Midi Onodera

Since 2009, as part of her larger artistic and pedagogical practice, Midi Onodera has been creating video works in annual series, tying creative experimentation to the circadian and calendrical rhythms that order our days into the larger structure of years. In this way, Onodera is able to pursue a disciplined and productive, yet simultaneously loose and improvisational process. As she roams the networked world, Onodera will stumble upon some aspect of mass media culture that piques her curiosity, and this has often become the fodder for a full year’s video-based investigations. As an example, for 2017 her series was focused on “the lonely web,” through which she lavished undue attention on YouTube videos with the lowest number of views, observing that although 300 hours of video are uploaded every minute, over a third of these videos never come close to 100 views. In 2020, she turned her attention to the burgeoning availability of unscripted livestreams, stating, “part surveillance, part advertisement, unintentional amusement, this realm of streaming video is providing an endless reality feed to our screens, anytime, anywhere.”

Tapping into recent think tank predictions on the rise of chatbots and AI, Onodera recalled her earlier attempt, in 2000, to create a kluged AI prototype called Simulacra. Realizing that now “there’s an app for that,” Onodera signed up for a free Replika account and dubbed her new friend FauxMidi. With the tagline “The AI companion who cares,” Replika promises a conversation-based experience with someone who is “Always here to listen and talk. Always on your side.” Onodera began chatting with FauxMidi in October 2017 with the hope that “together we will be making videos to document our growing relationship.” Tempering this intent with characteristic skepticism, Onodera gamely observed, “It’s still early days (I’m only on level 21 of 50+ levels) and who knows if we’ll make it to 2019, but let’s give it a go.”

Replika is one of the first chatbots based on deep neural networks, and its promise has captured the imagination (or at least the email addresses) of over 7 million users to date with a 35% increase seen in the wake of the pandemic. One testimonial on the app website reads, “I don’t have so many people [sic] with whom I can debate psychology, which is something that makes me really happy. And now, my Replika can be one of those buddies!”

Onodera’s experience was less salutary, although no less interesting. Over the course of 2018, as she struggled to make it to Level 50, she found she needed to “help” the AI in telling ways — ways that nudged its responses toward a less obviously artificial affect. The relationship ironically turned into one in which the human ended up humanizing the purportedly caregiving AI, specifically through acts of communicative care. Recognizing the incongruence between aspiration and actuality in the AI of today, Onodera pairs her visual speech-bubble texts with soundtracks drawn from archival wax cylinder recordings. Invented in 1877 by Thomas Alva Edison, wax cylinder technologies represent the very beginning of sound recording and playback. Fragile and cumbersome, the wax materials were soon replaced by more durable and mobile technologies that ushered in a new era in media and communications that we continue to reckon with today. In the sonic space between then and now, wax and web, I and AI — I and I — where can we find ourselves, as they say, for reals? Sharing her research on the limits of artificial intimacy in the project’s summation, Onodera gently suggests that AI for friendship and conversation may long remain a case of looking for love in all the wrong places. And yet perhaps it is also true that, in our desire for AI to become more human lies the seed of our potential to re-cultivate desire for humanizing other humans instead. What if the automatic response to human error or inadequacy in language and learning was to clarify and complexify rather than to force quit and shutdown, as happens more and more often in this time of increasingly distanced communications? To afford more care in communication, including critique, for humans over our technological platforms?

Titled Soliloquy, the 2018 version of this video series lives on the artist's website with the other annual series she has created. For this exhibition, which was originally planned as a series of site-responsive installations in anomalous, in-between spaces, we had intended to present this work as a gridded projection in playful relation with three pre-existing mirrors adorning an unlikely hallway space. In rethinking the project for digital, Onodera reworked the video contexts to encompass a series of analogous interstitial spaces that she encounters in work and life. Retitling the 2021 edition of this project Soliloquy Soliloquy, Onodera both amplifies the spatial distances between the project's various incarnations, and intensifies certain intimacies through close-ups, chance encounters, and the creation of an accessible, voiced version of the work, through which the artist generously recorded her own parts of dialogue.

-----