AI Ethics in Life Sciences: Universal Values and Cultural Diversity

A new paper on arXiv (2608.05436) argues that ethical concerns raised by artificial intelligence in life sciences and health research are real but not unique to AI. The authors contend that all sciences should be governed by values rooted in human brain architecture and socialization, not by anything specific to AI. They highlight that the human brain's computational architecture is far less costly than that of machines, achieved through a global neuronal workspace and a reward system described as a continuous cycle of wanting, liking, and satiety, rather than a quantity to be maximized. This creates a tension between the universality of ethical judgment and the diversity of morals. While brain networks for global workspace and emotion are universally shared, the content is shaped by epigenetic appropriation, leading to cultural diversity. The paper emphasizes the need for an 'architecture of care' in AI ethics, respecting cultural differences while upholding universal principles.