The shuffle function and depersonalisation

sufi

lala
I recently missed a talk on "Revolutionary Counter Optimisation Strategies" that sounded fantastic:
Optimization systems are everywhere: targeted advertising, personalized content feeds, location services, adaptive interfaces of software-as-a-service products, etc. In addition to their benefits, optimization systems can have negative economic, moral, social, and political effects on populations as well as their environments. Recent reports of neglect, unresponsiveness, and malevolence cast doubt on whether service providers are able to effectively reduce these effects on their own. Moreover, such negative outcomes are likely to predominantly impact the historically marginalized, minorities, and activists. In response, we propose Protective Optimization Technologies (POTs) that enable optimization subjects to autonomously or collectively defend against negative consequences of optimization systems without having to rely on service providers. We’ve seen many POTs in the wild. Uber drivers have been uniting to induce artificial price surges to counter the low wages caused by Uber optimizing for its own profit. Residents of neighbourhoods that suffer from heavy traffic routed there by Waze—a routing app that optimizes for convenience of its users at the cost of non-users—have fought back both with legislation and by reporting false accidents and traffic. AdNauseam is a browser add-on that automatically clicks on all web ads, effectively poisoning the behavioural profiling and data brokering systems—which optimize for click-through rates. As surveillance, control, and censorship increasingly become managed through optimization systems, one can imagine similar modes of resistance being part and parcel of the (digital) tool box of activists. In this session, we will introduce and discuss the negative effects of optimization systems, and provide examples of people resisting these. The rest of the session will be interactive and exploratory. During this second part, we want to gather accounts from those affected by these systems and discuss possible technological solutions to empower the affected individuals to counter such systems. By the end of the session, we hope to start conceptualizing realistic tools to fight back the discontents of optimization systems.
 

thirdform

pass the sick bucket
I must admit that the dehumanisation aspect appeals to me, (same as with the spine tinglers http://www.dissensus.com/showthread.php?t=14596 ) its a way of doing music that is antithetical to the whole elitist connoisseur approach & very undermining of it

Same tbh. if we ain't gonna have cockney geezers and geezerettes taking us to outer space then what's the point? noone wants to hear that capital fm presenting transplanted worldwide, actually capital extra presenting is still far more interactive than a lot of shows on nts, and nts has turned me onto a lot of stuff, but facts is facts...
 

sufi

lala
could be an interesting experiment. except the shuffle would pick from, as you say, selections that are based vaguely around my tastes. I'm talking about a DJ who doesn't know my tastes, and pulls in tracks/artists I either never imagined I'd like or just never heard of at all.
i suppose if we got hold of some mixes by algorithms we could do a blind comparison with mixes by human djs
 

Corpsey

bandz ahoy
UK garage is Spotify's Achilles heel

For imagine remix it brings up

Moving Too Fast
8 days a week (SFA SUNSHIP RMX)
Wookie - 2 us
SUNSHIP - I request
Misteeq - they'll never know
Jade - don't walk away

Then toddla t and m people oh dear

Like I said though Spotify hasn't got much garage
 

muser

Well-known member
Just recently got into the world of spotify when a climbing gym i go to started playing the entire discography of red hot chili peppers for hours on end and started driving me a bit nuts so I made a playlist for them. The algorithm for the shuffle is pretty impressive, goes a little further than grouping genres and I can only guess must use some data from what people have chosen to play before and after and links that with artists/songs in the playlist. I remember reading somewhere that winamp or itunes in the earlyish days of having the 'random' function purposefully made it less random because for some reason it being purely random made people think it wasn't actually that random.. still can't really get my head around that but it was explained with authority.
 

version

Well-known member
YouTube's algorithm is hammering me with this at the moment, seen it in related vids every day for the last three or four days.

 
Last edited:

Leo

Well-known member
YouTube's algorithm is hammering me with this at the moment, seen it in related vids every day for the last three or four days.

which begs the question: is it continually popping up because it's somehow related to other things you've watched, or simply because the artist or record label put some ad money behind it to MAKE it pop up in front of a lot of people?

few algorithms are "pure", platforms like YT, FB, Spotify, etc. make their money by selling visibility on our feeds.
 

luka

Well-known member
I've talked a bit about the sudden emergence of Midori Takada on everyone's YouTube suggestions a few years back before. That seems unlikely to be a corporate plot but it did lead to reissues of her records and the reignition of her career. Very odd. Very modern.
 

Corpsey

bandz ahoy
My own streaming conspiracy theory

When Michael Jackson's face appeared as the avatar for one of my "Discover" lists (plus loads of MJ choons) in the day or two following the broadcast of "Finding Neverland".
 

Leo

Well-known member
I've talked a bit about the sudden emergence of Midori Takada on everyone's YouTube suggestions a few years back before. That seems unlikely to be a corporate plot but it did lead to reissues of her records and the reignition of her career. Very odd. Very modern.

