EVENT | Seen by Machine: Computational Spectatorship and the BBC Archive

Film and television makers have been using computers in their practice for almost as long as computers have been around. Recently, they have incorporated machine learning techniques as creative tools in their craft. Creative machine learning holds the promise of the automation of the production and reproduction of visual culture, and this type of automated image-making presents to its audiences confounding pictures of authorship, authenticity and value. However, looking beyond the hype and the many misleading headlines about “creative machines”, there are powerful social and economic forces that have drawn artists and creators of all kinds to have an interest in machine learning.

Daniel Chavez Heras (King’s College London) collaborated with a small team of technologists at BBC R&D to create a system that generates sequences out of archive footage using machine learning. The results of these experiments were edited into the television programme Made by Machine: When AI met the Archive, which is now the first time machine learning has been used in this way to produce prime-time content for television. Through this example, Daniel will discuss the idea of audio-visual archives as “cultural big data”, and their automatic browsing as an instance of computational spectatorship: a way to understand how our visual regimes are increasingly mediated by machine-seers.

Bio: Daniel has been working with pictures and computers, in various capacities, for more than ten years. He trained as a designer in Mexico and has worked in creative roles in print media and television before joining the British Council as a digital manager, where he was responsible for the digital portfolio of the organisation’s operation in Mexico. In 2010 he was awarded a Jumex fellowship for the study of contemporary art, and he has since been awarded a Fulbright scholarship twice, although he ended up declining both times to come to study in the UK: first for an MA in Film Studies at King’s College London, and currently at the Department of Digital Humanities, also at King’s, where he is trying to teach computers to watch films for his PhD. Daniel is funded by Mexico’s Ministry of Education through its Science and Technology Research Council (CONACYT); he has published in English and Spanish in international peer-reviewed journals on films & computers, on videogames & art history, and has taught university courses, in both Mexico and the UK, on visual narrative, digital aesthetics, and most recently on the politics of online networks and social media. Daniel has also made a few personal short films (one of which was screened in the official selection at the UNAM International Film Festival in 2013).

This event is part of an ongoing seminar series on “critical inquiry with and about the digital” hosted by the Department of Digital Humanities, King’s College London. If you tweet about the event you can use the #kingsdhhashtag or mention @kingsdh. If you’d like to get notifications of future events you can sign up to this mailing list.