To put a finer point on it: if I put David Bowie in and set the date range to the 90s I should get Nine Inch Nails or Marilyn Manson, not Blue Oyster Cult or Talking Heads. If I set the date range to 2000's I should get St. Vincent perhaps. If I put Led Zepplin in and set the date to the 90s I should get Tool. For the 2000's perhaps Jack White. For Depeche Mode set to the 2010's I should get The Black Queen. Bob Dylan or Leonard Cohen transposed should include Nick Cave. Who is the 2010's version of Nick Cave? I'd really like to find out. I would be thrilled to have a machine that would help me find out. It's what I would use exclusively. I'm revealing my tastes here but you get the idea. I want to find the threads that exist between eras and genres, not within them.
... View more
I am desperately seeking new music (as in, music created within the last 5 years) that carries the temperamental torch of the music I like from previous generations. This means the personality characteristics of the music, not the transitory fashion of the era. I want to be able to input the vast list of artists I love from past eras into a pool to be assessed for personality characteristics (not so much genre affiliation) and click on a filter to specify how many years old the music that comes up is. This means that if I put in Nine Inch Nails, I'm not getting Marilyn Manson and Gravity Kills. I want the newest permutation of the temperamental characteristics of (to keep with the Nine Inch Nails example) asperity, dissonance, complexity, iconoclasm, intelligence/neuroticism, anti-pop with pop elements, anti-PC, etc. transposed into its modern expression, as opposed to the most fickle of radio genre relationships, that of 90's industrial. Maybe it will deliver the closest thing to a Trent Reznor of Jazz or hip hop, or at least the closest modern thing. Don't we have the artificial intelligence algorithms to be able to analyze on psychological personality traits, such as the big 5 temperamental scale, for example? Isn't that the primary power of artificial intelligence, to find associations that aren't simple and obvious? We clearly don't need complex algorithms to filter by release date. That should be dead simple. I have no idea why that isn't a basic feature in every music app. For me, the whole goal of an associative music suggestion bot is to help me discover novel, fresh music that resonates with my personality in a technological paradigm that has detached music from consensus culture and scattered it across an increasingly atomized and rarefied field. How is it that the output of these algorithms is still so simple and useless? I've listed 20 or 30 unique artists. The personality data there to be potentially analyzed to find the meaningful common threads to generate a smart filter should be close enough to infinite. Add a release date slider and I'd be listening to this all the time. Not like now. Only once in a very great while will I hear something new and exciting.
... View more