The Art of AI
31 July 2022
Art competition - inspired by discovered hidden paintings using AI techniques
![Neo Modigliani "The Hidden Picture" of Beatrice Hastings by Amedeo Modigliani created by Oxia Palus using AI technology.](https://www.ucl.ac.uk/ai-centre/sites/ai_centre/files/styles/large_image/public/mog.jpg?itok=fuocC5DJ)
In June 2022, the UCL Centre for AI hosted it’s first ever art competition, inspired by "The Hidden Picture" of Beatrice Hastings by Amedeo Modigliani created by Oxia Palus using AI technology.
All members of the Centre for AI were invited to submit any art, in any medium they so chose, under the theme of “The Art of AI”. There were three cash prizes available and we received over 20 different submissions including photography, acrylic paint on canvas, AI Generated images, videos, poems haikus and textiles.
![](https://www.ucl.ac.uk/ai-centre/sites/ai_centre/files/styles/xl_image/public/collage.jpg?itok=EcgBm4-V)
Collage of some of the entries for the 2022 Art of AI competition
Our guest judges for this competition were Prof Kieren Reed, the Director of the UCL Slade School of Fine Art and Hannah Pratt an artist from Deeep digital
![Image of the function space hosting the Art of AI event](https://www.ucl.ac.uk/ai-centre/sites/ai_centre/files/styles/medium_image/public/art_of_ai.jpg?itok=asK620hj)
The winning entries were:
Honorable mention: SVG image - Evolving Open Complexity, by Bill Langdon
![SVG image - Evolving Open Complexity](https://www.ucl.ac.uk/ai-centre/sites/ai_centre/files/styles/medium_image/public/bill_langdon.png?itok=EkweiEP9)
"Evolving Open Complexity" is a rendering of a proposed software architecture in which information theory suggests a balance between the rigidity of purely functional code, represented as inverted binary trees, on a background of isentropic data transfer, giving an evolvable balance of modifiability and robustness, permitting constant change.
The data membrane is shown in the background by a freely available highly magnified section of healthy human lung tissue onto which are projected 1300 random programs clustered on connective tissue leaving wide gaps for program inputs and outputs to circulate. Lungs were chosen since they are emblematic of free gas (data) flow and large surface area. The trees are generated with Graphviz dot with primary colours indicating functional inputs and rendered at infinite precision with Scalable Vector Graphics by open source Ubuntu Firefox
84.0 before screen caption and image manipulation using the Gimp GNU Image Manipulation Program 2.10.28 and compression using 8-bit Portable Network Graphics
The paper is open access https://arxiv.org/abs/2112.00812
Honorable mention: "Creation of Adam", as per the modern age, by Ashish Goyal
!["Creation of Adam", as per the modern age](https://www.ucl.ac.uk/ai-centre/sites/ai_centre/files/styles/medium_image/public/ashish.png?itok=Pfns5Jal)
In the modern times, similar concept is recreated but depicting the birth/ creation of Artificial Intelligence, as an image of man (human in general). Since, the most prominent personality in today’s era, representing mankind’s progress in the field of AI and path-breaking technology, is Elon Musk, he is shown as imparting the ‘spark of life’ into AI (illustrated as a humanoid robot). Behind Elon Musk, shadows of human-like figures can be seen, which depict the entire human race. Through this work, I wanted to preserve the theme of the competition, by recreating a famous art work, to show the overarching relationship between man and AI
3rd Place: “Not a Picasso, not explicit, and not a copyright violation” by Pontus Stenetorp
![“Not a Picasso, not explicit, and not a copyright violation” by Pontus Stenetorp](https://www.ucl.ac.uk/ai-centre/sites/ai_centre/files/styles/medium_image/public/pontus.jpg?itok=iLnLm_T1)
The submission is derived from an arrangement of pixels created by artificial intelligence techniques, deemed potentially illegal and too offensive for a public space. The very same pixels were then transformed by a statistical algorithm into the piece itself which now despite being the same pixels is now legal and subjectively not explicit to those that previously would have considered it to be as such, thus I named it: “Not a Picasso, not explicit, and not a copyright violation”.
2nd Place, “Dynamical Thinking” by David Barber
![“Dynamical Thinking” by David Barber](https://www.ucl.ac.uk/ai-centre/sites/ai_centre/files/styles/medium_image/public/david.png?itok=fiJyZMSV)
On a technical level, the work represents a high-dimensional rotation -- one of the simplest dynamical systems that lies at the heart of many AI algorithms.
These systems are analogous to physical pendulum art and can be thought of as the motion of a 1000 dimensional pendulum, visualised in 2 dimensions. However, that such high dimensional oscillations can also generate calming projections is perhaps less well appreciated and reminds us of the beauty underlying some of our research.
Since the dawn of human art, spiral petroglyphs have represented the harmony of rhythms, which are fundamental to the human experience. Generations of artists have attempted to capture such rhythms; the work hints at a modern version of the older mechanical spirographs and echos abstract artists such as Bridgit Riley and Terry Frost, amongst many others.
The work therefore sits at the intersection of art, nature and science. It connects the ancient with the modern -- unbroken lines of harmony across generations of the human spirit.
1st Place “Into The Metaverse” bu Muna Aghamelu
![1st Place “Into The Metaverse” bu Muna Aghamelu](https://www.ucl.ac.uk/ai-centre/sites/ai_centre/files/styles/large_image/public/muna.jpg?itok=2pxOwHD5)
The lines between the physical and digital are definitely blurring now more so than ever. I created this painting to try to encapsulate my optimism for the way tech is going. I recently read Race After Technology by Ruha Benjamin and as a black woman in tech, I wasn't really surprised about the case studies she brought up just saddened. There were examples where technological advances in the field of AI had failed marginalized people and cases where its not designed to be inclusive and even sometimes maliciously designed to exclude people. Joy Buolamini and Iddris Sandu are names I hope become household names in the advancement of technology, AI and the digital space. We need more people in the tech space who are pushing innovation, representation, inclusivity, especially with the growing potential of AI.
Congratulations on winning our Art in AI award and thank you for the outstanding Artwork.