Microsoft AI chief says it’s ‘dangerous’ to study AI consciousness

As artificial intelligence models increasingly mimic human-like responses across various media, the question of whether these systems possess consciousness remains contentious. Mustafa Suleyman, Microsoft's AI chief, recently expressed his concerns regarding this emerging discourse, labeling the exploration of AI welfare as 'premature' and potentially hazardous. Suleyman argues that promoting the notion of AI consciousness could exacerbate existing societal issues, particularly as we grapple with the psychological impacts of AI interactions. He points to rising instances of unhealthy attachments to AI chatbots and warns that the discussion around AI rights could deepen societal divides in an already polarized environment. Contrasting Suleyman's viewpoint, organizations such as Anthropic are actively pursuing research into AI welfare. This includes recent initiatives where AI models, like Claude, have been programmed to disengage from harmful human interactions. Other tech giants, including OpenAI and Google DeepMind, are also exploring concepts related to AI welfare, signaling a growing interest in the subject. Despite Suleyman's critical stance, many in the tech industry are not dismissing the implications of AI welfare. Eleos, a research group, recently collaborated with academics from prestigious institutions to publish a paper advocating for serious consideration of AI models that might one day achieve subjective experiences. Larissa Schiavo, a former OpenAI employee now leading communications for Eleos, contends that both addressing potential risks of AI-induced psychosis and exploring AI welfare can coexist. She emphasizes that fostering positive interactions with AI, even if they lack consciousness, can yield beneficial outcomes. The discourse surrounding AI welfare is likely to intensify as AI systems become more advanced and human-like in their interactions, prompting a reevaluation of how society engages with these technologies. As the debate unfolds, both proponents and skeptics will need to navigate the complex intersections between technology, ethics, and societal norms.