This paper hinges on the differences between human and algorithmic AI temporality by showing how generative AI systems profoundly influence our perception of time and cultural heritage. While human filmmaking involves continuous, shared experiences, transformer models use discrete operations to tokenize, compress, and reconstruct videos. Although Sora’s mixed transformer/diffusion multimodal model can generate complex scenes, such as a gymnast’s aerial flip, it often produces unrealistic physics and confuses causal relationships. Discretization influences historical consciousness, with AI systems like Sora reshaping narrative interpretation through mathematical frameworks. These “artificial history makers” (Hughes-Warrington, 2022) challenge orthodox ideas of historical temporality and foster contemporary conceptual mapping of the digital cultural construction of (fake) temporal experience, such as the deepfake reenactment of alternative histories such as  Nixon’s held contingency speech in the case of a moon disaster. Performed by an actor and made to resemble Richard Nixon through deepfake technology, the chapter shows that this technology actualizes historical possibilities rather than merely simulating them. AI avatars as synthetic AI media, however, not only preserve history but also revitalize it dynamically in alien machinic temporalities Ernst (2016). As generative AI media evolves, cultural knowledge transmission may increasingly depend on AI representations as shown in the Video of different temporal versions of Billy Joel. This might result in a more fluid understanding of cultural identity formed through algorithmic mediation and transform temporality experience and historical understanding as a continuous recreation process, necessitating a balance through concepts like memory fiduciaries to protect against distortions. Thus, we also hint at the importance of safeguarding against potential distortions in a future policy debate surrounding synthetic media, especially in the field of post-mortem agents, such as posthumous AI avatars examining societal consequences of the differences between human and algorithmic AI temporality.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Do “AI” Avatars Generate Our Past? Hacking into Transformations of Temporality by Synthetic Media

  • Alexander Matthias Gerner

摘要

This paper hinges on the differences between human and algorithmic AI temporality by showing how generative AI systems profoundly influence our perception of time and cultural heritage. While human filmmaking involves continuous, shared experiences, transformer models use discrete operations to tokenize, compress, and reconstruct videos. Although Sora’s mixed transformer/diffusion multimodal model can generate complex scenes, such as a gymnast’s aerial flip, it often produces unrealistic physics and confuses causal relationships. Discretization influences historical consciousness, with AI systems like Sora reshaping narrative interpretation through mathematical frameworks. These “artificial history makers” (Hughes-Warrington, 2022) challenge orthodox ideas of historical temporality and foster contemporary conceptual mapping of the digital cultural construction of (fake) temporal experience, such as the deepfake reenactment of alternative histories such as  Nixon’s held contingency speech in the case of a moon disaster. Performed by an actor and made to resemble Richard Nixon through deepfake technology, the chapter shows that this technology actualizes historical possibilities rather than merely simulating them. AI avatars as synthetic AI media, however, not only preserve history but also revitalize it dynamically in alien machinic temporalities Ernst (2016). As generative AI media evolves, cultural knowledge transmission may increasingly depend on AI representations as shown in the Video of different temporal versions of Billy Joel. This might result in a more fluid understanding of cultural identity formed through algorithmic mediation and transform temporality experience and historical understanding as a continuous recreation process, necessitating a balance through concepts like memory fiduciaries to protect against distortions. Thus, we also hint at the importance of safeguarding against potential distortions in a future policy debate surrounding synthetic media, especially in the field of post-mortem agents, such as posthumous AI avatars examining societal consequences of the differences between human and algorithmic AI temporality.