In a recent MIT Technology Review Roundtables discussion, renowned journalist Karen Hao delved into her new book, ´Empire of AI: Dreams and Nightmares in Sam Altman’s OpenAI,´ examining the meteoric rise of OpenAI and the far-reaching consequences of rapid Artificial Intelligence development. Hao recounted her initial in-depth reporting on OpenAI that revealed stark contradictions between the public-facing mission of democratizing Artificial Intelligence and the company’s internal operations, a theme which has since grown more pronounced as OpenAI evolved into one of Silicon Valley’s most capital-driven corporations. Despite presenting itself as nonprofit and mission-led, OpenAI leveraged unprecedented fundraising and a shift toward large-scale model development, intensifying commercial ambitions while reshaping public narratives to suit strategic ends.
Hao described how foundational choices—such as prioritizing scale over fundamental research—cemented OpenAI’s competitive edge against giants like Google but also triggered profound costs. She underscored how this relentless pursuit of scaling up Artificial Intelligence has led to soaring energy consumption, acceleration of the climate crisis, increased freshwater demand, and displacement in water-scarce communities. As a result, the social and ecological costs now mirror those of historical empires, making parallels between Artificial Intelligence firms and colonial enterprises compelling. Hao drew connections between resource extraction, labor exploitation in vulnerable countries, the monopolization of research and expertise, and race rhetoric that pits ´good´ and ´evil´ empires, all reinforcing a global structure where a handful of actors wield outsized technological and economic power.
The conversation also addressed the ambiguities and religious-like fervor surrounding artificial general intelligence (AGI) within OpenAI and the wider Artificial Intelligence community. Citing varied and often contradictory definitions, Hao challenged both utopian and dystopian ´boomer vs doomer´ rhetoric, instead advocating for an accountability-oriented perspective that foregrounds democratic oversight and ethical alternatives. She argued for shifting focus from massive general-purpose models to well-scoped, socially beneficial systems, such as using Artificial Intelligence for renewable energy integration. Hao urged both technologists and the public to demand greater transparency, regulation, and consent around Artificial Intelligence development, citing community-led initiatives like the Māori speech recognition project in New Zealand as inspiring models. Ultimately, she expressed optimism based on increasing public scrutiny and debate, emphasizing that collective action and local engagement are vital for steering Artificial Intelligence toward a more equitable and sustainable future.
