A Portrait of the Artist as a Young Industry

The work behind "AI" Art

This piece won't make attempts to solve the debates around what is 'art'? nor what is 'intelligence'? but is an attempt to honestly examine what is 'AI Art as is commonly being discussed (and used).

Machine Learning Models, not Artificial Intelligence

A history of religions suggests there is a tendency of the human mind to see patterns and extrapolate that an intelligent, independent force is behind it all. The mystery around the phrase "artificial intelligence" has added to this. From now on I will outline the process that leads to an algorithm being used to generate art.

Algorithms do a lot of work in modern times, some of these are strict sorting algorithms - as you read this, your browser has likely run a very simple algorithm to get you here. Chrome, which has for a long time been the dominant way people access the web [1], runs an algorithm for each site you go to, checking is there a https version of the site you are visiting available, and directs you to it [2], which we can consider a deployment of an algorithm.

More complex algorithms require a greater amount of involvement - to continue with Google, their famed 'pagerank' algorithm [3] was the dominant driver of what was seen on the internet for years.

Before expansion into email, data storage, digital office suites and mobile phones, this was Google's product. Yahoo was 'powered by Google' during a bried partnership, which terminated with Google denying a $3 Billion USD bid for the company and its algorithm.

Sources

[1] Browser Market share as recorded by StatCounter

[2] Blog Post from the Chromium developer team outlining the move of chrome browsers to https by default over http.

[3] How Google Finds Your Needle in the Web's Haystack from David Austin for the American Mathematical Society, which largely breaks down Bryan and Leise's 'THE $25,000,000,000 EIGENVECTOR' paper on the algorithm.