インタラクティビティを再定義する
Redefining Interactivity

About 70 years ago, a mathematician Norbert Wiener and his colleagues proposed the field of cybernetics. The idea was to create a unified theory for animal and machine functions from the perspectives of communication, control, and statistical mechanics. It was a bold question. At the same time, it was also a proposal of how the disciplines could overcome fragmentation. At the time, this idea made a significant impact on science, technology, industry, and culture. As a result, many works have been created that anticipate a world where humans and machines coexist in harmony. The most prominent of these works are interactive works in media art, games, and advertising. An interactive work takes information about a spectator's behavior and the environment as input and interprets this input. Then, an algorithm described by the author maps this interpreted information to outputs such as video, audio, and changes in the environment. The spectator experiences the work and interacts with the artist. At first glance, this may seem obvious. However, we came to realize that we must once again ask the question: What is interactive work; moreover, what is interactivity? Because the world where humans and machines live in harmony is becoming a reality.

Twenty years ago, Jim Campbell, a contemporary artist known for his low-resolution LED matrix installations, poses this question in a short paper. In this paper, Campbell raised several issues and sharply criticized interactive works. Most notably, he pointed out that unlike paintings and films, interactive works are framed by the artist's assumptions about how the spectator will experience the work. If the work can only be viewed within these limitations, it cannot be considered an open dialogue. Twenty years later, compared to the works that Campbell critiqued at the time (which were called “multimedia”), today's interactive works have evolved significantly in terms of technology. However, isn't his point still valid today? If an author's imagination and descriptive ability are constraints that define the behavior of spectators, then the dialogues are closed, and the spectators can only behave and interpret the work within the frame set by the artist. However, in the first place, an interactive work emerges when a spectator experiences it. Shouldn't there be more room for interpretation by spectators? Moreover, shouldn't it be okay if spectators misuse or divert the work from its intended purpose? We believe that we need to start by questioning and redefining the seemingly obvious concept of interactivity. For this reason, we focused on machine learning, one of the technologies used to implement AI.

In classical programming, people use programming languages to describe algorithms. By contrast, in machine learning, a machine learns rules from data and describes an algorithm. This new programming paradigm not only allows for writing beyond the author's ability to describe but may even allow the algorithm itself to be rewritten as it learns from feedback provided by the spectators who interact with it. We want to propose holding this workshop as the first step in redefining “interactive,” and to consider creativity and media in the age of living with machines.

基調講演

• 参加対象者はIAMASの受験を検討している人だけですか？
  ☞いいえ。今回のworkshopのテーマ「インタラクティビティを再定義する」に深く興味を持っていただける方の中で、この日程に参加していただける方が全てが対象となります。

• 最後の発表をYouTube Liveで配信予定とのことですが、ビデオをオンにして参加することは求められますか？
  ☞いいえ。チームメンバーの誰かが画面共有した状態での発表を想定しておりますので、発表時におけるビデオの使用は任意です。

• プログラミングや機械学習の経験は必須ですか？
  ☞いいえ。それぞれの経験に応じて事前に準備していただけるよう、YouTubeによるオンデマンド形式のチュートリアルとハンズオンを提供する予定です。

• ノーバート・ウィーナー『サイバネティックス―動物と機械における制御と通信』池原止戈夫：訳、彌永昌吉：訳、室賀三郎：訳、戸田巌：訳、岩波書店（2011年）

• ノーバート・ウィーナー『人間機械論―人間の人間的な利用【第2版】』鎮目恭夫：訳、池原止戈夫：訳、みすず書房（2014年）

• トマス・リッド『サイバネティクス全史―人類は思考するマシンに何を夢見たのか』松浦俊輔：訳、作品社（2017年）

• Campbell, Jim. "Delusions of dialogue: control and choice in interactive art." Leonardo 33.2 (2000): 133-136.

• 三宅陽一郎『人工知能のための哲学塾』ビー・エヌ・エヌ新社（2016年）

• 三宅陽一郎『人工知能のための哲学塾―東洋哲学篇』ビー・エヌ・エヌ新社（2018年）

• 三宅陽一郎、 大山匠『人工知能のための哲学塾―未来社会篇』ビー・エヌ・エヌ新社（2020年）