I’m building a modular system to model value change in agents through social and emotional exchange.
Applications:
- AI alignment experiments in moral development
- Empathy and cooperation modeling
- Social exchange and reciprocity research
- Long-term value drift studies
Core design:
- Interactions quantified via give / take / receive / allow
- Empathy and energy balance weight decisions
- Base (long-term) and current (short-term) states
- Decay and growth thresholds for evolving morality
- Outcomes emerge from accumulated experience
Status:
- Backend: JavaScript + SQLite
- Real-time UE5 integration
- Narrative and symbolic overlay
- Ready for insertion of research-driven behaviors and logging
Seeking:
- Simulation scientists
- Cognitive modelers
- AI alignment researchers
- UE5 developers
Why collaborate:
- Theory-agnostic
- Swap in new models without rewriting
- Test how agents learn morality from interaction
DM or comment if interested.
