Artificial Intelligence and Art
Machine learning as part of AI and how it has impacted on Art, investigating its creative potential.
produced by: James Tregaskis
Artificial Intelligence and Art
The field of AI/Machine learning has expanded so rapidly since 2011 conjoined with advances in Deep Learning and big data. AI/Machine Learning (ML) has had an uneven progress; in 1952 Arthur Lee Samuel coined ‘Machine Learning’. The rapid progress of computer hardware in the 50’s/60’s this may have led to an overblown prediction by Turing who said in 1950 that computers will have the same abilities as humans by 2000. LeCun, Hinton re-branded the term ‘neural nets’ with ‘deep learning’ publishing in 2006. This led on to a tsunami of research into AI today.
We mapped artworks (human,human-computer and computer) into y axis representing manual-autonomous and the x axis machine-human made. An assessment of the relative creativity of art works by computer can be found in work ‘Measuring Creativity’ by Elgammal and Saleh. How useful this is in creation of original art, is in doubt here. Do painting machines offer creativity combined with non-human created ‘art’ – or complex rules based programming?
Early work by Harold in self-generated art, AARON was based on rules-algorithms. Examples of art created by computers, free of human rules and based on ‘independent thought’ has yet to be achieved, indeed, is this ever possible, as humans are creating the machines in the first place? Examples of this are not purely confined to visual art but can be found in prose poetry and music (e.g. using Flowcomposer) as well. Although impressive enough, AI generated music, prose and poetry draws heavily on ‘mashing up’  an existing human created corpus of work.
Google plays a big part in the development of AI, creating open source Tensorflow. TF provides entry to AI, accelerating uptake.
Examples of Photo Style Transfer (as a subgenre of AI in art) exists, although a technically accomplished, relies on human creativity with little creative merit.
Karen Zack , questions the extent to which AI is Intelligent and not simply glorified pattern recognition. The recent rapid progress and refocus in AI research has led to startling jumps e.g. driverless cars etc. and this will create disruption in jobs. However, we must not get carried away with the idea that machines will put artists out of a job; photographs of brain cells, dendrites are casually juxtaposed against diagrams of multi layered software architectures, suggesting they are the same. As Krakauer points out, this does not explain the emergent behaviour of our brains, captured by Deep Learning or any other form of AI. AI is continuing to have a huge effect of the planet, hopefully with positive outcomes and its use in Art continues to take shape. It is questionable how AI on it own will out-run human beings in creativity but AI itself without doubt will change human culture immensely, thus in turn influencing the art humans make.
 Samuel, Arthur L. (1959). "Some Studies in Machine Learning Using the Game of Checkers". IBM Journal of Research and Development http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.368.2254&rep=rep1&type=pdf
 Turing, Alan M. 1950. Computing machinery and intelligence. Mind LIX:433-460.
 Gregory Piatetsky, ‘KDnuggets Exclusive: Interview with Yann LeCun, Deep Learning Expert, Director of Facebook AI Lab’ Feb 20, 2014. http://www.kdnuggets.com/2014/02/exclusive-yann-lecun-deep-learning-facebook-ai-lab.html
 LeCun, Y., Chopra, S., Hadsell, R., Ranzato, M., & Huang, F. (2006). A tutorial on energy-based learning. Predicting structured data, 1, 0.
 Blog page for diagram of Machine x creativity map https://docs.google.com/drawings/d/1QGvAiW3K8ciyIUhXkNf9dSB1PAxXaP1d8BS_-mRMvJ0/edit
 Quantifying creativity in art networks Ahmed Elgammal, Babak Saleh https://arxiv.org/abs/1506.00711
 E-David painting machine http://graphics.uni-konstanz.de/eDavid/
 Cohen, Harold. "Toward Diaper-Free Autonomy". aaronshome.com. Archived from the original on 1 July 2015. Retrieved 29 June 2015 https://web.archive.org/web/20150701130409/http://www.aaronshome.com/aaron/publications/mcasd.doc
 https://arxiv.org/pdf/1511.06349.pdf ‘Generating Sentences from a Continuous Space’ Bowman, Vilnis, Vinyals, Dai, Ozefowicz, Bengio.
 https://www.reuters.com/article/us-sony-algorithm/sony-develops-algorithm-based-ai-music-idUSKBN12H1ST Sony Paris development of AI music
 https://www.jukedeck.com/ UK based startup developing AI generated music
 http://www.gohomeproductions.co.uk/ Mark Vidler ‘Go Home Productions’ Fine proponent of mashups, not AI generated.
 https://www.tensorflow.org/ Open source library for machine intelligence
 http://cs231n.stanford.edu/reports/2017/pdfs/412.pdf Photo Style Transfer in Tensor Flow Kane, Lemionet, Shemaj
 http://databasecultures.irmielin.org/artificial-des-intelligence/ Artificial Des-Intelligence or Why machines will not take over the world. At least not now.
 http://www.cell.com/neuron/fulltext/S0896-6273(16)31040-6 Neuroscience Needs Behavior: Correcting a Reductionist Bias Krakauer, Ghazanfar, Gomez-Marin, MacIver Poeppel
 How Hard is Artificial Intelligence? Evolutionary Arguments and Selection Effects https://nickbostrom.com/aievolution.pdf
 https://www.wired.com/2017/04/the-myth-of-a-superhuman-ai/ The Myth of a Superhuman AI Kevin Kelly
 http://fortune.com/ai-artificial-intelligence-deep-machine-learning/ Roger Parloff Why deep learning is suddenly changing your life