Igor Rudan's Ideometrics Paper Links Consciousness to AI Feasibility
Consciousness may be reducible to three criteria: attractiveness, feasibility, and potential impact.
Igor Rudan's new paper on arXiv (2606.04011) introduces ideometrics as a formal approach to understanding consciousness. Rudan argues that consciousness reduces informational entropy by internally simulating multiple possible futures and then acting to realize preferred states. This process relies on three fundamental criteria: attractiveness, feasibility, and potential impact. Feasibility and impact can be computed by non-conscious systems including AI, but attractiveness is consciously and emotionally experienced. This suggests a potential role for AI in modeling some aspects of consciousness while leaving subjective valuation uniquely human.
The framework also redefines time and space through the lens of consciousness. Time is intertwined with consciousness to provide causal ordering to external changes; without conscious beings, time may lack meaning. Space provides the structured field for ideas to have causal impact across scales. Dreaming, according to Rudan, may be remnants of earlier evolutionary stages of internal modeling. The paper, at 38 pages with 95 references, is a dense theoretical contribution that could influence future research on consciousness and AI alignment.
- The ideometrics framework uses attractiveness, feasibility, and potential impact as three criteria.
- Consciousness is defined as the mechanism that reduces informational entropy by simulating and acting on alternative futures.
- The model suggests that AI can handle feasibility and potential impact but not attractiveness, which requires subjective experience.
Why It Matters
A formal framework for consciousness could inform AI safety and understanding of subjective experience.