Our series of talks with Federico Faggin goes on with another crucial question. Are robots capable of producing art?
As you previously explained, consciousness is the main property that distinguishes humans from robots. Within an imaginary competition between humans and robots, would consciousness be considered as an empowering tool for human beings or should it be seen as an element that increases their vulnerability against the “Artificial Intelligences”?
It depends on how the game is set up. It is obvious that we can create robots that can be much better than us in doing very specific skills. For example, a computer can do gazillions of multiplications in a second, when I would probably take one minute to do a regular multiplication and, if it was a division, even three minutes. Does that mean that a computer is superior to me because it makes so many computations a second? Do I feel myself inferior to a computer? No, because I created one, so the computer is my creation and, also, we are the ones that taught the computers how to compete with us. Do you think that the computer bit was able to figure out all by itself? Of course not, there have been thousands of people working on it, making it possible, and teaching the computer everything in order for it to do better.
Machines can improve thanks to “neuro-networks”, which are structures that humans have in their brain, imitated and reproduced within the computer. Artificial intelligences can learn things if we give them the opportunity to do so, and they can even learn better than we can, as long as the information is in the data the computer has access to. Creativity doesn’t work that way. It works as an inspiration, as a new tool that comes into us; it is something that didn’t even exist before. An original idea does not come in a symbolic form, but it is a comprehension that occurs and is transferred in our head. This comprehension is then translated into a symbolic form, which everybody can easily understand simply by seeing, or intuitively interpreting it. For example, Turing came up with the Turing machine, an unbelievably powerful idea, which is the one behind the computer. If Turing had explained the concept behind his machine only by words, very few people would have understood it, but, once he built it, everybody could see what he had in mind. Therefore, it is a process that goes from some kind of immateriality and non-symbolic understanding, which is something that we have inside us, into an expression in the form of symbols that becomes part to the world.
That’s how it works. Computers are simply stuff that we put together to do better than us things that we can describe and for which we can give him data. And that’s all.
As testified by many installations within the SG Dublin season: "Humans need not apply", machines are not only able to perform complex scientific functions in an equal, if not better way than human beings, but they may also show artistic capacities, which have always been related to an expression of feelings and consciousness. Installations like the Mindfulness machine proved to be able to realize paintings based on their "mood" (which is determined by a number of variables), others like Word.camera wrote poems based on caught images. Do you believe that this kind of art could be compared to humans' artistic expression?
It can, when humans act like machines, which they do most of the time. True humanity is about original creation; it is not about finding some pattern in Bach’s music and creating some sounds similar to it that can be heard in the elevator. These kinds of compositions have very low costs, and they are pleasant enough for most of the people, but I do not call that art, I mostly call that imitation. Maybe it goes a little bit beyond imitation, but that’s it.
To me, art is original creation, and how can we attribute it to the mood of a computer? It is simply an attribution, in the same way in Saudi Arabia the robot Sophia was recognized as conscious because it acted like a human being. We project consciousness on these mechanisms that don’t have any. We have to be careful that we don’t project something that we do feel into something that doesn’t have any feelings.
Yes, computers will be able to generate all kinds of pleasant art by proper training, by giving them sufficient examples, by getting them to understand styles of different artists and combining them in different ways. Maybe, thanks to computers we will be able to admire a Van-Rembrandt, or a Cezannes-Matisse painting. The result will be a combination of styles and the computer will do something even pleasant, but I don’t call that art.
by Valeria Sforzini - Young Voices Board