Switch off the editor's digest freed from charge
Roula Khalaf, editor of the FT, selects her favorite stories on this weekly newsletter.
There is a certain style of artist who longs for it to transcend the content of your individual head. For a protracted time I wanted a extremely formible painterly area that has the speed and dynamics of a fantastic abstract painting, which is just designed from the figuration and is made with pictures.
There can also be a story of artists that include technological innovation: screen printing, video, neon. Everything was latest once, even oil paint. It was assumed that photography signals the death of painting until artists who were as diverse as Thomas Eakins and Edgar Degas began of their work. Now we’ve artificial intelligence. Could Ai help to acknowledge my vision of a liquid visual space that was free of the gravity laws?
The AI art I had seen was not promising. The more futuristic the photographs, the more banaler the effect; The pictures didn’t appear to have inner life. Was it the fault of the technology or the user, a mistake of the automotive or driver? What did the machine not “understand” about visual art? On the one hand, digital images must not be mistaken for painting. Pixels don’t have any surface structure, no physical presence in any respect. Painting, irrespective of how cerebral, is an object on the earth. I’m a painter, not an engineer; I’m on the side of the things that an creator did by hand. And yet I actually have spent teaching a pc for the past two years to think like an artist.
Artificial intelligence is basically a average machine for the universe of images from the Internet. The basics of representation painting weren’t a part of the machine's neural network. Two things, each essentially for visual art, disappeared within the indiscriminately data scratch of AI: specificity and intentionality. The most important was not a sense of treating the place where two forms meet. How a painter dissolves the sting, wearing his style essentially. Is the sting within the burst and expressive or narrow and limited? In a digital picture, an edge is usually approached since the machine tries to guess the subsequent step, but in any case only different coloured pixels; The intention is missing and doesn’t house any bow.
The saying from the early calculation days “garbage in, garbage” was still relevant. To get somewhere, the machine would must be sent to the art school. Through the enterprise consultant Tom Cohen, I used to be introduced to a talented young technologist named Grant Davis. Together we began to design the curriculum.
We first convey the machine on a handful of paintings by Giorgio de Chirico, Edward Hopper and Arthur Dove, three masters of the twentieth century, the drama of perspective as an organizational device, the flexibility of the worth pattern, a sense of volume, and the lyrical use of the colour black. For these lights, we’ve added a closely edited choice of my paintings from the Eighties and 90s, which highlight a collagen -like juxtaposition for composition. The visual IQ of the machine rose by several points, but the sides were still weak.
We re -elected the art history hour and rewritten the machine in quite a lot of my paintings from 2001: extravagant, fragmented compositions loosely on a nineteenth century opera set. This romantic, pastoral picture-a man and a lady who sits on a dam with a lake in the center ground and mountains in the gap was painted in striking coloured pitches. As a final step, we added my paintings from me, wherein forms are rendered in crucial, expressive brushstroke.
While we finely coordinated the model, Davis developed a software program for visual artists called “Magic Wand”, which eliminates the necessity for verbal input requests. This works like a joystick, a lever that moves from “similar” like “dissimilar”. We fed a fancy “pastoral” painting to the machine, placed the dial between the 2 extremes and held their breath.
The machine now proved to be a superb student. My compositions were melted down, put in a blender and walked through a sieve. The results were clearly derived from my work with many known elements, however the reconstructions were also different from anything that was previously seen. Since the machine had no idea what it did, she was capable of break the foundations of continuity, scaling, anatomy and visual logic with impunity. (The meaning of the machine that doesn’t know its violations was the important thing; more artificial is to the purpose.)
The pictures were still just pixels on a screen. I chosen essentially the most promising scenes and printed them on canvas. These prints then formed the backgrounds that I normally painted, and sent a collage of figurative elements that result from the background vortex. Male and feminine torsos; gesticulating hands; Precarious balanced tea cups are painted with a big brush with heavily saturated colours. The compositions are dynamic; Everything sails through the liquid, top-system-turbulent room. The paintings speak to the Gravitas of Art. But also they are related to an uncanny “future presentation”. They are Palimpsest from a dialogue with my earlier work – akin to “a duet of 1”, as a curator friend expressed it.
AI is a strong tool, nevertheless it's still only a tool. It is the creator of anything in the meanwhile. It works with what it was taught, because it was directed. The “considering” a part of the machine, the algorithm, seems to have a drive to stop pictures and recombinate their component parts, to look their spatial orientation, but to preserve and even increase their emotional undertext. What could possibly be more expressive this moment?
