AI Ethics Expert: The AI Myths You SHOULDN’T Believe

In the 'AI Ethics Expert: The AI Myths You SHOULDN'T Believe' episode of 'Slo Mo: A Podcast with Mo Gawdat,' Mo Gawdat converses with Rumman Chowdhury, an AI ethics expert and CEO of Humane Intelligence. The episode unfolds at the Sync conference in Saudi Arabia, which focuses on digital well-being and the challenges posed by the digital world and AI. Chowdhury, an MIT graduate and former US envoy of AI, shares her insights on the intersection of technology and humanity, and the need for a more engaged and ethical approach to AI development.

Mo and Rumman delve into the role of corporations in technology and their promises to society. They discuss the evolution of companies like Google, where the initial mission-driven culture gradually gave way to a focus on profits, leading to a potential disconnect between the companies' promises and their actions. Chowdhury emphasizes that corporations are just one part of the ecosystem and that fulfilling a humanistic vision cannot be purely corporate-driven.

The conversation shifts to the topic of AI and its reflection of humanity. Chowdhury explains that AI systems are built on data collected about human beings, often without the intention of training AI models. She points out the limitations of data from the internet, which is not fully representative of the world's diversity. The two discuss the challenges of creating AI models that are fair and contextually aware, given the complexity and fluidity of human values and cultures.

Rumman also touches on the risks of AI-generated content and the importance of critical thinking in the face of AI that presents synthesized information with confidence. Mo and Rumman explore the concept of intelligence, both in AI and more broadly, discussing the limitations of defining intelligence as merely the performance of intellectual tasks and the danger of ideological homogenization.

Throughout the episode, they underscore the importance of diverse perspectives and the role of individuals and communities in shaping the future of AI. Rumman introduces her organization's work on red teaming exercises and algorithmic assessment, aiming to foster a community that can critically evaluate and improve AI systems.