As i have already told, simple beginning is to break the already written code into pieces and combine them in various ways, will give huge number of creative solutions to a problem. Surely, machines can create art much more fascinating than picasso himself. Just feed Robots with various strokes, styles, colors. The finer the variety and details, better the output. Robots should be taught tastes by separate codes, so that it is guided by the tastes and output comes accordingly. Nobody, even humans create anything out of nothing. They are always guided by their tastes. Supposing, i decide to draw something. I decide that my taste is something, like i want to draw a bird. Then through strokes, styles and colors and much more of my choice, i draw it. AI can draw your art in seconds, with innumerable varieties, u can pick. Robots should draw art, without human guidance. It will, if you ask them to pick up random styles, tastes, etc. Not everyone likes other's arts. Or you can give them particular taste, it will produce art of that taste, learn slowly by ratings, how much humans like a particular style and taste, just like how a child learns, or an artist learn art. Creativity is just part of AI. We will get into creativity much more in the later series.
Suppose, i wanna draw Monalisa. I will tell style as real life woman, who half smiles and looks happy and worried at the same time. If u ask an artist to paint such a picture, think about the cost and effort. But if you have coded very finely, in the sense, breaking face into finer parts, how a person's face changes and overlaps, to emotions. If you combine both happiness and sadness, you have created a marvelous art, at the click of a button.
Let's get into another scenario. I want to create a natural scenery. There are 4 items in that. Mountain, waterfall, plants and birds.
So A1, A2, A3, A4 are four types of mountain types. Each segment of code knows very well about it's type and it's relations with other code. Now B 1-4, is four types of waterfalls, C 1-4 are four types of plants, D1-4 are four types of birds. A1, B2, C4, D3 will give you a scenary, created from nowhere. If you have ever done coding, you will know one piece of code's output type will match exactly other piece's input type,in the sense, A,B,C,D should collaborate,allowing each other certain positions for location and ability to change each other, to a certain level. If one doesn't know other type nothing can be done. But coming to the issue, it is possible to draw picasso drawings, if first you draw 4 women thru women's code. Next mix it with next segment, whose job is to find the drawing of a women and slightly straighten it's curves. Then send it to a segment which fills the body with strokes. A1,A2,A4,A3( 4 women), B2( straighten it's curves), C4( fill the body with strokes). Break into pieces( if you already don't have), mix it with other segments. Art can be created by AI. Not necessarily photographs.
Ok. Let's leave supernatural things. Not only machines, but humans also do mix and match, with some experience and guidance of the past, while creating art. We can inject sophisticated codes to make machines choose certain types of art, based on the input. If u say, humans like certain art. They do it either because of their gene and lineancy, or from past experience. Anyways, we can make certain machines behave the same way, thru code lineancy towards certain art form.
AI is gonna dictate art industry in the future. Wroom.
