电影艺术2024,Issue(5) :28-34.

人工智能影像生成的模型诗学

Model Poetics of Image Generation in Artificial Intelligence

朱恬骅
电影艺术2024,Issue(5) :28-34.

人工智能影像生成的模型诗学

Model Poetics of Image Generation in Artificial Intelligence

朱恬骅1
扫码查看

作者信息

  • 1. 上海社会科学院文学研究所
  • 折叠

摘要

通过人工智能(AI)开展的影像生成,在艺术创作与研究中面临表述危机,"模型诗学"有助于应对这一状况.模型诗学建立在对AI的历史溯源上,它以施行性为导向,从技术操作的视角审视模型的"类型"与"实例".它结合了针对代码和文档的细读与操作者经验的采集分析两种操作方法,前者为AI模型提供"部件解释",后者则以"接合表述"的方式呈现出AI模型承载的思想观念.模型诗学揭示出AI模型自身可以是艺术表达的载体.

Abstract

The use of artificial intelligence(AI)in image generation has posed a challenge to artistic creation and research,leading to a crisis in representation."Model poetics"offers a solution to this situation.Drawing on the historical origins of Al in cybernetics,model poetics takes a performative approach to examining the"classes"and"instances"of Al models in the context of technical operations.It combines close reading of code and documentation with the analysis of operator experience,providing"parts explanation"for Al models and presenting the conceptual ideas they embody through"articulation",both in Kaufmann's terms.Model poetics reveals that Al models per se can serve as vehicles for artistic expression.

关键词

模型诗学/计算性诗学/控制论/黑箱/细读

Key words

model poetics/computational poetics/cybernetics/black box/close reading

引用本文复制引用

基金项目

国家社科基金艺术学青年项目(21CA169)

出版年

2024
电影艺术
中国电影家协会

电影艺术

CSTPCDCSSCICHSSCD北大核心
影响因子:0.411
ISSN:0257-0181
参考文献量2
段落导航相关论文