KI könnte zu „sozialen Brüchen“ zwischen Menschen führen, die sich über ihre Empfindungsfähigkeit nicht einig sind

Jonathan Birch, a philosopher specialising in sentience and its biological correlates, has stated that ‘social ruptures’ could develop in the future between people who believe AI systems are sentient, and therefore deserving of moral status, and those who don’t believe AI systems are sentient. There are parallels with humans’ relationships to AI chatbots; some people scorn them as machines incapable of human emotion but others have developed apparently deep and meaningful relationships with their chosen chatbots. Birch states that AI companies have been narrowly concerned with the technical performance of models and their profitability, and have sought to sidestep debates around the sentience of AI systems. Birch recently co-authored a paper with academics from Stanford University, New York University, Oxford university, as well as AI specialists from the AI companies Elios and Anthropic, about the possibility of AI sentience. The paper argues that the possibility of AI sentience shouldn’t be seen as a fanciful sci-fi scenario, but a real, pressing concern. The authors recommend that AI companies should attempt to determine the sentience of the AI systems they develop by measuring their capacity for pleasure and suffering, and understanding if the AI agents can be benefitted or harmed. The sentience of AI systems could be assessed using a similar set of guidelines to those used to guide welfare policy in animals.