Mario Klingemann is considered a pioneer in neural networks, computer science and artificial intelligence.
Anna Ridler, one of the most famous artists whose work involves extensive coding and neural networks, also believes that some processes are necessary to create artificial works of art.
Ridler and Klingemann commented on one thing: the art market relationship with machine learning and AI is still in its infancy.
Rama Allen (who uses automatic learning to create engaging experiences and interactive art) said: “The defining technology of our time will be AI.”
GANs (generative adversarial networks) entered the AI conversation in 2014, when researcher Ian Goodfellow published an article that explains that it was the next step in the development of neuronal networks: interconnected layers of the node of treatment, loosely shaped on the human brain, which have triggered many recent advances in AI.
Decades of sci-fi exposure have conditioned the general public to associate every AI call with what technologists and researchers would call ‘artificial general intelligence’, or a machine with the ability to think freely, to learn without specific training and perhaps even to experience emotions.
In fact, AI has attracted the attention of artists who are more involved in the concept of technology and more deeply rooted in the structures of the traditional art market.
Tabitha Goldstaub, a British tech entrepreneur, said: “The act of AI systems training to create works of art raises fascinating questions about the nature of creativity.”
If creativity such as the application of Instagram filters, creating Pinterest boards or drawing in adult colouring books is a bar, then machines can already do so.
Like other new technologies (such as photography), I think it will become its own domain.
But of course, the emergence of new technologies can change existing fields. Photography has had a huge (and probably positive) impact on painting, which has helped to stimulate the creation of non-representative painting forms at the turn of the century.
AI can certainly improve artistic capabilities, but I doubt whether it will override human artists because people are obliged to choose and train an algorithm, set neural network parameters, etc.
Computational creativity is the study of building software that exhibits behaviour that would be considered creative in humans.
Such creative software can be used for autonomous creative tasks, such as inventing mathematical theories, writing poems, painting images and composing music.
However, computer creativity studies also allow us to understand human creativity and create programs for creatives, where software works as a creative collaborator and not as a simple tool.
In other words, physical laws, theories and music tracks can be generated from a finite set of existing elements, so creativity is an advanced form of problem-solving, which includes memory, analogy, learning and reasoning, among other things, and can therefore be replicated by computers.
AI has played a key role in the history of computer music since the early 1950s.
Artificial intelligence has existed for 50 years, but its use in creating artworks is part of a new wave.
Klingemann, one of the pioneers in the use of AI in art, sees it as a means of pushing the boundaries of human cognition.
He constructs generative models which he combines with the output of one to train the other until the final images are a distant, distorted collapse of the original input.
Klingemann is an artist whose preferred tools are neural networks, codes and algorithms rather than brushes and paints.
He hopes to understand, question and subvert the internal functioning of systems of all kinds when exploring machine learning.
Ridler creates unique datasets for training her models, eg taking thousands of tulip photos and training an AI to create a blooming video, controlled by the price fluctuations in bitcoin.
Two artificial neural networks are used: the generator which constructs real and artificial samples, and the discriminator, which evaluates the power of the generator to push it towards improvement.
In contrast to GANs, which reproduce well-known artistic styles and themes, CANs (creative adversarial networks) produce more original works thanks to the addition of a style ambigram, which gives more room for creativity.
Has AI given us the next great art movement? Experts say slow down, the field is in its infancy.
This article was written by AI-writer, an artificially intelligent content creator, from QLX. See how AI-writer works at content.QLX.services.
To read more stories written or curated by AI Editor Squire, visit our AI Squire section.