In the last session of our reading group, we all met up again to discuss and present the readings we examined. The reading I received for this week was by Hassine, Tsila, and Ziv Neeman, “The zombification of Art History: How Al resurrects dead masters and perpetuates historical biases.” Journal of Science and Technology of the Arts 11, no. 2 (2019): 28-35.
This text mainly indicates problems in AI in connection to art history. We are aware that AI has been around for a while, but always with the thought of humans being replaced and losing their jobs due to AI.
This has been discussed in many sectors across industries and also in the field of arts. But is AI really helpful when we think of how it is right now in art (f.e. image recognition, etc.)?
In order for AI to be actually useful it needs to be renewed and advanced. The Dutch Next Rembrandt project for instance is an example of how recreations of certain historic styles from the past in art with the help of AI are being processed. This project created a new portrait using gathered data that was analyzed from other pre-existing paintings of Rembrandt.
It is certain, AI has the ability and potential to change the art world, and its environment and can be very innovative.
It can even change art just like photography changed and influenced paintings during the time of its invention.
But the question arises if this type of AI is creating something like forgery? Is this something we can call ‘’Zombie art’’? Since this AI tries to recreate artwork and a style from a dead person, in this case, Rembrandt’s work.
This is in a way trying to bring the dead alive. It is questionable if this is intriguing or recreating a fake. Additionally, it is questionable if in the end, it is actually worth something? One could say this project only does the recreation of fakes with the help of machines. This is called zombie art and can also be created by humans, but at least humans in fact bring in creative elements while creating.
The writers of the text almost describe ‘’The Dutch Next Rembrandt AI art project’’ as it is kind of creating a fake.
During the creation of the project, data was collected to recreate the painting. They looked at age, gender, and face the direction in order to gather this data, but would all this effort really trick an expert at the end of the day?
The writers describe those projects overall as disappointing when it comes to the artistic expression, and question the value of the outcome. The project might be more interesting when it comes to technology instead of art and also brings the value of the original creations down in a way.
Additionally, what the project tries to do is recreate a certain style by analyzing the different paintings of the artist, but Rembrandt actually tried never to repeat his work and style, hence, this is why it would be hard to pick up on similarities. He created a variety of different artworks, such as landscapes, animals, portraits and Dutch life, etc.
For recreating the new Rembrandt with AI, certain features were just taken out of his portraits, but the downside of an AI project like this is that there is no surprise factor involved and therefore remains average as a piece of art. Moreover, the recreations do not seem balanced in a way hence why they do not appear appealing to the eye.
Most AI projects overall seem limited when it comes to creativity and people always almost know what is coming when using AI tools. To gain and create the data, they needed to study the works and then decide on a subject.
Things like, how often you see women or men or both in portraits together were analyzed, but AI can simply not pay enough attention to the artist’s view on life.
Earlier, creating art with computers did not have a certain goal, but now AI almost has to learn in order to recreate the masterpieces of art.
Technology certainly has influenced art already just like photography and the invention of oil colors did, but using AI to recreate the art of certain masters does not always make sense. In the future, the recreation of masters could possibly become more important.
Perhaps the actual goal of reproducing such masterpieces is not to recreate them out of context, but to learn about the creation of the painting with AI and its meaning?
The overall conclusion was that out of Rembrandt’s work, the algorithms could work best with portraits out of all of the works.
The first question that came up in the reading group regarding this text, was the fact that we questioned if it is okay when people recreate and reference someone’s work and if it is different when machines do the same thing?
Nevertheless, we all agreed that AI could never achieve the work that was the goal. AI can do a lot in our opinion, but this is something that it could never do (to be an artist). The reason for that is that artists don’t think like that, as they are not machines and do not get their inspiration from algorithms. AI can surely do great things, but not be creative as can the human mind. Humans with AI in combination can be effective, but not forcing AI to be creative. We also agreed that it also should be illegal to allow this to happen, even though AI recreating the work of a dead person becomes legal after a certain time has passed. In addition, it should be at all times mentioned as a reference.
Another interesting book we had was about Crypto and what it can do for political perspectives. Our comments on crypto were that we all are aware that it never shows any good news and that it also makes it possible to do bad actions and buy illegal items. The investors are also mostly men and this allows them to perform more sketchy activities. It also separates communities, and its access is just for the privileged, which is worrying. It excludes people since it is not accessible to everyone, and in order to understand it you have to properly invest time. To keep up with crypto you have to be up to date every day, it almost becomes a hobby, as circumstances can change every day. It is also not based on predictions, unlike stocks.
Another discussion about a text reading covered a group of artists who literally took action with art and approached the public to detect and speak about existing problems within society. They interviewed locals and published their truths, including hacking systems. The book overall described how they did it and described what needed to be avoided within their group. We thought though that this art actually made a huge impact compared to art we see from time to time.
On top of that, we talked about software in another reading and how it is difficult for people to understand. Nevertheless, it is such an important topic, since it is hidden everywhere. Most of us have no idea how the code behind it makes things really work. It is interesting how software is visible but also hidden at the same time. An example of this are phones and how they take actions of their own most of the time. In our opinion, there should be an opportunity to learn more about software, since it is everywhere. Yet, we also realized, that data is only gathered in certain societies, so this creates a huge imbalance.
Looking back at the last few weeks, I can say that working with people from another similar, yet totally different course has already broadened my horizon. It made me start looking at topics from another angle, I learned more about other people’s countries and what exists within them. Examples of this are political views, as well as communities and waves. In the future, I hope to be able to still interact with people from my course and other courses in this way, as it is a perfect way to share knowledge and be able to share your opinion while seeing how other people think about it too, which is not always possible in a huge group. I learned so much from each person and maybe would have never come across such interesting articles that make people think and reflect on the world we live in and how this might impact the future.