Computers are useless. They can only give answers
— Pablo Picasso

Humans are continuously being knocked out of their helm in the gaming world. AI agent AlphaGo defeated the best of human players Lee Sedol and Ke Jie in the game of GO, soon after, the AI moved on to learning other games like StarCraft II where human pros like TLO and MaNa ate its dust. The AI agent, named AlphaStar, managed to pick up 10 wins against the two StarCraft II pros in two separate five-game series that originally took place back in December. After racking up 10 straight losses, the pros finally scored a win against the AI when MaNa took on AlphaStar in a live match streamed by Blizzard and DeepMind.

Strategic activities like gaming have been very earnestly practiced in human societies since a long time. Another activity which humans have embraced extremely close to their hearts is Art.

Art in the age of AI

The opening portrait of this article depicts a portly gentleman, possibly French. The work appears unfinished, the facial features are slightly indistinct and there are blank areas of canvas. Oddly, the whole composition is displaced off the center of the canvas towards the top left corner. Though the giveaway clue to the origins of the work is the artist’s signature at the bottom right. In a cursive script it reads —

Yes, this is a portrait painted by Artificial Intelligence and the equation above is a part of the algorithm which was used to make the portrait. This work was auctioned for a whopping $432,500. At first look it is not entirely obvious that this portrait was not made by a human being. In fact this is not the only artistic feat achieved by AI . In recent years AI has managed to —

• Compose a piece of music (Transits — Into an Abyss) that was performed by the London Symphony Orchestra and received praise from reviewers. You can hear the performance in this link
• Identify emotions in photographs of people, and create an abstract painting that conveys these emotions to the viewer. The AI can even analyze the painting as it is being created, and decide whether it’s achieving its objectives.
• Write a short novel that almost won a literary prize in Japan.
• Create a movie trailer — watch it here.
• In Rio Olympics 2012 Washington Post had used AI to make news more interesting and informative

AI has the ability to learn and improve itself by experience by gorging upon huge amounts of data. The agent learning the games AlphaGo and StarCraft II played a million games and learned, improved and evolved its strategy to beat human players.

The AI which created the portrait of Edmond Belamy was trained on thousands of images which led it to create this portrait. The portrait still had some human intervention but largely it was created using an algorithm called Generative Adversarial Network or simply GAN. The algorithm’s invention is credited to Ian Goodfellow.

While this portrait though generated by an algorithm is uncannily similar to a work by a human artist, can this portrait be thought of as a work of art? Should there be a threshold of influence over the final product or the inputs to these algorithms that the artist or designer needs to wield for the product to qualify as art? Do these machines possess creativity?

Cre(AI)tivity

The GANs used for the portrait above used many images as inputs to the network. The network which created this portrait was shown curated portraits made by lot of different artists. From a macro perspective the network created a different version of what it was shown, a deformed portrait, it did not create something entirely different or novel. Though the generative algorithm can produce images that surprise even the artist or designer presiding over the process of generating those paintings. Should novelty be the only parameter for creativity or art?

Psychologist Daniel E. Berlyne has studied the psychology of aesthetics for several decades. He found that novelty, surprise, complexity, ambiguity and eccentricity tend to be the most powerful stimuli in works of art.

The portrait created by AI is certainly novel, surprising and bizarre and so are paintings created by the British artist Francis Bacon —

The difference between the two deformed portraits, AI made Edmond Belamy and Francis Bacon’s triptychs, lies maybe not much in aesthetics but in intent. It was Francis Bacon’s intent to make his paintings deformed while Edmond Belamy, the outcome of a GAN may not necessarily be intentional. Which raises another question whether the intent of the artist really matters? A really nice video exploring some aspects of this question by TedEd.

Future of AI and Art

Surely AI has created a lot of high valued art works but currently AI in my opinion is no way close to the level and works of most experienced human artists. But the ability to learn and learn at lightning fast pace leaves no reason why AI will not be able to achieve the feats of the gaming world in the realm of arts. In the next decade, we’ll see AI composing music and even poems, drawing abstract paintings, and writing books and movie scripts. And it’ll get better at it all the time.

While Artificial Intelligence is developing at a rapid pace and as technology starts to intertwine with humans, we have a harder task at our hands. To keep answering and asking the fundamental questions — What is Art? What makes us humans? Some of those questions are relevant for the future of humanity, some are just food for thought.

Art makes us human, music makes us human, and I deeply feel that science makes us human.
- Brian Greene

X8 aims to organize and build a community for AI that not only is open source but also looks at the ethical and political aspects of it. Look out for a simplified explanation of Random Forests on X8 this Friday. If you liked this or have some feedback or follow-up questions please comment below.

Thanks for Reading!