The new possibility of creating an image via a text prompt, raises new questions about the boundaries of art and its future. A post about Dall-e 2 and AI based art
A few months ago, at a regional fair in the state of Colorado, Jason Allen won first prize for a painting he submitted to a competition at the fair. The title of the work was “The Space Opera Theater” and it was created by a new tool, DALL-E 2, (pronounced like the name of the artist Dali and not by chance), which enables the translation of verbal instructions into an image using artificial intelligence. This is exactly how the image at the top of this post was created.
Artificial intelligence is a product of the accelerated technological development of our time, which allows a computer to imitate human actions while accessing an infinite pool of information on the network. The computer can process huge amounts of information in a matter of seconds and probably in many cases, much more successfully than humans. Such as, for example, the software Alpha Zero, which is based on artificial intelligence, that completely disrupted the world of chess, and to this day no one has been able to beat it.
Computer generated art has existed since the early computer days but obviously, the aesthetic quality of such images has changed and improved and advanced greatly over the years. Theoretically, all the information in the internet is at our disposal, but humans are limited in their accessability, compared to sophisticated computer systems, which are able to process, analyze and extract specific information based on huge data in a matter of seconds.
The next Rembrandt
Back in 2016 I wrote about a groundbreaking project in the field of big data – the next Rembrandt. A so-called “new” painting by Rembrandt was revealed in the Netherlands, but it was actually a computer-generated image using artificial intelligence analysis. 160,000 fragments were extracted from the 346 paintings that 17th century Rembrandt created, to determine what is a typical Rembrandt painting. All this data was fed into a 3D printer that, with the help of 13 layers of ink, created a new portrait that even has a texture typical of Rembrandt. It was not a Rembrandt forgery but rather a painting that definitely could have been created by Rembrandt himself.
140 million pixels of artificial intelligence managed to deceive even Rembrandt experts, who were convinced that this was a newly discovered painting….
DALL-E 2
Six years have passed since the last Rembrandt – an eternity in today’s technological age and now, anyone can create complex, abstract or photorealistic works simply by typing a few words in a text box and pressing enter. I must admit that it is fascinating and exciting. Here above is an image I created using the DALL-E 2 software, with the following prompt: a Picasso-style painting of a woman frustrated by technology. Within seconds I received four samples and this was the most successful, although not quite in the style of Picasso. I’m sure that if I told you that this is a painting by a real life painter, you would certainly believe it, and so would I.
Many concerns
After it became known that the digital creation that won first prize at the state fair was an AI generated image, the net went berserk. Art is dead! Artificial intelligence is about to take over us!
If this sounds familiar, it’s because such reactions are heard whenever a new technological component is added to our cultural repertoire. Just a year ago it was the NFT, which also caused a big stir. Now it’s DALL-E2 and other AI apps.
This exactly was the scenario when photography was invented. The first camera designed for commercial distribution was invented in 1839. But for decades, photography was not considered art at all. Even in the second half of the twentieth century there are still remnants of the debate whether photography is art. Not every photographer (we are all photographers) is an artist but a camera, like any tool, can create art depending on how it is used. Today, in the digital age, we have amazing tools at our disposal, but most of the art they produce is far from amazing. Yet.
Pioneering AI generated art
The endless possibilities of big data have long attracted Rafik Anadol, a Turkish artist living in Los Angeles. Anadol is a pioneer in the field, using data as other artists use paint and creating mesmerizing immersive video works based on vast datasets of nature, history, human activity and art history. In the work he exhibited in New York, for example, he collected from the Internet all the images of the city – Central Park, the Brooklyn Bridge, skyscrapers. This multi-vision depiction of New York is, according to Anadol, the broadest view of the city.
In the technological age we live in, it is only natural that technology be an integral part of art. This, of course, does not negate the art of the past but adds a new layer of possibilities, but it takes time to assimilate a new artistic category. Slowly (or maybe quite quickly), the serious artists will stand out in a large mass of amateur work. At this point, art produced by artificial intelligence is in such an embryonic stage that it is not really yet possible to understand it.
What next?
We already realize that the technology that enables artificial intelligence has far-reaching consequences for humanity, primarily the way we experience ourselves as human beings. It raises the fundamental question “What does it mean to be human”? – What is the human advantage over the machine?
Artificial intelligence will definitely disrupt all areas of our lives and change them radically. The absurdity is that humans themselves are the generators of technology and as we strive to improve it, we may lose control of the machines we ourselves have created.
What are the implications for the future of art? Some people believe that human creativity will always be necessary to create truly original works of art. After all, machines can only do what they are programmed to do. They can’t come up with new ideas on their own (yet). Others believe that machine learning will eventually surpass human creativity. They claim that machines have the ability to learn and evolve much faster than humans. So it’s only a matter of time before they start creating their own masterpieces.
The main problem for the art world, as I foresee, will be to distinguish between an authentic creation and a computer generated image. The consequences of this issue will be many, first of all the question of the creator- who should sign the work and who does it belong to? Another implication is regarding the trading of these works and digital works in general – if a person can go to a website and create something custom, why would they buy art from a gallery or at an auction? Again, this doesn’t mean that there won’t still be collectors and people who acquire art as a status symbol (or as an investment) but what will happen in the future? In the meantime, I continue to have fun creating surreal images, such as the flying saucer in the style of Renaissance art, which is the product of a prompt I gave on another site. Although I didn’t actually create her, I feel a warm bond between us.
Would you care to share this post? Sharing is caring. Thanks
[addthis tool=”at-below-post”]
