The opinion of the mathematician of Oxford: can the AI do?

17

The game of go, in which a computer program DeepMind defeated the champion among people, has created a kind of confusion for Marcus du Sautoy, a mathematician and Professor at Oxford University. “I always thought math with the game of go,” he says. And should not be a game that the computer is so easy to play because it requires intuition and creativity. So when du Sautoy saw AlphaGo from DeepMind won Lee Sedalia, he thought that in the field of artificial intelligence there have been changes, which will affect other creative industries.

The scientist decided to investigate the role that could be played by AI in our attempt to understand creativity, and wrote the book ‘The Code of Creativity: Art and Innovation in the Age of AI’, which was published in the edition of government at Harvard University.

Artificial intelligence and creativity: who?

The Verge discussed with du Southem different types of art, as the AI helps people become more creative (instead of replacing them), as well as creative areas in which AI is faced with the greatest difficulties.

Let’s first look at what creativity, or artistic creativity. In the book you talk about three types of creativity. What is it and what does it mean for the role of AI?

Many people think that art is an expression of what it means to be human, and if so, how AI can come close to that? I look at many artists and show that quite a lot of art pieces have a pattern and structure, which is highly mathematical in nature. That’s why I believe that artistic creativity can be more about the templates and algorithms than we think, and very often these patterns are hidden. Maybe AI could detect because he’s very good at finding hidden patterns.

There is no research work that takes the rules of the game and brings them to the extreme, as did Bach. There are combinatorial creativity when you take two ideas that have nothing to do with each other, to see how the Association of one can help to stimulate new ideas in another. The third work, which is somehow the most mysterious, it’s those moments that occur out of nowhere — like phases are, when you boil water, the water turns to steam and the state of matter changes completely.

How AI fits into these schemes?

Each of these creative approaches offers different challenges for AI. Research work seems perfect for a computer because it is able to produce much more calculations than the human brain. Combinatorial creativity is interesting — the AI can learn the patterns and apply them in new areas. But the most difficult for him to create something new and to escape from the system.

Normally thought: “How the AI can break the rules? Isn’t he stuck in the system because it is programmed to work a certain way? How to jump out?”. But if AI say: you gotta break the rules, it will also be the rule. You have a meta code that tells the program to break the code behind it.

In your book you talk a lot about creative projects AI. Which ones were particularly interesting to you?

One of the most interesting was jazz Continuator, who took the music of a jazz musician, studied the patterns and started playing independently. Striking was the reaction of a jazz musician. He said, “I understand everything they hear. This is my world of music. He plays just like me, except for those things that I never thought of before in my musical world.”

So I think this is one of the challenging roles of AI in the future. People often begin to repeat patterns of behavior. Ironically, we become more like machines, since they simply repeat something, so I’m impressed that the jazz Continuator made the musician a little think about their native behavior. He helped to awaken his creativity, showing that it is possible to rearrange the ingredients that he already had, and that’s about it not even thought of. I wanted to show that the role of AI in the works, perhaps, is to increase the creative potential of a person that is a partnership for the future that together we can make things more interesting than if they worked separately.

Another interesting story that, in my opinion, important, connected with the world of fine art and DeepDream from Google. Google gave a task to its software visual recognition to consider a random array of pixels and describe what he saw. By this means we learned something about how it was programmed artificial intelligence and how it is seen.

What’s the point?

One of the problems with modern AI is that a machine learning generate code, but we don’t quite understand how it works. The Google project DeepDream helps us to find a way to understand how this happens. Therefore, as for us humans, art is a way to break into the consciousness of another person, perhaps the art created by AI, will help to penetrate into the essence of this code, very mysterious.

Take Microsoft project Rembrandt, which creates the generated AI image in the style of Rembrandt. You could say: “Why do we need another Rembrandt? Do we yet have a fantastic Rembrandt?”. The point is that all of this helps to understand new works of art. If you look at the work of Jackson Pollock from a mathematical point of view, we see new things we missed before. So the AI might play an interesting role in discovering new structures that we may have missed in works of art and now take for granted.

This search for patterns is not restricted to only fine art, right?

Well, in the world of cinema can take the algorithm Netflix, which recommends movies that we like. He may divide films in new ways. Some of the groups we would identify as “all of the Comedy together,” but sometimes the movies are grouped according to the expression of people “like” and “dislike”, and then the overall theme eludes us. It looks like the AI has defined a new genre of film for which we have even no name. We can say that “there is a new scent that you need to call”. Maybe the AI takes our creative works and sees in them something that we can Express, but be aware — no. He could help us consciously to articulate the essence of creativity.

There are many creative fields. Name one where the AI has the hardest?

One of the surprises for me was how difficult it is to write words. In artificial intelligence there is so much written for learning. I was quite surprised that even though the AI is pretty good and writes literature in brief, he is still unable to write something for a long time. He doesn’t have a good sense of narrative line, for example. I haven’t seen anything that would extend the coherent story over three pages. Perhaps the AI is very difficult to formulate language constructs as weII as we do. Maybe he needs to go through the evolution that we passed. And then the question is: how long will it take?

To this question you can answer in our chat in Telegram.