Computer Animation/Film/VFX

Anerkennung - Honorary Mentions

YouTube Smash Up

Parag K. Mital (US)




URL:
http://pkmital.com/home/projects/youtube-smash-up/

YouTube Smash Up attempts to generatively produce viral content using video material from the Top Ten most viewed videos on YouTube. Each week, the Number One video of the week is resynthesized using a computational algorithm matching its sonic and visual content to material only from the remaining Top Ten videos. This other material is then re-assembled to look and sound like the Number One video. The process does not copy the file, but synthesizes it as a collage of fragments segmented from entirely different material.

Using YouTube’s interface, the videos are also textually tagged with popular culture’s “most viewed” artifacts, i.e. the database containing the Top Ten YouTube videos. This process attempts to inject the video into the community, masquerading as an innocent tribute video. The video’s audience, often viewers hoping to find the original Number One video, are almost certainly disturbed by the videos, as illustrated by the video’s overwhelmingly negative “like” ratio, and by comments such as, “now im [sic] blind”, “Will someone kill me in my sleep because I watched this video?” and another commenter’s reply to the previous comment, “me 2 [sic]”.

Despite their poor reception, likely due to their cut-up and abstract nature, most smashups have been the subject of copyright violations from YouTube’s automated copyright infringement detection system, Content ID. In each case, Content ID flags the videos as duplicates of the Number One video, rather than flagging any of the content actually used from the Number Two to Ten videos. This automated system attempts to automatically discover copyrighted content in newly uploaded videos, informing the original content holders if it finds anything. Most likely the content-rights holders never watch the supposedly infringing videos, and instead forward a cease-and-desist notice threatening a lawsuit. Despite the powerful language used by the content-rights holders, the videos were all put back online after multiple rounds of fair-use arguments and even more cease-and-desist notices.

The videos manipulate a level of representation indistinguishable by a robot perception, a space between pixels and perception, juxtaposing cultural fragments at a proto-object layer in an entirely automated process: Miley Cyrus’s lips collaged against the background of a troupe of dancing animals or Psy’s forehead dancing without the remaining pieces of Psy. Within this space, a disjunct between a state-of-the-art robot perception and those of unsuspecting YouTubers is revealed, asking what constitutes a copyrightable cultural artifact, as algorithms become increasingly more intelligent and as data continues to be manipulated by even more complex pattern-recognition and information-retrieval algorithms. Finally, the videos attempt to probe a dystopian future of automated content generation, when computer algorithms are not only capable of modeling cultural artifacts but also producing them, further embracing their present role as mere content curators.

Biography:

Parag K. Mital

Parag K. Mital (US) is an artist and interdisciplinary researcher obsessed with the nature of information, representation and attention. Using film, eye-tracking, EEG, and fMRI recordings, he has worked on computational models of audiovisual perception from the perspective of both robots and humans, often revealing the disjunct between the two, through generative film experiences, augmented reality hallucinations and expressive control of large audiovisual corpora. Through this process, he balances his scientific and arts practice, with both reflecting on each other: the science driving the theories, and the artwork re-defining the questions asked within the research.

Credits:
Programming, concept, execution: Parag K. Mital
Carried out during a PhD on “Audiovisual Scene Synthesis”, funded by the Department of Computing, Goldsmiths, University of London, under the supervision of Mick Grierson and Tim Smith.