[FRIAM] Schwill Rock?

[Jon Zingale]

Thanks for asking, Frank. On Friday, Glen mentioned that he was attempting
to train his music streaming service to identify *Schwill Rock*[0]. This is
amidst the ongoing discussions we have, on the server, regarding the
foundations of statistics[1], the use of AI in the legal system, and the
recurring realization that our suggestion engines are leading us into
deeper and deeper silos.

[0] The first question, the one about the Dwarves, is mostly asked toward
Glen, though the question is on the forum and so clearly open to whomever.
>From what I gather, Glen classified Schwill rock as music that rocks but
without apparent intellectual value. He gave two examples. In the Schwill
category we might have White Zombie:
https://www.youtube.com/watch?v=yPNFVj-pISU&ab_channel=WhiteZombieVEVO

and in the other, Tool:
https://www.youtube.com/watch?v=Y7JG63IuaWs&list=RDY7JG63IuaWs&start_radio=1&ab_channel=TOOLVEVO

which for those playing at home, invites analysis:
https://www.youtube.com/watch?v=uOHkeH2VaE0&ab_channel=Polyphonic

As for me, if I happen to (for some god-forsaken reason) find myself
listening to early Bob Dylan, I am very unlikely to want to listen to Joan
Baez or Woody Guthrie or Nina Simone's "folk" record next. Seriously
though, must *all* roads lead to Morrissey? I agree that it is possible
that someone somewhere does, but I sure as shit don't.

This leads me to wonder... if training AI on the current legal system
produces racist AI judges, then what hope do we have that (once those old
enough to have admired seeing live Dylan are finally gone) the rest of us
won't be stuck with *that* legacy "baked" into the AI engines? Let's face
it, it's not just the biosphere, look at that wasteland we still call
radio. (Thanks, Telecom act, wherever you are).

That said, I made some effort a few weeks ago (as well as on Friday) to get
a conversation started around what suggestion engines and search engines
could potentially do differently. One promising idea I have been
researching a lot lately is to adapt weighted ensemble methods (those used
for protein folding, say) to get better estimates on the all too
volatile energy landscape that is our individually fickle musical tastes.
Other ideas might include methods for rewarding actions in favor of
tolerance. But hey, just about anything would be better than the "geodesic
humping", "least-action loving" and "k-cluster fucked" nightmare we have
presently.

So there I am at Bosque Brewing and Spotify has gone off the rails,
everyone in the bar can tell, and no one is quite sure what can be done
about it. In some places I visit, the wait staff has already been defeated
by the domineering sounds of AI deejays and corporate policy. I can
empathize when I hear workers ask, "who would want to hear music on their
day off"?

Anyway, I wasn't sure about posting this response, but Grammarly assures me
that this post "may sound to readers": admiringly, appreciative, and
formal. All good things. mmm... spicy.

[1] Sorry, Nick, you missed your chance.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://redfish.com/pipermail/friam_redfish.com/attachments/20211011/3a93d10f/attachment.html>