Okay, Alexa gives me unparalleled access to music and can play a couple of cool trivia type games around bands. Spotify makes excellent recommendations every Monday and has turned me on to a LOT of cool bands I otherwise never would have found. However, I really feel like IBM’s Watson and I could be best buds, sitting around drinking beer and analyzing tunes for hours on end. I normally loathe infographics as tools for those with limited understanding to misinform those with limited attention spans (sorry, my profession is writing very large, very thorough analysis documents, and I enjoy it) but this one seems benign enough. I’m not sure what measurements and metrics a computer algorithm
I’m not sure what measurements and metrics a computer algorithm uses to infer the meanings of songs and assign emotional scores to those meanings, but I’ll keep searching. Case in point, Roy Orbison’s “Crying” can really only be objectively considered significantly Sadder than NIN’s “Hurt” if you measure it based on the tone of Roy’s voice (far sadder sounding than Trent’s), or the the fact that it repeats the “sad-related” word “crying” about a bajillion times, where Trent uses lesser-relatable words like pain, hurt, feel, kill only once each. Otherwise, bumping into your ex (as in Roy’s song) hardly seems on the same plane of sadness as stripping yourself down to your emotional core and not knowing if what’s left is even truly alive (as in Trent’s song).
And don’t get me started on Lovely Rita, a song about getting a parking ticket and realizing you have a homoerotic fetish for guys (or masculine women) in uniforms, being the happiest thing the Beatles ever put out?!