Happens To Be Performing
鄭先喻 X 陳武康 CHENG Hsien-Yu X CHEN Wu-Kang
2022
軟體視覺化、機器學習、印表機、螢幕、現場演出
Visualized software, machine learning, printer, screen, on-site performance
本次作品『Happens to be performing』將由2019至今的軟體所蒐集到的網路留言資料庫中,建置文本語義模型,規劃每一週產出一次劇本生成,並視覺化文本再以現場的事務印表機裝置去藉由表演者,以隨機觸發劇本生成的方式,在現場取得列印出的文本,做出臨時並且即興的演出嘗試。
*作品源自2019年個展同化者中的無名作品,該作品前身也與張碩尹與廖銘和的臺北機電人2.0一展中,作為控制現場裝置與聲音以及顯示網路民眾即時留言與對話訊息展出,在發展時,原計劃為將網路民眾的留言與即時聊天訊息資料藉由機器學習,去制定出語意的模型與說話的方式,再由機器學習產出劇本。
Happens to Be Performing originated from an untitled work in Cheng Hsien-Yu’s 2019 solo exhibition “Assimilator.” The earlier version of the work was also presented during Chang Ting-Tong and Liao Ming-Ho’s 2020 exhibition “Taipei Robot Man 2.0.” The work uses internet installations to extract messages and conversations of people on the internet, which are retracted and transcoded into a driving force for the on-site installation and sound.
This presentation takes a step further and incorporates machine learning to construct a semantic text model through a database of online messages that are collected from 2019 to the present moment, to produce instant play scripts. This work was created in collaboration with choreographer/interdisciplinary artist Chen Wu-Kang, who activates the printers on site and plays chess with printed AI visual texts, leading to random and impromptu performances.