Governing and sustaining the AI Commons? Finding models to pass the pilot phase

Ramya Chandrasekhar, Michelle Thorne, Daniel Brumund, Solana Larsen

Click here to view the external video (YouTube). More information

Summary
Big Tech’s “one-model-fits-all” approach to AI development deepens data and knowledge inequities, automates oppression of data workers, and fuels the climate crisis. Federated AI commons offer a sustainable and participative alternative. How can we make this alternate vision a reality?
Stage 7
Panel
English
Conference

Big Tech dominates AI development. They extract data and cultural resources from individuals, communities and from the open web, without giving back to the maintenance of these resources and with no regard to data protection principles. They valourise a “one-model-fits-all” approach, which increases exploitation of data workers. And they pay no regard to the worsening climate crisis. In response, many stakeholders are advocating for ‘enclosure’. Global AI policy is driven by imaginaries of ‘threat’ and ‘competitiveness’. Content creators are adopting restrictive interpretations of intellectual property to limit sharing and re-use of data and content, which threatens the open movement. A global push is needed to combat Big Tech's digital colonialism, resist a monoculture of thought, and foster participative and sustainable AI. Our panel of technology researchers and activists from the majority world will explore federated AI commons as a solution. We’ll discuss data governance, labour, economics and ecology. We’ll also present some design, regulatory and political interventions to enrich and sustain the AI commons.

This programme session is supported by Stiftung Mercator. / Dieser Programmpunkt wird durch die Stiftung Mercator unterstützt.

Solana Larsen
Writer and Editor