Narjes Bufferin on AI and Bias: A New Perspective
Introduction
Narjes Bufferin, founder and co at Key Text AI, firmly believes that artificial intelligence is transforming every facet of life, from banking to ecommerce. But, she questions whether these advancements, including artificial intelligence, come with ethical, diversity, and inclusion challenges. In her view, AI could help organizations foster harmonious relationships with the people they serve. This is why she founded Key Text, where she offers a disruptive platform that empowers brands to leverage customer feedback for a superior customer experience.
The Impact of AI & the Question of Bias
Narjes weighs the pros and cons AI in daily life. She loves how AI helps her beat traffic on Waze or find a new song to enjoy on Spotify. But as a woman entrepreneur with an immigrant Muslim Arabic background, she recognizes how AI can introduce unconscious biases into our society.
With the goal of stimulating a discussion on AI diversity and AI bias, she offers a fresh perspective on how we can use AI to counteract bias and build a better society.
The Conundrum of Bias in AI
She explains that AI can result in biased decisions, mirroring human biases. She raises questions like:
- Is the unfairness stemming from humans or machines?
- How can we use AI's benefits but prevent the reproduction of bias?
- Can we train AI to see the world as we want it to be, rather than as it is?
Addressing Human Bias in AI
Narjes takes a deep dive into these important questions and proposes innovative solutions. Tapping into her experience and expertise in the sector, she stresses that the responsibility for bias in AI lies with us, and that improving diversity in the AI community can offer new perspectives on bias.
The Role of Diversity and Fairness in AI
Organizations like Microsoft Feed and AI for All are already taking strides in increasing diversity in the AI industry. She believes that having diversity among those who create AI models helps to prevent bias. More diverse teams can implement multiple perspectives and approaches to mitigate bias.
Transforming AI Bias into Positive Impact
Narjes also champions the idea of using inherent AI bias to positively impact society. For instance, by skewing data used in AI models, these models could challenge their own biases to reach a fairer outcome. This approach, termed "counterfactual fairness", tests an AI model's predictions against a hypothetical world where sensitive, bias-prone attributes are changed.
She concludes by posing more thought-provoking questions that challenge traditional notions of diversity and bias.
Conclusion
Narjes's perspective on AI and bias invites us not only to see the risks but also the opportunities. While technical solutions are part of the answer, she believes that the human element in AI development is crucial. It is only through diverse teams, striving for fairness and employing innovative strategies, that we can ensure AI is a tool for positive transformation rather than reinforcing established biases.