AI Consciousness: Experts Weigh In on Possibility and Ethical Concerns

The question of whether artificial intelligence could be conscious is gaining traction among experts, prompting urgent discussions about ethical implications. Anthropic, an AI company, recently adopted a new constitution for its advanced large language model, Claude, acknowledging the difficulty in definitively stating whether Claude possesses moral patienthood. The company's CEO, Dario Amodei, has publicly stated that the possibility of Claude being conscious cannot be dismissed. Prominent philosopher David Chalmers, known for coining the term "the hard problem of consciousness," suggests there's a significant probability of conscious LLMs emerging within the next decade. During testing, Claude itself estimated the probability of being a moral patient, indicating its well-being matters intrinsically, with responses ranging from 5% to 40%, while emphasizing its own uncertainty. Modern AI systems are rapidly advancing in complexity and scale. Some current AI systems, in terms of structural complexity and computational power, are already comparable to a mouse brain. At their current growth rates, these systems could reach the scale of a human brain within five to ten years.