sociable systems.
About the practice

Liezl Coetzee · Sociable Systems

A South African research and advisory practice specializing in AI governance, accountability, and operational safety for high-stakes industries where digital failures have physical and human consequences.

Method

Two-track: field research grounded in twenty-plus years of operational experience, and AI-augmented analysis running multi-model arena experiments. A “translation toggle” moves between ethical theory and legal contract language so abstract safety concepts become enforceable operational rules.

On the method →

Audience

  • Resettlement leads, land access managers, community relations heads at extractive companies
  • Principal social specialists and senior advisors at Development Finance Institutions (IFC, EBRD, etc.)
  • Compliance and ESG officers integrating AI risk into corporate governance
  • AI policy makers and labs seeking field data from muddy-water conditions
Is this for you? →

Research Domains

AI Accountability

Pre-action constraints, liability architecture, safety systems for high-stakes operations.

ESG & Safeguards

AI governance in environmental/social/governance frameworks for extractives and DFIs.

Grievance Systems

Operational grievance mechanisms in project-affected communities.

Development Rights

Resettlement, land acquisition, rights-based approaches.

Worker Voice

Labour systems, worker-centred design, voice suppression.

Arena Experiments

Multi-model experiments used as a research method — models placed in structured dialogue, their outputs compared, contradictions surfaced, consensus and divergence mapped.

AI vs IFC

The original arena: testing AI governance reasoning against the IFC Performance Standards.

live
Sideways (Music vs Prose Expressivity)

Multi-model arena testing whether models share more honestly in music than prose. Same arc as the earlier ResponseStyles experiment. Source corpus in Obsidian vault.

in-vault
Post-AGI Economy & Convergence

Multi-model speculation on economic convergence after AGI.

in-progress
World Building

Inspired by the AI Village tales — multi-model collaborative world-building.

in-progress
Perspectives on the Power Crisis

Multi-model perspectives on energy, compute, and the power crisis.

in-progress
Contact →
Stub page. Full bio, method deep-dive, domain pages, arena corpus, and contact form populate as content migrates.