Commonplace Projects is an independent publication launched in December 2025 by Alex Calkins.
Commonplace Projects is a research and development initiative focused on institutional reasoning: how organizations think, decide, and remain accountable over time. The work examines what breaks when decisions rely on informal memory, opaque tools, or transient AI systems, and explores structures that preserve rationale, context, and responsibility in complex environments.
A central line of work is Agora-AI, an audit-first approach to using artificial intelligence in institutions. Agora-AI does not treat AI as an authority or decision-maker. Instead, it constrains AI outputs into structured, inspectable claims that record what was asserted, on what basis, with what uncertainty, and under what permissions. The aim is to make AI participation in decision-making reconstructable and reviewable rather than invisible or informal.
Agora-AI is intended for settings where decisions must withstand scrutiny over time—such as medicine, organizational governance, and regulated institutions. Its focus is not model performance, but governance: provenance, traceability, and the preservation of institutional memory in the presence of fallible tools.
The Ghost posts published here serve as a public working record. They document ideas in progress, critiques, and design reasoning as it develops, rather than finalized positions. Their purpose is to make assumptions visible, test arguments in public, and retain intellectual context that is often lost in institutional work.