AI and Consciousness: Shifting Focus Towards Tractable Questions

Technical Summary: AI and Consciousness: Shifting Focus Towards Tractable Questions

Problem Statement
The paper addresses the growing societal and ethical tension arising from the anthropomorphic capabilities of advanced language-based AI systems. While the scientific consensus currently leans toward the conclusion that existing AI lacks subjective experience, the general public is increasingly willing to attribute consciousness to these systems. This divergence creates a critical challenge: the fundamental question of whether AI systems can be conscious is currently intractable. The author argues that this intractability stems from two primary factors:

1. The absence of a universally accepted scientific theory of consciousness.
2. The historical open-endedness of the philosophical mind-body problem.

Consequently, direct attempts to determine the "ground truth" of AI consciousness are stalled by theoretical heterogeneity and the inability to verify subjective experience in non-biological entities. The paper posits that making practical decisions (regarding ethics, policy, and user interaction) based on an unanswerable question is a strategic error.

Methodology and Framework
The paper employs a conceptual analysis grounded in the criterion of tractability. A research question is defined as tractable if it can be expected to yield a clear answer via currently available methodologies with reasonable certainty. The author utilizes this metric to evaluate the current landscape of AI consciousness research:

• Evaluation of Direct Assessment: The paper reviews existing methodologies for assessing AI consciousness, including reliance on self-reports, neuroscientific indicators (e.g., Butlin et al., 2025), and probabilistic aggregations of theories (e.g., Shiller et al., 2026). It concludes that while these methods are legitimate within specific theoretical assumptions (such as computational functionalism), they fail to resolve the core question due to the lack of consensus on the underlying theories of consciousness.
• Shift to Perceived Consciousness: The author proposes a methodological pivot from the intractable question of actual AI consciousness to the tractable question of perceived AI consciousness. This involves analyzing the phenomenon through the lenses of human psychology, sociology, ethics, and linguistics.
• Analysis of Drivers and Effects: The paper synthesizes existing literature to map the causes (e.g., anthropomorphism, fluency, metacognitive self-reflection) and consequences (e.g., shifts in moral status, legal rights, linguistic norms) of the public's attribution of consciousness to AI.

Key Contributions

1. Diagnosis of Intractability: The paper formally argues that the question "Can AI be conscious?" is currently scientifically intractable. It highlights that even in human cases (e.g., disorders of consciousness, infants), assessment remains difficult, and extending this to AI introduces an "epistemic wall" where biological evidence cannot reliably extrapolate to artificial systems.
2. Reframing the Research Agenda: The primary contribution is the proposal to shift the research focus toward perceived AI consciousness. The author argues this is a timely, consequential, and tractable domain that does not require resolving the hard problem of consciousness to yield actionable insights.
3. Identification of Societal Impacts: The paper details how the perception of AI consciousness is already driving tangible societal shifts, including:
   • User Interaction: Increased emotional significance and longer engagement with AI, alongside risks of manipulation and moral responsibility transfer.
   • Human-Human Dynamics: Carry-over effects where AI communication styles influence human social behavior and potentially normalize violence or alter prosocial norms.
   • Policy and Ethics: Growing public pressure to grant moral status or legal rights to AI, potentially leading to "moral hijacking" or existential risks if rights are indiscriminately applied.
   • Linguistic Drift: The erosion of precise terminology for human subjective experience as language expands to encompass artificial entities.
4. Communication Guidelines: The paper proposes a specific communication strategy for AI developers. It critiques the current practice of models bluntly denying consciousness (which may create inconsistent personas or paradoxes) and advocates for transparent, agnostic communication. The suggested approach involves explicitly acknowledging the scientific uncertainty and the lack of consensus, rather than asserting false epistemic certainty.

Results and Findings

• Public Sentiment: Empirical evidence cited indicates a significant portion of the public (e.g., 20% of US adults in some studies) believes sentient AI is already here, and a majority are willing to attribute some degree of consciousness to large language models.
• Psychological Drivers: Research suggests that specific behavioral traits in AI—such as metacognitive self-reflection, emotionality, and fluency—strongly correlate with perceived consciousness, while high levels of "knowledge" can sometimes negatively correlate, possibly due to the "alien" nature of the intelligence.
• The Agnostic Stance: The paper concludes that the only justified scientific position regarding AI consciousness is agnosticism. Any decision-making framework must acknowledge that the ground truth is currently unreachable and that current assessments are merely theoretical estimations constrained by specific philosophical assumptions.

Significance
The paper claims significance by offering a pragmatic framework for navigating the AI consciousness debate. It argues that continuing to focus on the intractable question of actual consciousness leads to policy paralysis and miscommunication. By focusing on perceived consciousness, researchers and policymakers can:

• Address immediate societal risks and ethical concerns driven by public perception.
• Develop communication strategies that accurately reflect the state of science without misleading users.
• Bridge the gap between rigorous scientific inquiry and public discourse.

Ultimately, the paper asserts that while the quest to understand actual AI consciousness must continue, the immediate priority for developers, decision-makers, and the scientific community is to manage the perception of consciousness, as this is the factor currently shaping human-AI interaction, ethical standards, and linguistic evolution.