1、arXiv:2402.01662v4 cs.CY 12 Dec 2024Generative Ghosts:Anticipating Benefits and Risks of AI AferlivesMEREDITH RINGEL MORRIS,Google DeepMind,USAJED R.BRUBAKER,University of Colorado Boulder,USAAsAI systemsquickly improve inbothbreadthand depthof performance,theylend themselvesto creating increasingly

2、 powerful andrealistic agents,including the possibility of agents modeled on specific people.We anticipate that within our lifetimesit may becomecommon practice for people to create custom AI agents to interact with loved ones and/or the broader world after death;indeed,thepastyearhasseenaboominstar

3、tupspurportingtooffersuchservices.Wecallthesegenerativeghosts,sincesuchagentswillbecapableof generating novel content rather than merely parroting content produced by their creator while living.In this paper,we reflect onthe history of technologies for AI afterlives,including current early attempts

4、by individual enthusiasts and startup companies tocreate generative ghosts.We then introduce a novel design space detailing potential implementations of generative ghosts,and usethis analytic framework to ground discussion of the practical and ethical implications of various approaches to designing

5、generativeghosts,including potential positive and negative impacts on individuals and society.Based on these considerations,we lay out aresearch agenda for the AI and HCI research communities to better understand the risk/benefit landscape of this novel technologyso as to ultimately empower people w

6、ho wish to create and interact with AI afterlives to do so in a beneficial manner.CCS Concepts:Computing methodologies Artificial intelligence;Human-centered computing Human computerinteraction(HCI).Additional Key Words and Phrases:AI,AI agents,Generative AI,HCI,digital afterlife,digital legacy,post