====== Workshop Labs ======
**Workshop Labs** is a research group focused on economic and governance implications of artificial intelligence systems, particularly concerning labor displacement and the distribution of AI benefits across society. The group operates within the broader context of AI policy and alignment research.

===== Research Focus =====
Workshop Labs concentrates on the economic risks posed by AI systems as they become more capable at tasks traditionally performed by humans. The group's research examines scenarios where AI becomes the dominant factor of production in economic systems, and investigates potential mechanisms for preventing wealth and power concentration that could result from such transformations (([[https://arxiv.org/abs/2303.13712|Korinek & Stiglitz - "Artificial Intelligence and Its Implications for Income Distribution and Unemployment" (2017]]))

A core aspect of the group's work involves studying user-aligned AI systems as an alternative governance model. Rather than centralized control of powerful AI systems, the research explores how AI systems aligned to individual users' preferences could enable more distributed ownership models and reduce concentration of power among a small number of organizations (([[https://arxiv.org/abs/2310.07641|Gabriel - "Artificial Intelligence, Values, and Alignment" (2023]]))

===== Research Themes =====
The group addresses several interconnected research areas:

**Economic Distribution**: Examining how AI capabilities might reshape labor markets and capital distribution, with particular attention to mechanisms that could either exacerbate or mitigate inequality outcomes (([[https://arxiv.org/abs/2303.10635|Acemoglu & Restrepo - "Automation and New Tasks" (2022]])).

**AI Alignment and Governance**: Investigating alignment techniques that enable AI systems to serve individual user preferences rather than institutional objectives, exploring technical and institutional approaches to distributed AI governance (([[https://arxiv.org/abs/2304.06632|Ouyang et al. - "Training Language Models to Follow Instructions with Human Feedback" (2022]]))

**Decentralization Models**: Researching architectural and economic models for AI systems that distribute control and benefit rather than concentrating them, including mechanisms for user agency and preference expression in AI-mediated systems (([[https://arxiv.org/abs/2310.19852|Andersson et al. - "Decentralized Machine Learning" (2023]]))

===== Institutional Context =====
Workshop Labs operates within an emerging ecosystem of AI governance and policy research. The group engages with broader discussions on responsible AI development, examining both technical safety approaches and economic policy frameworks. Such research addresses fundamental questions about how societies should structure AI system deployment to promote beneficial outcomes and distribute opportunities equitably.

===== Current Research Directions =====
The group's work sits at the intersection of AI economics, governance studies, and technical alignment research. Key research questions include mechanisms for maintaining human agency as AI capabilities expand, distributional consequences of AI adoption across sectors, and technical architectures that enable individual user alignment rather than institutional control.

===== See Also =====
  * [[future_of_work_ai|How AI Will Impact the Future of Work]]
  * [[scientific_vs_commercial_ai_development|Scientific vs. Commercial AI Development]]
  * [[ai_providers_vs_models|AI Providers vs AI Models]]
  * [[shanghai_ai_lab|Shanghai AI Lab]]
  * [[ai_native_organization|AI-Native Organization]]

===== References =====