Is the AI ​​capable of producing art and expressing creativity in its works?

Is the AI ​​capable of producing art and expressing creativity in its works?
Spread the love

Artificial Intelligence increasingly adds more achievements and improvements, but creativity and the creation of artistic works has always been an exclusive role of human beings. So, can you develop the painting machines in an original way and awaken our interest?

We have always thought that creativity and art belonged only to human beings, but now Artificial Intelligence puts this thought in check. The creation of numerous works through learning processes of different artistic techniques leads us to ask ourselves if it is possible for AI to produce art and to develop in machines the same creativity that characterizes humans.

The first works of rock art that our ancestors made are about 73,000 years old, and now it is the turn of the machines. Since 1960 there have been numerous studies to determine if a computer could produce art independently.

For this reason, the first drawing programs were created for machines based on codes, such as those of the Hungarian Vera Molnár, or the creation in the 80s of simulators that evolved organic forms animated without being directed by a human, which created the British artist William Latham.


Creativity in the AI currently

At present there are numerous cases in which the Artificial Intelligence generates artistic contents. For example, Spotify hired an AI researcher named François Pachet to work on AI software that wrote music.

In Artnet , a website in which works of art are sold, a Parisian art collector bought an image made by AI, that is, using a computer program by Obvious, a collective of French artists who defend this art under the motto ‘creativity is not just for humans.’

This software that created the works of art did so using other historical paintings as references, and although it is a completely new painting, it resembles an 18th century image.

This work of art has been generated thanks to the use of ‘adverse generative networks’ (GAN). These GANs are neural networks that teach themselves through their own experimentation, without the need for a human being to program them.

That is, they make use of two programs: one that contains the database of something real (old works of art as reference), and another that seeks to ‘cheat’ the first program trying to constitute an example with what ‘knows’ but that be something new. Thus, the second program creates invented images, but with a ‘style’ of those that have been entered into the database.

So, will they never be works that are not based on preexisting art?

In the end what makes a work created by a human being is creative has a lot to do with the intentionality of the artist in carrying out that work. Your thoughts, passions, hopes or ideas, is what characterizes an original work. That is, in most cases, humans create works by a need for self-expression.

But is this the only valid creativity? Maybe machines also have their own form of expression, generated from their own needs.

As the first humans painted their hands on the stones in the first expression of their existence to themselves, it is possible that the machines reach that creativity in the future and perhaps their originality will be compared to that of humans in all areas.

Leave a Reply

Your email address will not be published. Required fields are marked *