Can Artificial Intelligence Create Art?

You probably know modern-day robots are incredibly sophisticated and algorithms increasingly complex. Artificial intelligence, or AI, is taking over many tasks humans used to fulfil. Ranging from facial recognition to proofreading contracts to driverless cars, our daily activities are gradually automated and algorithmicised.

You probably also know that the rise of AI and robots is causing numerous concerns. Some say robots are coming for our jobs, not only replacing personnel who routinely perform identical tasks, such as factory workers or secretaries, but also lawyers and doctors. Whilst renowned public figures such as Richard Susskind have predicted the downfall of the legal service some time ago, increasingly complicated chirurgical operations are being performed by intelligent robots as well.

Others go even further, and believe machines will soon become more intelligent than their human creators. Stephen Hawking said that it is only a matter of time before AI defeats human superiority, as algorithms continuously learn and improve themselves from experience. In short, AI may soon encompass all of human life. But what does ‘human life’ mean?

When we say something is human, we mean that it requires intelligence to be achieved. AI enables machines to recognise faces, drive cars, and guess what films we might like based on what we have previously browsed on Netflix – all things which require what we consider ‘intelligence.’ We teach machines this intelligence, which makes us fundamentally ‘human,’ through algorithms, thus making them appear or become ‘human,’ too.

Until recently, one of the few areas on which AI had relatively little influence was the art world. It has proven most difficult to teach machines emotions and imagination, and one of the last strongholds of human exceptionality seems to be creativity of the mind. But even this is changing. On the 16th of October 2018, Christies sold a work of art created by algorithms for almost half a million dollars, making it the first auction house to sell AI art. The painting in its gilded frame portrays a gentleman, possibly French, and judging by his dark frock coat and plain white neckline, a man of the church. The computer software that “trained” itself using a set of historical paintings for reference, resulting in this new image, resembling an 18th-century portrait, which was created by a French art collective called Obvious. Its slogan boasts that “creativity is not only for humans.” As such, the sale was symbolic for the future of art, signalling a change in the art community’s approach towards AI.

Logically, this caused significant controversy in the art world. Originality is not necessarily an issue; it may be acceptable that machines, just like humans, can be ‘inspired’ by previous artworks, as the Obvious software was. The controversy arises more around authorship, morality, and authenticity, as well as the idea that art needs a reason.

The truth is that fear drives a significant amount of the criticism the art world has expressed towards robots and AI. Just like fears of automation taking away industrial jobs from humans, artists increasingly consider AI too close to their comfort zone. But there are other factors which differentiate the work of humans from robots, and most art specialists confirm that for the time being, these differences cannot be overlooked. Many experts believe that until AI can be emotionally intelligent enough to create art without human intervention, it is merely a tool to assist artists in their creative process. It seems that learning from other (human) artists and data analytics to reproduce art, as algorithms do, is not considered enough to give the robot credit for the creation of art itself – at least, not yet.

Leave a Comment