The undesirability of AI tailored content

What is Art?

Many people think they know what art is… but many people are wrong. Art is about defying vulgarity.

If your whole Paleolithic tribe is talking about a bison hunt and you can transform that tale into an image on a wall, you are defying vulgarity.

If centuries later most of the cavemen are mechanically painting bisons on a wall , but you incorporate the cracks of the stone wall as strokes to outline your drawing, adapting the shape of your bison to the shape of the wall, you are defying vulgarity.

Vulgarity is not bad. It is just ordinary. Any form of Art is created in an effort to overcome vulgarity.

Composition with lines

Piet Mondrian defied the vulgarity as one of the pioneers of abstract art by moving away from representational forms in favour of geometric shapes. His geometric abstractions had a huge impact on architecture, fashion and graphic design. This is one of his paintings from 1917: “Composition with Lines”.

Explorers vs Tourists

Art is about exploration. Exploration is done by explorers. Explorers are experts in the field. Artists are explorers. The role of artists (thinkers, writers, makers, designers…) has historically been to move ahead of the population, discover something new and wait for common people to catch up at some point, if ever. That is the difference between explorers and tourists.

Vanguards are the product of intellectuals who run away from vulgarity faster and sooner than the rest of us. Of course, Mondrian was an explorer.

The Noll experiment

A. Michael Noll, an American electrical engineer and pioneering computer artist at Bell Labs in Murray Hill, New Jersey, created in 1965 “Computer Composition With Lines”. He generated this artwork algorithmically with pseudo-random processes to mimic Piet Mondrian’s Composition With Lines. In what became a classic experiment in aesthetics, copies of both works were shown to people, a majority of whom expressed a preference for the computer work and thought it was by Mondrian. The work won first prize in August 1965 in the contest held by Computers and Automation magazine.

Missing the point

There is no MVP for Neoplasticism, Abstract Art or Cubism because artists do not care about users when trying to discover new domains. There is no previous survey to guarantee that users will like it. There is no usability test to check that people understand it. There is only the instinct of a creator who is convinced that there is a harbour awaiting.

Asking people whether they prefer the Mondrian image or the computer-generated one is, to put it softly, missing the point. People happened to like the one by the machine better. So what? The machine did not break the laws of art 50 years before. The machine did not show people another type of images to like. The machine did not influence a whole generation and several disciplines. But Mondiran did. He arrived in one place where no one had ever been before and he waited for common people to get there waaaay later. Will the machine be able to do it in the future? Maybe. But this is also not the point.

What this experiment and AI image generation tell us about art is that imitation continues to be easy, and also that defying vulgarity continues to be art (and not so easy).

Tailored content · A use case

What is all this “exploration reasoning” about? Bear with us. We are getting there.

Imagine a future scenario where all kinds of data are taken from wearables: watches, wrist bracelets, ankle bracelets, collar microphones, glasses… An AI would process all that data taken from your speech and biometry. It will know you better (more objectively) than yourself.

Now you want to buy some trousers. Your AI assistant will know exactly what type of clothes you like: they should be produced of recycled material in fair conditions. Size 40. Your thighs and bum are larger than your waist. Ideally transportation routes are minimised. Cut: Straight large, high waist, in black. Secondhand is preferred to new. It will even know you gained two kilos!

After some questions to refine the search, the AI is done. You unfold your screen and look at a UI full of auto-generated models, posing with your body shape, wearing the kind of trousers that you asked for. You can see exactly how they would look on you. AWE-SOME

Now. Imagine you very well know what you want, but your taste has not changed in the last few years. Your preferences will show up all over the place. Wouldn’t you end up being tired of yourself? You would want to check the trends. Leave your island. At some point you would be willing to forget about your own taste and expand it upon the exploration of others.

AI prophets sing their praises for personalised content but it is easy to envision a scenario in which you get sick of it and want to switch back to a discovery mode that allows you to re-connect with the designers and search for new trends. Artists will, again, save us from ourselves.

And here is where the analogy between Mondrian’s act of creation and the algorithmic act of imitation from Bell Labs comes into place.

Final words of wisdom

Of course tailored content is desirable. What is undesirable is how Western society becomes obsessed with the latest trends and forgets that there is a whole world out there; a whole world full of explorers specialised in their different terrains who do not care about what you like now but rather what you will like next. A whole world inviting you to leave your comfort zone; inviting you to take a quick look at what they are proposing.

AI is amazing and it will get even better. Our aim is to help people get there and encourage them to use it cautiously and consciously. We aim to reflect, ponder and decide what kind of world will we build with it.

And remember: we are the good guys.