Agnosticism About Artificial Consciousness (arxiv.org)

A new paper argues for an agnostic position on artificial consciousness, contending that our understanding must rely on solid scientific evidence rather than intuition or speculation. The author, Tom McClelland, highlights the significant divide in current debates between those who are skeptical of AI consciousness (biological views) and those who lean towards its plausibility (functional views). He asserts that both sides overreach the available evidence, creating a dilemma for consciousness researchers: either pass judgment without sufficient backing or adhere strictly to evidentialism and remain undecided on AI's consciousness potential. This stance is particularly relevant for the AI and machine learning community, as it emphasizes the need for cautious interpretation of consciousness principles when applied to artificial entities. McClelland suggests that the extension of findings from conscious biological organisms to AI systems presents inherent challenges. By advocating for an agnostic perspective, he calls for a more nuanced approach to understanding consciousness in AI, urging researchers to rigorously evaluate their assumptions and claims in light of the complexities involved. This discourse could influence future research directions and the ethical frameworks surrounding AI developments.