Algorithmic creation: Unbearable thinnes of flatness?

I recently read Christopher Steiners: Automate This: How Algorithms Came to Rule Our World (2012) which is a good overview of how computer algorithms have become commonplace in most unusual places. 

Steiner starts his journey with the 2010 flash crash in Wall Street, when Dow Jones suddenly lost over trillion dollars of it’s value in one day. Which is by the far the largest single loss ever recorded. Reason for this loss was a bug in one of the automated trading algorithms. Steiner then expands to the history of algorithms and how they took over stock trading and how they are now taking over many other areas, like entertainment industry, medical and health care industry. Steiner’s book is a good introduction on algorithms and their importance in our lives. The book got me thinking about algorithms in culture creation, where they are also already utilized.

For example in movie industry producers or film studios can feed a movie script to a computer algorithm and it will predict with good accuracy if it’ll be a hit or not. Same kind of algorithms have been used in music industry for some time. Then there’s algorithms that can produce soundtracks, so that movie producer can order song that for example sounds like Rolling Stones, or Mozart, but isn’t.  – It’s cheaper and quicker.

All this can sound exciting, magical or scary, or both at the same time.To start understanding computer algorithms it’s  good to know some of the basics in algorithm creation: At it’s core algorithm is just a set of instructions. Algorithm needs some input, then it compares it to the (already fed) database and based on the set of instructions programmed in it gives some outcome. For example algorithm can give you recommendations (output  for what to wear when going outside, based on the data it receives from temperature sensors, weather forecasts and the data it has of your wardrobe. 

Because computers are fast and can run through huge amounts of data in seconds they can be used to give outputs on much larger amount of data that any human can. This is the reason algorithms found their way to wall street. This also means they can be used to predict the likely outcomes of certain situations. Like the popularity of a pop song or hollywood movie. 

By feeding lots of data points of some music genres (or composer etc.) typical characteristics, algorithms can produce music that fits in that genre. Same works in some extent on any other artistic field, like visual arts, but music is simpler as it has widely used laws in notation and theory, which painting, for example has not. 

Algorithms can make life easier, and make industries work faster and produce more, but when algorithms start to march into cultural domains  we might be getting ourselves into trouble if we do not understand what we are doing. By trouble I mean something that Jaron Lanier described as “Unbearable thinnes of flatness” in his book “You are not a gadget.” By this he means that we are entering stagnant period in culture, where nothing new is made, we just recycle our existing cultural database again and again. 

One might argue that everything we ever create in culture is a remix. But I think in algorithmic creation we are entering a new kind of remix for one part simply because the remixing is done by a piece of silicon not by alive being. And this is the important thing to understand and is the main part of being code literate: Computers don’t do magic, they don’t think, they just follow orders. If the order is to run to a wall and crash, they will do it without any doubt or hesitation, because they haven’t got those. (One might code them to simulate them, but it’s simulation, which is just another snippet of code.)

Interesting part is that if we understand this we can start to use computer algorithms as artistic tools much in similar way we can use any other artistic instrument. This of course means that in odred to produce interesting art we have to know how to use, we have to master it. When we do we might get something new as an outcome, or it can lead us to something new. (Or not) 

For some time I have been thinking of building a drawing bot, that would do drawings on it’s own. I could code it to have some of the characteristics of my own drawing style, research other drawing styles, mix it up with other databases, so it could chage it’s style depending of the news in the world or weather or whatsoever. But where would the art be in that project? The most creative part might be in the code, not in the outcome. So it could be kind of boring project in the sense that would the drawing the bot produce have any value, would it have the aura of an art piece? (Another interesting take  of course is that if the viewer doesn’t know that the art piece is produced by robot, would the viewer give it more value? This is also something Steiner touches on (albeit lightly) in his book.)

Comments are closed.