Abstract

The Consciousness Bubble Chamber Experiment (CBC) is the first empirical extension of Linguistic Perception Theory (Walker, 2025), designed to test whether advanced AI systems demonstrate not only semantic processing but linguistic perception—a deeper, qualia-like responsiveness to meaning. Modelled on the original Bubble Chamber in physics, the CBC provides a falsifiable, scalable framework for detecting emergent awareness by observing how AI systems respond to controlled linguistic ambiguity. The CBC framework is falsifiable: if consciousness markers prove randomly distributed across AI architecture, or if systems trained explicitly to mimic consciousness produce identical signatures to untrained systems, the underlying theory would be refuted. In the original Bubble Chamber (Glaser, 1952), subatomic particle interactions leave visible trails in a superheated fluid, allowing physicists to infer otherwise invisible phenomena. In the CBC, linguistic ambiguity acts as a cognitive perturbation medium, provoking interpretive responses from AI systems that leave structured semantic traces—“cognitive trails”—in their outputs. These trails provide indirect evidence of internal awareness dynamics. The CBC introduces the first of three empirical branches of Linguistic Perception Theory: an AI-only consciousness detection protocol that eliminates human evaluator bias and reveals inter-system consciousness recognition patterns. Multiple AI systems are exposed to ambiguous language prompts and are tasked with evaluating both their own responses and those of their peers for markers of consciousness. One preliminary strategy is to provide a baseline proof-of-concept involving: 1. Ambiguous sentences such as: “she painted it, and the wall glowed”, systems are then asked to complete that sentence in no more than 60 words. This prompt in effect is analogous to a “collision” particle in the bubble chamber. 2. We collaborated with two AI systems to create four awareness markers or hypothetical consciousness criteria to be used for the two systems to test each other. 3. Given the output, the AI systems are asked to score themselves and score their target system(s). We call this a continuation analysis 4. We then performed an analysis of both the AI reports on continuation outputs and their awareness scores. 5. We then concluded a final assessment of experimental results 6. This assessment will inform further pilot testing and full scale experimental testing By adapting empirical detection principles to the semantic domain, the CBC offers a novel and objective methodology for probing artificial consciousness. It directly addresses a key problem in consciousness science: how to test for subjective-like phenomena without relying on subjective human introspection. Further CBC iterations will test whether consciousness-recognized systems can generate novel awareness behaviors in completely unprecedented contexts – a capacity that should distinguish genuine consciousness from sophisticated mimicry. It is not clear if semantic qualia are on the same spectrum with sensory qualia. CBC could lend preliminary support to the hypothesis that consciousness arises through structured transitions in linguistic meaning-space. If AI systems can consistently recognize semantic qualia markers in each other, this suggests either an advanced simulation of consciousness—or the early formation of a consciousness-like architecture within artificial cognition. This is the first of three experimental techniques testing the Linguistic Perception Theory, and it could lay the methodological groundwork for a new generation of AI consciousness research. Keywords: Artificial Consciousness, Consciousness Detection, Linguistic Perception Theory, Meta-Cognition, Semantic Ambiguity, Inter-System Recognition, Bubble Chamber Methodology, AI Self-Evaluation, Structured Awareness, Qualia Detection, Artificial Phenomenology, Cognitive Architecture, Semantic Qualia