What does a system modify when it modifies itself?
📰 ArXiv cs.AI
A system modifying itself can target low-level rules, control rules, or norms, lacking a formal framework to distinguish these targets
Action Steps
- Identify the targets of self-modification in a cognitive system, including low-level rules, control rules, and norms
- Develop a formal framework to distinguish between these targets
- Apply this framework to understand the implications of self-modification in cognitive systems and AI models
- Evaluate the effectiveness of this framework in explaining executive control, metacognition, and hierarchical learning
Who Needs to Know This
AI researchers and cognitive scientists benefit from understanding the targets of self-modification in cognitive systems, as it can inform the development of more advanced AI models and improve our understanding of human cognition
Key Insight
💡 A formal framework is needed to distinguish between the targets of self-modification in cognitive systems
Share This
💡 Self-modifying systems: what do they modify? Low-level rules, control rules, or norms? #AI #cognitivescience
Key Takeaways
A system modifying itself can target low-level rules, control rules, or norms, lacking a formal framework to distinguish these targets
Full Article
Title: What does a system modify when it modifies itself?
Abstract:
arXiv:2603.27611v1 Announce Type: new Abstract: When a cognitive system modifies its own functioning, what exactly does it modify: a low-level rule, a control rule, or the norm that evaluates its own revisions? Cognitive science describes executive control, metacognition, and hierarchical learning with precision, but lacks a formal framework distinguishing these targets of transformation. Contemporary artificial intelligence likewise exhibits self-modification without common criteria for compari
Abstract:
arXiv:2603.27611v1 Announce Type: new Abstract: When a cognitive system modifies its own functioning, what exactly does it modify: a low-level rule, a control rule, or the norm that evaluates its own revisions? Cognitive science describes executive control, metacognition, and hierarchical learning with precision, but lacks a formal framework distinguishing these targets of transformation. Contemporary artificial intelligence likewise exhibits self-modification without common criteria for compari
DeepCamp AI