World models, trust-native AI, and policy wind tunnels—explained for leaders, not labs.
About Simon Muflier
Simon Muflier is an IEEE member, researcher, consultant, and the author of The Architecture of Trust, exploring how AI systems must evolve from opaque, probabilistic models into verifiable, governable, trust-native architectures.
At The Oyez, Simon helps companies, governments, and institutions understand and deploy AI systems that are accountable by design. His work focuses on three core pillars:
- Knowledge Engines — systems that replace brittle RAG pilots with structured, governed retrieval and reasoning
- World Models — simulation frameworks that let leaders test strategies and policies before they impact the real world
- Trust-Native Architecture — cryptographic provenance, verifiable retrieval, auditability, and governance baked directly into the AI stack
Simon works at the intersection of technical architecture, institutional decision-making, and policy. He translates advanced AI system design into frameworks that leaders can use to make safer, more strategic decisions.
Speaking topics
1. From RAG to Knowledge Engines
Ideal for: Engineering orgs, CTOs, enterprise AI teams
Most organizations deploy AI using brittle RAG systems that hallucinate, drift, or fail silently. This talk explains:
- Why RAG breaks at scale
- What a “knowledge engine” is and how it differs
- How provenance, schemas, and trust layers stabilize AI systems
- A practical 5-layer “Trust Stack” leaders can demand from vendors or internal teams
Talk Title:
From RAG Hacks to Knowledge Engines: Designing AI You Can Actually Trust
2. World Models & Policy Wind Tunnels
Ideal for: Governments, policy leaders, corporate strategy, think tanks
World models are becoming the defining architecture of the next decade — not just predicting text, but simulating economies, infrastructures, policies, and organizational decisions.
This talk covers:
- What world models are, and why industry is shifting toward them
- How policy wind tunnels allow leaders to test decisions before real-world impact
- Simulation governance: preventing bias, overconfidence, or misuse
- Practical adoption patterns for governments and large companies
Talk Title:
World Models as Policy Wind Tunnels: Crash-Testing Strategy and Regulation
3. Trust-Native AI Architecture
Ideal for: Boards, regulators, public-sector agencies, risk teams
AI governance fails when it’s layered on top of opaque systems. Trust must be architectural.
This talk explains:
- The difference between trust theater and trust-native design
- Cryptographic provenance, verifiable retrieval, and secure enclaves
- Evaluation harnesses and audit trails
- How to build AI systems that regulators and courts can scrutinize
Talk Title:
Trust-Native AI: How to Build Systems You Can Audit, Govern, and Defend
4. The Post-Parameter Era
Ideal for: Industry executives, journalists, innovation teams
The scaling race is reaching diminishing returns. The future belongs to:
- Small, specialized models
- World models as cognitive infrastructure
- Hybrid systems with verifiable logic
- AI portfolios that emphasize governance, not just capability
Talk Title:
Small Models, Big World: What Comes After the Scaling Race
Sample Questions
- What exactly is a world model, and why are companies betting their futures on them?
- How can governments use world models to safely test policies before implementation?
- What is “AI trust theater,” and how do institutions avoid it?
- How can boards identify when an AI system is governable vs. dangerous?
- Why do most RAG systems fail, and what should replace them?
- What does “trust-native architecture” mean in practice?
- What is the “post-parameter era,” and how should leaders prepare?
SHORT Bio
Simon Muflier is the author of The Architecture of Trust, a foundational text on the future of AI governance and architecture. An IEEE member and consultant, Simon works at the frontier of world models, trust-native design, and institutional decision-making.
As a founder at The Oyez, he collaborates with organizations seeking a deeper understanding of AI systems and how to deploy them responsibly. His work focuses on knowledge engines, policy wind tunnels, and architectural approaches to trust and safety. Simon speaks internationally on world models, AI governance, and the transition toward a post-parameter era of intelligent systems.
CONTACT
For all booking or interview inquiries please contact: research@theoyez.org