Cat B

And that had me think how a predictive model of an LLM will always lack the nuance of the immediate context. It may exceed us in the synthesis of the past, but the moment is owned by us. And we forfeit the richness of the moment and context when we are tied to consumption on a screen.

Not sure we can - but reading Questlove's "Hip hop is history" made me think about the conversation jazz has w past and present - musicians musically conversing on stage with each other as well as w the time and place of the composer.

“The real threat is if he convinces the rest of us to believe he has power he does not have.”
So appreciated this breakdown.

Your post had me curious about these recordings - i found an nytimes article from 2018 talking about how upsetting it would be to Gould to have these released but that they’re an interesting documentation of his process…but they got you angry so I’m gonna side w you and Gould.

alas what gravity has wrought is a continual RBF. i have a fondness for dust and watching it dance in a sun beam.

What a contextless ML/AI engine can spit out as its massive data pattern dump

ha - in the brief obsession w 3d modeling one of my favorite things were there were dust texture packs that could be applied to a surface that gave the image more real world feel.

cursory beauty
veneer wiped off with all the nonsensical

it’s funny how different imperfections are tells for real and fake

the differences between true world context and predictive conjures