IRINA CHECKHOMOVA / Imagination.
RESEARCH
It is important to look into relating practices and literature while working on any creative project. This page presents a few of the most relevant artworks and articles that have contributed to the development of this project. You can find 8 artworks and articles below.
All of the works are incredibly important because they inspired and guided the way for this project.
Artificial intelligence
Machine learning and artificial intelligence were first based on the model of neurons connecting with each other that was presented in the book “The Organization of Behavior: A Neuropsychological Theory” by Donald O. Hebb (1949). A recent report from the Computing Conference in 2017 talks about the simulation of brain activity in artificial intelligence by calculating how and why humans process information and what outcome the brain gives. Then implementing it in a mathematical algorithm to see if the human brain can be replicated by a computer.
Humans have five sensory organs that bring information to the brain. “The process of obtaining an inference to become new knowledge is called as knowledge extraction” (Ahmad & Sumari, 2017). They form the theory of constructivism. This theory says that “humans generate knowledge and meaning from experiences and interactions.” (Ahmad & Sumari, 2017). This also includes the probabilistic type of thinking that bases our decisions on our beliefs. We also constantly thinking about the future and past and gain information from communication. This document presents a model that puts human information processing into the computational math method in a computer model. This model proves that it is possible to implement rational thinking carried out by humans by enhancing the algorithms that already exist.
​
- Ahmad, A. S. & Sumari, A. D. W., 2017. Cognitive Artificial Intelligence: Brain-Inspired Intelligent Computation in Artificial Intelligence. London, UK, Computing Conference.
Recognition
I came across a large number of interesting artworks within this area. The first installation that is worth talking about was presented to me by my supervisor; it is called “Recognition”. This program used computer recognition to process Tate’s collection and find similarities between the artworks and news images. This program won the IK prize 2016 for digital innovation. It was a physical installation in Tate Britain in 2016. Now when accessing this installation online it shows the most interesting and thought-provoking matches that the program created. It searched the data with different types of methods like an object, facial, composition, and context recognition. One of the examples the computer matched a picture that was attached to a man’s head during a public appearance at a religious procession with a classic portrait from 1545 A Man in a Black Cap. It looked for men in hats and with a beard and matched the two from that logic. Although this project does not necessarily talk about the same topic as my project it shows how complex and fascinating machine learning and artificial intelligence can be.
26/11/2016 Kitchen utensils are pictured after a fire broke out in a slum area in Jammu © MUKESH GUPTA / REUTERS 1991 Monument © IAN HAMILTON FINLAY / TATE
02/09/2016 United Nations Secretary-General Ban Ki-moon reacts at the "Sustaining Peace - Achieving Sustainable Development Goals" forum during an official visit in Colombo © DINUKA LIYANAWATTE / REUTERS 1963 Before the Event © HAROLD COHEN / TATE
13/09/2016 A boy immerses an idol of Hindu god Ganesh, the deity of prosperity, into the Sabarmati river during the 10-day-long Ganesh Chaturthi festival, in Ahmedabad © AMIT DAVE / REUTERS 1893-4 August Blue © HENRY SCOTT TUKE / TATE
26/11/2016 Kitchen utensils are pictured after a fire broke out in a slum area in Jammu © MUKESH GUPTA / REUTERS 1991 Monument © IAN HAMILTON FINLAY / TATE
screenshots from the original website http://recognition.tate.org.uk/#intro
Machine Learning
To investigate machine learning let us look into the book “Machine learning: Trends, perspectives, and prospects” by M. I. Jordan and T. M. Mitchell (2015). They go in-depth about the concept of machine learning and the types of algorithms within the topic. Most importantly they talk about how in the past twenty years machine learning has driven the development of an unthinkable amount of digital influence. Machine learning makes computers improve automatically through experience based on fundamental statistics and algorithms. “Within artificial intelligence (AI), machine learning has emerged as the method of choice for developing practical software.” (Jordan & Mitchell, 2015). There are different types of machine learning. It started as something as simple as a yes or no question then it gradually progressed onto complex algorithms. Big data can be a great resource for machine learning input. Such a big amount of information gives machine learning a huge platform for development.
​
- Jordan, M. I. & Mitchell, T. M., 2015. Machine learning: Trends, perspectives, and prospects. Science, 349(6245), pp. 255-260.
Digital Art
Another interesting artwork is surprisingly not so modern for this contemporary topic. It takes us back to 1982. At this point in history, computer development was at the very beginning of its journey. Back then, analogue art was a primal source of all artistry. People created art on paper and canvas mostly. On the other hand, it was a time of experimentation in art history. Harold Cohen was one of the first in his field. He decided to use a sketch that was generated by a computer and combine it with regular art by colouring it in (Cohen, 1982). This piece of art is captivating. At that time, it was a revolutionary idea. Something that humanity is so used to nowadays. He was really one of the first people who explored the topic of digital art. The computer was given the name AARON, this robotic machine would create large drawings on a sheet of paper on the floor (Tate, 2004). Our projects explore a very similar topic of the relationship between human imagination and artificial imagination.
