The ‘World-System-AI’ Cybernetic Intelligence Alignment Problem скачать в хорошем качестве

The ‘World-System-AI’ Cybernetic Intelligence Alignment Problem

Speaker: Max Ramsahoye Session Description: This talk explores a redefinition of the AI alignment problem, expanding it beyond isolated technological systems to include the broader collective intelligence that drives AI development. The speaker introduces the concept of the "Cybernetic Intelligence Alignment Problem," situating AI within the context of the global socio-technical system—referred to as the (capitalist) world-system—viewed as an emergent superintelligence requiring alignment. The presentation critiques the suitability of current power- and profit-driven institutional designs as safe frameworks for AI governance. It argues for reframing AI alignment as inherently tied to the alignment of the world-system itself. The talk concludes by proposing a new subfield, "Critical Systems Catastrophic Risk," which integrates critical and systems theory to address the generative dynamics of global risks and establish solutions for aligning collective and artificial intelligences as a unified, adaptive system. Talk's presentation is available at https://docs.google.com/presentation/...