For a few weeks now, I’ve been using ChatGPT to debug my old Google Hacks, which were regularly crashing over the past few years. From around 2001 to 2007, I focused on this artistic trend that I had helped to invent and explore at the time. Most of these hacks consisted of a few lines of code (“leeches” as they are called) that diverted Google from its utilitarian dimension. It all started around the events of September 11, 2001.
The very first one, from August 2001, was Epiphanies (ChatGPT just fixed it so you can play with it). It was both a Google Hack and a misappropriation of comments on James Joyce by Jacques Lacan. This piece then gave birth to Gogolchat, in collaboration with Jimpunk (2002), and to Human Browser, a performance that I produced from 2004 to 2017. Today, it resembles a crazy ancestor of LLMs and ChatGPT (so why not have it fixed by ChatGPT itself), very rude and politically incorrect.
In December 2001, if I remember correctly, I made Fascinum (see picture above). The influence of the events of September 11 is obvious: the piece shows in real time the most viewed news images on the various national news feeds. The viewer rides at the top of the infotainment wave and experiences the paradoxes of one-track thinking in a panoptic vision. Fascinum won the @ARCO 2007 New Media Prize at the ARCO Contemporary Art Fair in Madrid and was subsequently sold as an installation to contemporary art collectors (I sold 4 out of 5 items). The fact that I had to repair the piece every time Google changed the code of its web pages is quite interesting: this art piece is a critique of humanity’s imprisonment within its own media landscape, and the very fact that the piece has survived is a sign of the failure of its critique ; conversely, if criticism had won, the spectacular media landscape would have collapsed and the work of art would have died (like a parasite, a saprophyte more precisely, would die if it killed the ecosystem on which it feeds). This paradox is quite interesting in the context of media archeology (I have written a lot on this topic) and its relationship to the narcissism of the artist, torn between aesthetics and politics.
The last piece that ChatGPT fixed is Non-Weddings (January 2002), the first Google Hack for image search, and, once again, inspired by Lacan (at least by one of his drawings). I was hesitant to revive this piece because it seems so ridiculously flat compared to the visual power of newer digital art installations or AI image generation (but as I often says, my practice is about “conceptual art without ideas”). However, something still fascinates me in this piece, especially since I reactivated my Dreamlogs research project, which tries to take a step further in the direction of a “topography of the unconscious of capitalism”.
Since 2001/2002, all these pieces have broken down several times and I have had to repair them regularly. I can tell you that it’s quite desperate, especially when you participate in the invention of a new form of art and you would like a still non-existent public to accept it as art, at least (although sometimes it’s nice not to be considered an artist, as I will learn much later). Today, the situation is such that digital information structures have become so complex and secure to ensure the stability of what I called Semantic Capitalism, that it is almost impossible for a simple human being and a bad programmer like me to repair these works of art. It’s like the administrative structures of developed Western countries: bare life has become more and more difficult there.
Christophe Bruno 