Can Robots Replace Human Artists?

Feb 17, 2018 · 5 min read

Imagine this future scenario where robots and humans equally stroll the planet: You are invited to attend the opening of a significant art gallery showcasing creative artworks, only they are generated by a robot, or in other words “Artist Robot.” Does this thought disturb you by any chance? If not then we might be on the same page.

Automation is, in fact, a considerable concern to many people, the subject of artificial intelligence replacing jobs normally operated by humans has been investigated, studied and calculated very thoroughly to inform us that automation is “inevitable.” It is said that jobs that require “less than five seconds of thinking” will disappear, leaving routine, repetitive and predictable ones more in danger of automation. This leads us to think:

How can we appropriate automation and artificial intelligence in the creative fields without the risk of competition?

“Originality is what makes us different.” This was my initial argument because I am a human, and as humans, we like to believe that creativity is singular to our kind, don’t we?

Origins of originality have always been a very controversial subject. Innovation is mainly defined by the ability to think independently and creatively. Thus many believe that original ideas do not exist anymore, that creativity is necessarily a blend of prior knowledge, altered, associated and reproduced to form newer works of art. As a result, a combination of processes, borrowed ideas, and altered ones create what we call “the creative idea.” It was widely agreed upon in the past, that computers can not generate creativity, and in attempts to support this assumption, several studies, articles, and books explained how creativity is the outcome of complex, mysterious processes taking place in the right section of the “human brain.”

Today we conclude that these assumptions are no longer valid. Machines are rapidly developing and will eventually be able to adopt creative thinking processes. Bottomline, we can’t beat robots, so we might as well join them.

Patrick Tresset, a French scientist who has been developing creative robots for more than seven years. Tresset was fascinated by the human behavior of perceiving art. Through surveys and various other research methods, Tresset was able to define how we see art and project it ambiguously. As a part of his research, he designed a machine that could support his findings and embody his interpretations. This machine is the sketch-bot “Artist Robot” I mentioned earlier. Allow me to introduce him properly to you; his name is Paul-IX, and he can draw an outline still-life better than most of us. Paul-IX was manufactured to balance computer vision and artificial intelligence with human qualities such as the way we interact with art. Paul-IX’s skills might be unpolished, but machines are growing smarter and developing better skills at imitating human creativity. However, this doesn’t necessarily mean it should be replaced. A sketch-bot like Paul-IX has the potential to assist artists and educators — such as myself with initial sketching, replicating objects in correct proportions, as well as demonstrating basic drawing techniques to students. Automated design technologies can also be applied to firms to produce generative posters for small ad campaigns. I believe it is an efficient way to cut down costs, consume less time and effort collectively.

One of Google’s Artificial Intelligence (AI) experiments is AutoDraw, a platform that allows people to create a fast drawing, AutoDraw is smart enough to guess what the picture is, and later enable the person to choose the matching illustration from a more sophisticated list of drawings generated by the software itself. On the other hand, initiatives such as “NIIPS for Creativity” are using art produced by neural nets and experimental machine learning systems. Their work includes collage, photography, digital art and music among other disciplines.

So it is fair to say that robots have all it takes to imitate human creativity, but can they embody human emotions and experience and further replace them?

The other day, I walked into Greene Naftali, an art gallery in Chelsea, Manhattan. They were displaying ten large-scale monochromatic artworks by Jacqueline Humphries. From a distance, I could see graffiti brush strokes and blown-up emojis on the surface, but when I took a step closer, I realized that the paintings were covered with micro size cutout and stenciled letters deriving from computer text and coding. My impression was that the artist was communicating her feelings and thoughts on the contrast of the bold abstract expressionism of the past, and the modern-day digitalized emotions. The artist was suggesting that, in our visual world, nothing is what it exactly seems. I must say, I instantly related to her because I have experienced this change in behavior myself, which eventually brought me to appreciate her art even more. Such a connection is very improbable to exist or form between a human and a creative robot.

The way I see automation being successfully situated in the creative realm is as the following:

Artificial intelligence can be used either to purely increase work efficiency by replicating styles and artistic techniques to save time and effort, or to elevate our creative capabilities and forms of artistic expression to newer levels.

For some, this is a daunting trajectory in art, but these systems emerging today bring more than just generic algorithmic interpretations, they are providing a more comfortable mean of artistic communication and articulation. Hence, helping people better frame their creative expression.

All these creative approaches to AI in art and design reassure us that the existence of artificial creativity is indeed inevitable in a way or another. However, they do not suggest the end of human creativity. Creative processes can be automated, but the authenticity and emotional dimension to art produced by us “humans” is much more complicated and personal, it is irreplaceable.