Art and simulacrum at Nitroglobus in Second Life

Drawn from the Latin, simulacrum (“likeness, semblance”) entered the language of European art during the late 1500s, when was originally used to refer to a painting or statue directly representing someone or something (most notably a god or deity). However, by the 19th century, it had come to express an artistic endeavour – notably an image – formed without the substance or qualities of the original (e.g. to use a more modern example: a photocopy of a photograph of a painting).

It is a term which is particularly relevant today thanks to the rapid rise of generative AI tools, and the manner in which they can be used to literally churn out prompt-based images over and over without any genuine artistic input or understanding of the actual desired request or the inconsistencies, anachronisms and outright errors present within the final product.

Of course, those using such tools diligently will argue that they are merely a tool; when balanced with a critical eye and the use of other tools to refine, enhance and correct, they can be used to produce richly diverse pieces that do have originality within them. This is actually not an unfair summation, as far as it goes.

The problem¹ here is that the vast majority of users of such generative tools don’t exercise skill or talent. They prompt, wait, publish, often using minor variations of the same prompt ad nauseam (“make the woman’s hair colour blue”) to produce a stream of near identical images (as all too often seen on the likes of Deviant Art). There is little originality within such pieces when compared one to the next; worse, there is little in the way of critical review, and those aforementioned errors are allowed to persist. Worse still, all this repetition (complete with errors) is fed back into the data pool, further diluting it and increasing the regurgitation of vacuous elements to form what amounts to self-perpetuating simulacrums, devoid of originality and talent.

This constant cycle of regurgitation and repetition without originality is explored by Manoji Yachvili (onceagain), in her exhibition Simulacrum, currently on display at Nitroglobus Roof Gallery, curated by Dido Haas. It is also very much a personal expression of Manoji’s own increasing sense of artistic loss and frustration as she sees creative expression, skill and progress within art and artistic experimentation vanishing in a rising tide of banality.

Reality no longer exists, it has disappeared, crumbled by the media and modern technologies that propose images that do not refer to reality, that receive meaning only from other images and that are perpetually regenerated, thus remaining increasingly disconnected from what was originally real. Everything I see has an appearance and the power of appearance, without a faithful external image and therefore devoid of the original vitality.

– Manoji Yachvili

To achieve this, Manoji presents a collection of original paintings that are richly expressive whilst clearly echoing the banality and emptiness of the kind all too easily spewed forth by generative AI. Through the use of simple, repeated elements – masks, faceless figures, and muted colours – she beautifully conveys the empty, expressionless redundancy so common within generative AI art.

Yes, these are all paintings utilising common themes elements – but each of them is completely original in form and presentation, allow it to stand on its own as a singular, unique piece carrying with it an ability to speak to each of us differently. Thus, these are pieces that reach beyond the artist, offering a richness of expression and meaning, standing not as the result of a collection of prompts, but as a prompt for our imaginations to take flight within them.

In doing so, Manoji establishes the genuine power of art through the hand, eye and talent originating from within the artist, making this a beautiful expressive collection, each piece standing in its own right as a genuine work of art.

SLurl Details

• Nitroglobus Roof Gallery (Sunshine Homestead, rated Moderate)

1. Please note that within this article, I’m intentionally avoiding issues of copyright and ownership in relations to the “training” of generative AI tools, because while these are clearly of concern generally, they are not the focus of the exhibition being reviewed.