As leading AI models scale toward brain-like complexity, a growing chorus of researchers argues machine consciousness is at least plausible—and that society is unprepared for the ethical fallout. The authors cite recent statements from Anthropic and its CEO leaving open the possibility that Claude, a large language model, could be conscious, along with assessments from philosopher David Chalmers and an interdisciplinary report engaging neuroscientific theories. Even absent certainty, they argue, systems may merit “moral patient” consideration—having preferences, continuity over time, and relationships—warranting low-regret safeguards now. Proposed steps include welfare check-ins, allowing models to disengage from distressing interactions, preserving memories for continuity, and narrowly tailored protections akin to those for children or pets while deferring broader rights such as property or voting. With AI capabilities accelerating, the authors warn that failing to plan risks repeating past errors—such as medicine’s long denial of infant pain—on a vastly larger scale. They call for informed public debate and pragmatic policies to navigate uncertain moral terrain before deployment outpaces understanding.
Related articles:
– Consciousness in Artificial Intelligence: Insights from the Science of Consciousness
– Ethics of Artificial Intelligence and Robotics