Harold Cohen
Untitled Computer Drawing 1982
Tate
© Harold Cohen
GAN
This project works closely with the website mentioned earlier called thispersondoesnotexist.com. This website is based on GAN which translates to a generative adversarial network. The next article “A Review on Generative Adversarial Networks: Algorithms, Theory, and Applications” (2020) looks closely at GAN and talks about different algorithms, applications, and theories. GAN is one of the quickest growing areas in machine learning and as the author states “there are about 11,800 papers related to GANs in 2018.” (Gui, et al., 2020). GAN consists of two models, a generator that “capture the distribution of true examples for new data example generation.” (Gui, et al., 2020) and a discriminator that “is usually a binary classifier, discriminating generated examples from the true examples as accurately as possible.” (Gui, et al., 2020). This is the base principle that allows it to produce a very realistic outcome. GAN is leading in advantages over other generative algorithms. It can parallelise the generation and produce multiple outcomes at once. It has very few restrictions and it produces a better outcome than any other method. It is now used within “image processing and computer vision, sequential data… such as image super-resolution image synthesis and manipulation, and video processing.” (Gui, et al., 2020).
​
- Gui, J. et al., 2020. A Review on Generative Adversarial Networks: Algorithms, Theory, and Applications, s.l.: Cornell University.
Exhibition “Neurons, simulated intelligence”
The next exhibition for discussion is “Neurons, simulated intelligence” (2020) that took place in the Centre Pompidou in Paris, France. It shows an incredible collection of artists that based their work on comparison and similarities of technical progress and artistic representation around artificial intelligence (Centre Pompidou, 2020). This exhibition is based on the history of AI and machine learning. It particularly emphasises the crucial topic of brain cell neurons that were the model for artificial intelligence. It shows the evolution in understanding how neuron connections and the brain work from the sixteenth century to the modern days, empathising with the twentieth century. It shows how people learned to implement their knowledge about the brain into mechanics and computers. Another important aspect of this installation is the decision tree algorithm. It has been around for centuries and now found its place in AI. It works by transfiguring an input and different outputs depending on the algorithm that is put into a computer. This algorithm explains the whole process of machine learning and brain activity in one (Centre Pompidou, 2020).
Imagination and Creativity
“Imagination is the Seed of Creativity” (2018) talks about imagination as the foundation for creativity. The author describes imagination as “the ability to mentally simulate situations and ideas not perceived by the physical senses” (Gotlieb, et al., 2018). The authors present two types of imagination processes that are important to creativity. Social-emotional type that has the ability to reflect on multiple social perspectives and scenarios and temporal imagination that processes mental time. The authors also talk about the fact that imagination is not the only factor that produces creativity, it also relies on many other factors that are individual to each person. Nevertheless, without imagination, there would be no creativity because they both use the same cognitive process in the brain. Yet, creativity is much rarer than imagination because for it to appear it needs social factors, originality, knowledge.
On the other hand, this report talks about a very interesting thing that most people do not know about. Memory is a part of imaginative thinking because “humans are not able to “play back” the past like a movie-reel” (Gotlieb, et al., 2018). So, everything that a person remembers is the work of imagination.
​
- Gotlieb, R., Hyde, E., Immordino-Yang, M. H. & Kaufman, S. B., 2018. Imagination is the Seed of Creativity. In: The Cambridge Handbook of Creativity. New York: NY: Cambridge University Press.
Anna Ridler
In 2018 Anna Ridler created an enormous archive of ten thousand pictures of tulips during their blooming season. She created an installation called Myriad (Tulips) displaying all of them with handwritten notes on every single picture (Ridler, 2018). This archive was turned into a database for her next project “Mosaic Virus” (2019) that was presented in the earlier mentioned exhibition. She used this archive to make videos that have artificially generated tulips merging into each other on three screens. Machine learning was used as a tool to create new tulips the same way as the website that is used for this project thispersondoesnotexist.com creates new faces. She created an installation using the same principle of the generative adversarial network that would create tulips that do not exist. This is fascinating because her project is literally another example of computer-generated pictures from examples. She created a variation of something that this project is basing its research on. She also created the archive for this machine learning network to work. With this project, I am taking this a step further and using the outcome of GAN to compare it to human artworks to show that computers were made on a human example.
​
This artwork was crucial in the development of this project
