Digital Art and AI

Digital Art and AI

When one thinks of AI (Artificial Intelligence), their first thought tends to escalate into a science fiction futuristic nightmare where rogue, metallic robots have taken over humanity. However, when one approaches AI from a scientific perspective, the process of developing an algorithm is far less dramatic and far lengthier.

AI is an intelligence demonstrated by machines as opposed to the emotional, natural intelligence shown by humans.  

Advances have been made in Artificial intelligence in numerous fields, the inclusion of deceased actors posthumously in film and hologram technology to resurrect musicians at music festivals are just a couple of prominent examples. Artificial intelligence has more recently been used to restore and replicate old or lost masterpieces from a bygone era. By using this digital platform, scientists have been able to restore and recreate masterpieces that would have remained lost forever. However, is it right to manipulate history in such a way?  

Most recent examples of AI art restoration were the resurrections of three lost Gustav Klimt paintings, these paintings were destroyed during Nazi looting in World War Two. The three faculty paintings entitled Medicine, Philosophy and Jurisprudence existed only as black and white photographs until Google Arts & Culture collaborated with Belvedere Museum to develop a tool to colourize and display the paintings. After 6 months of developing an algorithm, the scientists were able to teach the computer to digitally revive the artwork. Klimt had been dead for almost two decades when the paintings were stolen and therefore would never have known his artwork was taken or that decades later the pieces could once again exist. Morally the idea of recreating lost art is challenging as by restoring them, one could argue that history is somehow being undone. History is sometimes overlooked but a plaque detailing the paintings history would rectify this.  

Another notable example of restoration was when Rembrandt Van Rijn’s famous 1642 Night Watch canvas was completed. The canvas had been trimmed down in 1715 to fit between two doors at Amsterdam’s city hall. Using high tech scanners, Xray’s and digital photography – the missing pieces of the artwork were restored. This monumental achievement allowed the missing pieces to be restored after nearly three decades. Therefore, this allowed generations of art enthusiasts to witness the missing pieces of the completed artwork but also omitted the story of it being trimmed. It is miraculous and exciting to be able to see the completed artwork but also feels uneasy that is has undone the timeline of the artwork’s history. While the original artwork still exists, there is now a completed version that contradicts it. This multi-textual blending of worlds would be better suited in a Science Fiction novel than real life as it appears to propose a multi-verse. 

Digital art has been a predominant means of creating artwork over the last few years therefore Artists who are established now will be most likely using a digital platform to display and compose their pieces. Digital artwork appears seamless and while it may take just as long to create as a painted piece, it is easily replicable and distributed upon completion. Being that it is digital, the artwork doesn’t need to be handled with protective gloves or transported with great care as painted pieces would which reduce the travel costs of the artwork (if it is to be transported in a physical form).  While a digital piece can be just as awe-inspiring as an archaic masterpiece- the way in which it has been made differs significantly. Instead of paint, the artist uses a stylus and a screen; the artist’s tools are therefore far more compact and transportable. The stereotypical image of an artist’s apron splattered with numerous colours still exists but when looking at digital art this image has altered to one an artist creating digital art using lines and patterns.  

Using AI to recreate an aged masterpiece or lost artwork is nothing short of incredible. Technology is always evolving and by developing these algorithms; it means that damaged or missing artwork is retrievable. Does this not de-value the original piece however as it can be replicated any number of times. On one hand this allows multiple collectors to own a ‘one off’ masterpiece of art history while the original remains housed safely in a museum. But it also means that these priceless pieces may lose value as they can be mass produced based on market demand. When handling technology of this ability, the scientist needs to consider the ethical issues that may arise. To create an almost exact copy of a timeless piece of art to some artists might appear like copying or forgery as it crudely can resemble postcard prints and souvenirs that one might purchase at a museum. Thus far, this technology has only been used to restore rather than market but if it is possible to teach a computer to create a lost piece of art based on historical evidence alone- it wouldn’t be too difficult to teach it to replicate DaVinci’s Mona Lisa or Munch’s The Scream. Posters of both pieces are available to purchase from most museums, yet where would this differ from an AI produced version. An AI version would be grandiose and exhibited at a specially organised event and would then find a permanent residency at a gallery or museum.  These events often require a ticket to attend which differs from the often-free admission to a museum. 

If numerous museums were able to house the same piece of classical artwork, would it then mean that Museums like the Louvre and Mauritshuis lose their appeal?  This might also affect tourism and traffic as people wouldn’t need to travel to see rare artwork. It is unlikely that this would happen as assumedly the Museum would have to express consent for the artwork to be replicated. 

AI has often been at the centre of debate as some fear that it will render humanity redundant if a machine can do the work more efficiently and without payment.  Many of these fears however originate from a place of anxiety and fiction based on novels, movies and television. Yet these fears are also valid and require discussion. When an AI programme produces artwork, it is without emotion. An artist requires these feelings and stylistic individuality in order to create impactful pieces of artwork. If an AI can create these same pieces without any feelings whatsoever, does this not disregard the struggles of the original artist?  For instance, Vincent Van Gogh suffered debilitating mental illness including hallucinations and psychosis throughout his life up until his suicide; while institutionalised in an asylum he continued to paint some of his most iridescent and striking pieces. Through history and letters, we can better understand what Van Gogh experienced and find deeper meaning in the artwork- by digitally recreating these pieces are we not dismissing the devastation and unease from each brushstroke?  The Van Gogh alive exhibit was a digital experience that displayed his work; it was both immersive and emotive despite being entirely created by projection. While telling the story of the artist’s life, the artworks were enlarged to cover the walls in the dimly lit rooms. Praised by many as mesmerising there were still some who found the exhibit more a spectacle than an appreciation of the artwork. The debate is ongoing but for some who had studied the artist in detail, they found the exhibit showcased the material in an inventive way. 

It is imperative for the world to evolve; it is necessary and expected that technology plays a significant role in this. While some art is still painted on a canvas or sprayed onto a wall- digital does take precedent in the current market. It is a dizzying thought but an assuring one that artwork can be created using software and AI. Digital art however is not without risk as the software might corrupt or break; just as an easel might chip or fade.  

 NFT’s (Non Fungible Tokens) are another way in which an artist’s signature has changed, by using a digital platform to create a token which will represent the artist’s identity.  The metaverse enhances reality by proposing a virtual world of tokens and iconography enclosed behind a screen. When purchasing an NFT the individual has permission to copy and print the image onto merchandise. This is one of the ways in which the artist/collector relationship has shifted.  Art is still auctioned off in some cases but with NFT’s the purchase is made through crypto- currency. In recent years this has been beneficial given the imposition of travel restrictions and social distancing. Each NFT has the same basic principle but is layered to feature different clothing or accessories.  

AI in art has made strides and broken new ground which will establish the groundwork for future projects. AI is not however restricted in the realms of artistic development; it too can be used in other fields of interest. Medicine, Law and Media to name a few.  By teaching a machine to think consciously, we are further reducing a need to humanity. While this thought is disconcerting, it does inspire discussions about the evolution of technology. 

Overall, by using AI to restore artwork, it presents opportunities for missing pieces to be seen. As moral and ethical issues arise; it is without a doubt a spectacular triumph of man-made technology that will most likely change the course of how art is appreciated and shown.  

Article by Dana Archer.