My research focuses on the mathematical, ontological, and game-theoretic foundations of safe Artificial General Intelligence (AGI) and existential hope — bridging formal models of agency with real-world safety and governance.
1. Ontological Anti-Crisis (OAC) Theory
Investigating how advanced AI systems represent reality, agency, and goals when their world models undergo fundamental ontological shifts — and how to preserve alignment across conceptual revolutions.
2. Real-World Risks of LLMs Through the Lens of Ontological Completeness
Examining failure modes, value alignment dynamics, sabotage reasoning, and mechanistic architectures of frontier language models and agentic platforms.
3. Safe AGI Game: Mechanism Design and Computational Complexity
Studying multi-stakeholder incentives, compositional game theory, undecidability boundaries, and governance architectures for safe and beneficial AGI development.
