Re:An experiment in crowd-sourced songwriting
by SwellJoe on Thursday September 01, 2016 @01:05PM
(#12406990)
The majority of the people are completely lacking in taste the majority of the time.
Also, the majority of people have no musical training; not even a few hours tinkering with a music app to have a vague understanding of what makes a song meaningful; basic stuff like tension and resolution and song structure. This is effectively a random selection of notes within the key, and I bet "no note" will never win the vote, so there will be no interesting rhythmic variation. A simple script would work as well as these 65,000 people.
That said, there probably are interesting (though questionable in their artistic merit) ways to crowd source an artistic work. Focus groups for films and TV shows and such have been happening for a couple decades. There have been interesting surveys of popular harmonies and melodies, which could be applied to scientifically build songs that fit what is popular (and I'm sure it has been done). The uncanny valley is probably where you'd end up in most cases, though; at least for now.
Some interesting work in the area of understanding what crowds like:
Also, the majority of people have no musical training; not even a few hours tinkering with a music app to have a vague understanding of what makes a song meaningful; basic stuff like tension and resolution and song structure. This is effectively a random selection of notes within the key, and I bet "no note" will never win the vote, so there will be no interesting rhythmic variation. A simple script would work as well as these 65,000 people.
That said, there probably are interesting (though questionable in their artistic merit) ways to crowd source an artistic work. Focus groups for films and TV shows and such have been happening for a couple decades. There have been interesting surveys of popular harmonies and melodies, which could be applied to scientifically build songs that fit what is popular (and I'm sure it has been done). The uncanny valley is probably where you'd end up in most cases, though; at least for now.
Some interesting work in the area of understanding what crowds like:
http://www.hooktheory.com/blog/i-analyzed-the-chords-of-1300...
http://articles.latimes.com/2012/feb/03/science/la-sci-hit-s...
https://en.wikipedia.org/wiki/Hit_Song_Science