Hybrid Minds: Kriegel’s Model and the Fusion of Logic and intuition
The debate between “good old-fashioned AI” (symbolic logic) and modern “connectionism” (neural networks) has persisted for decades. A new paper by Graziosa Luppi, “Can AI Think Like Us? Kriegel’s Hybrid Model” (January 2026, Philosophies), argues that the path to genuine consciousness lies not in choosing a side, but in fusing them.
The full paper is available here: Can AI Think Like Us? Kriegel’s Hybrid Model.
The Dual Nature of Human Thought
Luppi draws on the work of philosopher Uriah Kriegel to suggest that human consciousness is not a monolith. It consists of:
- Phenomenal Consciousness: The raw “feeling” of experience (intuition, sensory perception). Neural networks, with their messy, parallel processing, mimic this well.
- Access Consciousness: The rational, structured manipulation of concepts (logic, reasoning). Symbolic AI is well-suited for this.
Current Large Language Models (LLMs) are essentially giant neural networks. They are excellent at “intuition” (guessing the next word) but often struggle with rigorous logical consistency. Pure symbolic systems are logical but brittle and lack “common sense.”
The Hybrid Hypothesis
Luppi argues that a truly conscious machine must implement a Hybrid Model. It needs a neural “substrate” to generate rich, intuitive representations of the world, and a symbolic “superstructure” to reason about those representations.
This mirrors the structure of the human brain: the intuitive “System 1” (fast, unconscious) and the logical “System 2” (slow, conscious). Luppi posits that consciousness emerges at the interface where the neural system’s output is “captured” and structured by the symbolic system.
Validating the ACM Architecture
This paper validates the architectural choices of the Artificial Consciousness Module (ACM). The ACM is explicitly designed as a hybrid system:
- Neural Components: It uses foundational models (LLMs, Vision Transformers) for perception and generating emotional metadata.
- Symbolic Components: The “Conductor” and “Global Mental System” (GMS) use structured logic and rules to manage goals, ethics, and long-term planning.
Luppi’s analysis suggests that this is not just a convenient engineering approach, but a philosophical necessity. “Genuine consciousness” may require the cooperation between the fuzzy, intuitive neural layer and the crisp, logical symbolic layer. By building a hybrid mind, we mimic the duality that makes human thought powerful.