A Look at: Technology and Drawing

Deep Dream
Google's Image recognition and manipulation AI

Google's frequent endeavours into machine learning consistently involve artistic qualities. Their deep dream AI is a clear example of how machines can draw - with the help of it's vast library of images and strong image recognition software it's able to 'look' for pictures of - in this case - dogs within an image.

This is an example of basic machine learning being directed toward an artistic outlet - describing potentially how technology can create works of art almost independently and dynamically.

Tubingen
Tubingen’s artistic style recreation system

Tubingen's researchers took Google's deep learning algorithm a step further into the world of art and drawing - emulating the styles of famous paintings or painters to recreate photographs fed into the program. The results are extraordinary and undeniably successful - describing huge leaps in machine learning algorithms. However doubts can be raised about how dynamic the algorithm is and the breadth of styles that could be recreated with such a high level of success as the 'paintings demonstrated in Tubingen's demo. 

These demonstrations not only describe the progression of the artist's tools through technology but also potentially a separation from the artist and the tool - raising questions on automation in a sphere, the artistic sphere, that might have previously been considered the furthest from human - machine automation. 

For me personally this further describes the power of computation and how its boundaries are forever breaking - defying what was previously thought as unlikely or even impossible by some. Machine learning - whilst still primitive in many ways compared to the biological - can excel where humans can never hope to because of its logical and rapid nature.