art review











Art of the Machine

Nov 16th – Dec 7th

Can AI be truly creative? Can algorithms and machines write poetry or produce art? Is AI an existential threat, an opportunity for greater innovation and artistic collaboration …. or both? In this show we display the work of some artists who have worked with AI in a collaborative fashion to explore these creative possibilities. 


In recent years AI has taken the world by storm and raised all kinds of challenging questions about human self-identity.  Humanity got used to the earth not being the centre of the universe. Then we found we are not so distinct from animals. But still we have our unique intellect: “Cogito ergo Sum”.

But now Artificial Intelligence challenges another human frontier – rational thought. Already AI has defeated all the great games players and is gradually learning to drive, talk, walk and compete with humans in the intellectual workplace. 

Thankfully creativity will always be a human domain ...won’t it?

The pioneer of AI in Art is Harold Cohen (1928 – 2012), a British Artist who moved to California and started working with some of the early AI pioneers in 1968. He built a program called AARON which could control a drawing machine and spent the next 40 years developing and evolving this program, coding it with his own principles of artistic practice and addressing the engineering challenge of realising the work on paper or canvas. It was, in the parlance of the day, an ‘expert system’: an autonomous system programmed with knowledge of the expert, in this case Cohen. 

Pindar Van Arman has been developing his own artist collaboration with his painting machine, ‘Autonymo.us’, for many years. He has been able to harness the new wave of computational algorithms that emerged in the 1990’s when new algorithms for artificial neural networks (ANN), genetic algorithms, support vector machines and many others became better understood. 
These neural networks were the precursors of ‘Deep Learning’ that arrived a few years ago with some additional refinements to the learning algorithms. Pindar uses GANs, ‘Generative Adversarial Neural Networks’, amongst others, in a process of generation and challenge between neural network until a final state is reached. The work is then rendered from the digital representation onto the canvas by the painting machine. 
Pindar regards his work as a process of dissecting his own artistic process and teaching it to robots, providing a better understanding of creativity. His most recent robots, using deep learning, artificial intelligence and generative algorithms are able to make a surprising amount of independent aesthetic decisions, he says.  

“They are beyond being simple assistants. My robots are now effectively augmenting my own creativity.”
The work on display are facial images generated by the GAN algorithm. The GAN is used to generate better and better faces, iteration after iteration. Once the system is able to recognise the image as a face it stops and renders the image.  

Patrick Tresset is a Brussels based artist who develops theatrical installations with robotic agents as actors. Tresset’s installations use computational systems that aim to introduce artistic, expressive and obsessive aspects to robots’ behaviour. These systems are influenced by research into human behaviour, more specifically how humans make marks, depict other humans, how we perceive artworks and relate to robots. 

Originally a painter, Tresset is part of a generation of artists coming out of Goldsmiths’ College (London). His work has been internationally exhibited in association with major museums. He has also published research papers in the fields of computational aesthetics, social robotics, drawing research and AI.

The show starts on 16 November

Follow us:
Instagram @dekka_arts
Facebook @dekka-arts




6 Lombard st Margate CT9 1EJ



16 November

7 December