从价值认同的信任内涵及维度看,生成式人工智能 Sora 的出现,使价值认同在主体乐信、文本可信、时空能信三个维度上经历着从复制到定制、从抽象到具象、从在场到离场的叙事逻辑转换.但文生视频Sora也面临着算法推荐迟滞主体能动性发挥、言语偏见阉割图像传播效度、网络圈层阻隔社会公信力生成等问题,造成生成式人工智能时代价值认同主体信任关系失序、文本信任关系失真、时空信任关系失衡等价值认同风险.为化解风险,通过价值注入与技术赋能、伦理审视与法律审查、破壁入圈与连圈接层等方法,扩大信任视角下生成式人工智能时代价值认同的范围并提升其效度.
The Logic and Approach of Value Identification in the Era of Generative Artificial Intelligence from the Perspective of Trust:Based on Thoughts on Sora
From the perspective of the trust connotation and dimensions of value identification,generative artificial intelligence Sora has witnessed the narrative logic transformation of value identification in three dimensions:from replica-tion to customization in the dimension of the subject's willingness to trust,from abstraction to concreteness in the dimen-sion of text credibility,and from presence to absence in the dimension of spatio-temporal trustability.However,text-to-video Sora also faces dilemmas such as delaying the exertion of the subject's initiative by algorithmic recommendation,emasculating the validity of image dissemination by speech bias,and blocking the generation of social credibility by net-work circles,which result in risks of value identification,such as disorder in the trust relationship of the subject,distor-tion in the trust relationship of the text,and imbalance in the trust relationship of space and time in the era of generative artificial intelligence.To mitigate these risks,approaches such as value infusion and technological empowerment,ethical review and legal examination,and breaking barriers and integrating circles and layers can be adopted to expand the scope and enhance the validity of value identification in the era of generative artificial intelligence from the perspective of trust.
NETL
NSTL
万方数据
