(ANTIMEDIA) — This year Anti-Media has been tracking the progression of artificial intelligence (AI) as an economic force and, more specifically, the development of algorithmic AI as an artistic movement. This year alone, AI artists have been prolific, producing short stories, movies, paintings, and musical albums. The latest offering is a black metal album from the mind of an artificial neural network.
The five-track album, entitled Coditany of Timeness, marks a genre first for AI and, according to the project’s creators, Zack Zukowski and CJ Carr, known as the Dadabots, should be considered a significant achievement in the field. Unlike previous AI music that derived from classical music, this album was charged with recreating black metal.
“Characterized by its ultra-long progressive sections, textural rhythms, deep screams, and melodic weaving over a grid of steady, aggressive rhythmic attacks,” say Zukowski and Carr, “[black metal has] extreme characteristics [that] make it an outlier in human music.”
They created the music by feeding 3,200 eight-second segments of raw audio data from the album Diotima by Krallice into a neural network.
Neural networks work by gradually improving machine performance through example analysis. In this case, after 5 million guesses in 3 days, the Dadabots’ AI created a full album of algorithmically generated black metal music. The titles for the songs, the album title, and even the cover artwork were all algorithmically generated by the AI using a probability equation known as a Markov Chain.
Jon Christian, author of The Outline magazine, which interviewed Zukowski and Carr, remarked, “If I didn’t know it was generated by an algorithm, I’m not sure I’d be able to tell the difference.”
Before the year’s end, Anti-Media will be releasing a Best of AI Art special, replete with superlative artwork created by artificial intelligence in 2017. Until then, take a break from your day to listen to some true black metal machine noise.