Are Deep Neural Networks Creative?
Gregory Piatetsky-Shapiro
Part-time philosopher, Retired, Data Scientist, KDD and KDnuggets Founder, was LinkedIn Top Voice on Data Science & Analytics. Currently helping Ukrainian refugees in MA.
Here is a recent KDnuggets blog by Zachary Chase Lipton, Ph.D. student at UCSD, working on Deep Learning and Neural Networks.
Are deep neural networks creative? It seems like a reasonable question. Google's "Inceptionism" technique transforms images, iteratively modifying them to enhance the activation of specific neurons in a deep net. The images appear trippy, transforming rocks into buildings or leaves into insects. Another neural generative model, introduced by Leon Gatys of the University of Tubingen in Germany, can extract the style from one image (say a painting by Van Gogh), and apply it to the content of another image (say a photograph).
Generative adversarial networks (GANs), introduced by Ian Goodfellow, are capable of synthesizing novel images by modeling the distribution of seen images. Additionally, character-level recurrent neural network (RNN) language models now permeate the internet, appearing to hallucinate passages of Shakespeare, Linux source code, and even Donald Trump's hateful Twitter ejaculations. Clearly, these advances emanate from interesting research and deserve the fascination they inspire.
In this blog post, rather than address the quality of the work (which is admirable), or explain the methods (which has been done, ad nauseum), we'll instead address the question, can these nets reasonably be called "creative"?Already, some make the claim. The landing page for deepart.io, a site which commercializes the "Deep Style" work, proclaims "TURN YOUR PHOTOS INTO ART". If we accept creativity as a prerequisite for art, the claim is made here implicitly.
In an article on Engadget.com, Aaron Souppouris described a character-based RNN, suggesting that higher sampling temperatures make the network "more creative". In this view, creativity is the opposite of coherence.
Can we accept a view of creativity which consists primarily of stochasticity? Does it make any sense to accept a definition by which creativity is maximized by a random number generator? In this view, creativity reduces to entropy maximization. Interestingly, this seems opposite to views on creativity espoused by deep learning pioneer Juergen Schmidhuber, who suggests that low entropy, in the form of short description length is a defining characteristic of art. More importantly, it seems at odds with the notion of creativity we attribute to humans. It seems uncontroversial to label Charlie Parker, Beethoven, Dostoevsky, and Picasso, as creative, and yet they their work is clearly coherent.
Read the rest of the blog on KDnuggets
Are Deep Neural Networks Creative?
https://www.kdnuggets.com/2016/05/deep-neural-networks-creative-deep-learning-art.html
Managing Partner, Rice Analytics - AI Architects, Author Calculus of Thought
8 年The biggest difference is that the backprop ANNs that may look on the surface to write in the style of Shakespeare and Linux code are random meanderings and jibberish without deeper associative and logical connection (the code usually does not compile and even if it does it does not solve problems; the Shakespeare does not tell stories). Creative outputs from human cognition have deep associative, logical and meaningful connection between surface elements so code compiles and solves problems and Shakespeare tells stories. "Deep" learning is a misnomer for these backprop ANNs that do not model real deep neural networks because backprop has never been observed in real deep NNs. "Hallucination" was the right word though, as they do make many random insertions even when learning photo images much like happens when a dysfunctional brain no longer engages in deep learning and instead hallucinates.
Founder & Lead Developer at academicum.ai | Building AI-Powered Tools to Enhance Academic Research and Learning
8 年Not convinced yet. A Turing test for art should be built: pieces of artwork, half for unknown artists and half from computers are shown to a panel of art critics... you know the rest.