginalized groups," contends Taher.

This imbalance not only stifles innovation but also raises concerns about the inclusivity and fairness of AI technologies. The lack of diverse perspectives can lead to AI systems that perpetuate existing biases, making it imperative to address these disparities head-on.

Striking a balance between the dangerous insensitivity of the New York City Health Department algorithm and the bizarre bias overcorrection of the Google Gemini platform will not be easy. But, for Hassan Taher,

Imperative to rectify this imbalance has never been clearer, prompting a closer examination of the importance of diversity in AI."

The implications of DEI efforts (or a lack thereof) are bound to have profound repercussions in every corner of the AI world. And given the increased prevalence of AI applications and DEI focus in human society, these two massive forces are bound to both compliment and conflict with each other for the foreseeable future. Therefore, AI professionals who fail to take DEI measures into account certainly do so at their own peril.

Diversity in AI is not just a moral imperative but a practical necessity,"writes Hassan Taher.

No matter how difficult, we must strive to get the DEI balance right in the AI industry.

Looking ahead, the road to diversity and inclusion in AI is long, but the direction is clear," writes Taher. It's up to everyone - companies, developers, policymakers - to keep pushing for a future where AI is as diverse as the world it serves. After all, the goal isn't just to avoid the next scandal. It's to create AI that's as groundbreaking, equitable, and inclusive as the world we aspire to live in."

The post Hassan Taher Discusses Embracing Diversity in AI appeared first on The Tech Report.