“The Latent Future” is an installation that visualizes the latent future by generating fictional news that could occur in the near future, using an AI that has learned from a large amount of past news. We presented this work at the “Open Space 2017: Re-envisioning the Future” exhibition held at the NTT InterCommunication Center [ICC] in Tokyo.
In recent years, events have occurred in the world where reality and fiction are mixed through incorrect facts and fake news. This work attempts to shed light on the uncertain “present” by generating fictional news that could occur in the near “future” based on a probabilistic model of language learned from past news.
Countless news articles mapped based on their semantic distance are displayed in a 3D space on the screen. The real news that actually happened in the past and new news that flows in from Twitter are mapped, and new articles/new news are generated from the spaces between their semantic proximity or other news.
Through such news that “could have happened/could happen”, this work crosses the boundaries between reality and fiction, past and future.
Exhibition
2017/05/27 – 2018/03/11 | Open Space 2017: Re-envisioning the Future | NTT InterCommunication Center [ICC]
Credits
Concept / Programming: Nao Tokui (Qosmo, Inc.) Visualization: Shoya Dozono (Qosmo, Inc.) Sound: Taeji Sawai (Qosmo, Inc.) Assistant: Yuma Kajihara (Qosmo, Inc.), Robin Jungers (Qosmo, Inc.)