I've worked peripherally in the digital marketing world, the only reason platforms like YouTube and Facebook are "free" to use is because they transitioned early on from organic to paid posts. not 100% of course, but a sizable percentage. started out that a company would open its FB or YT account, upload/post content, and everyone who followed those companies would see it on their feeds or it would show up on organic searches. nowadays, companies can still post/upload content, but just a small percentage of people actually see it. in order to get the content into our feeds, companies have to "boost" the post with some spending.

just guessing but wouldn't be surprised if the record label seeded the market with paid boosts for the Takada reissue to generate awareness in the months leading up to announcement of the reissue. or could have been an organized PR campaign to get advanced content to influencers/tastemakers (select hipsters, journalists, etc) in order to generate a buzz. to the unaware outsider (aka, the average music fan), the whole things appears to be organic in that an artist somehow enters the zeitgeist and then a reissue appears four months later to "answer" the demand in interest...when in actuality it was all astroturf instead of grassroots demand.

not saying that definitely was the case with Takada, but wouldn't be surprised.
 

Corpsey

bandz ahoy
You know what probably influences our thinking these days, subliminally or otherwise, is the prominence of 'networks' in our social relations as well as technology.

I mean, most people are now are of everything being connected economically ('kevin bacon effect') - and on spotify, say, you have this chain of linkages now between - I dunno - fleetwood mac and justin bieber - there are a certain number of steps you can take via spotify or youtube to get from one to the other.

There's no sense of anything existing in isolation from anything else.
 

yyaldrin

in je ogen waait de wind
which begs the question: is it continually popping up because it's somehow related to other things you've watched, or simply because the artist or record label put some ad money behind it to MAKE it pop up in front of a lot of people?

few algorithms are "pure", platforms like YT, FB, Spotify, etc. make their money by selling visibility on our feeds.
i also got those takada and blue gas recommendations on youtube. what you say is of course true but i thought that maybe there are some songs that sort of apply to everyone's taste. songs that nobody consciously dislikes, and somehow i think that takada music fits into that category. i have no idea how these algorithms work but isn't it possible that some songs are build up of exactly the right elements for them to show up in everyone's recommendations?
 

Leo

Well-known member
i also got those takada and blue gas recommendations on youtube. what you say is of course true but i thought that maybe there are some songs that sort of apply to everyone's taste. songs that nobody consciously dislikes, and somehow i think that takada music fits into that category. i have no idea how these algorithms work but isn't it possible that some songs are build up of exactly the right elements for them to show up in everyone's recommendations?

technically it's possible and I'm sure it happens sometimes, but probably not that often. most online suggestions are somehow based on past viewing or interactions (liking a post, friending someone, etc.). watching some underground resistance videos will probably cause some Robert hood and Jeff mills videos to be suggested in the righthand column on youtube (since they both played in UR), but it'll probably also get you some suggestions on new electronic music artists who are currently being hyped by the record label through paid boosting.
 

jenks

thread death
You know what probably influences our thinking these days, subliminally or otherwise, is the prominence of 'networks' in our social relations as well as technology.

I mean, most people are now are of everything being connected economically ('kevin bacon effect') - and on spotify, say, you have this chain of linkages now between - I dunno - fleetwood mac and justin bieber - there are a certain number of steps you can take via spotify or youtube to get from one to the other.

There's no sense of anything existing in isolation from anything else.

That’s what I always think when it says ‘if you like this, you will like this as well...’ That constant sense of wanting things which are similar, if not the same. I use Apple Music and I’m often appalled by what they think I’ll like because I’ve played a particular artist - I wonder with supposedly with the world of music at its disposal why it’s offer is often so narrow. Which is probably what they want.
 

Leo

Well-known member
That’s what I always think when it says ‘if you like this, you will like this as well...’ That constant sense of wanting things which are similar, if not the same. I use Apple Music and I’m often appalled by what they think I’ll like because I’ve played a particular artist - I wonder with supposedly with the world of music at its disposal why it’s offer is often so narrow. Which is probably what they want.

like I mentioned upthread, this is the reason an algorithm will never replace a good DJ.
 

jenks

thread death
like I mentioned upthread, this is the reason an algorithm will never replace a good DJ.

Whilst true, what happens if they either don’t know a good DJ or if DJs themselves become seduced by the algorithm- rather than exercising their own tastes and predilections they try to guess what the audience wants, almost aping the algorithm? A kind of branding out as everything tends to the middle.
And if we listen to more and more music alone in our own headphone world - who is making those choices? Curated playlists and relying on the shuffle feature of iTunes or whatever music service we’re paying for. Are we abnegarpting control not to the ‘good DJ’ but to an algorithm and relying we won’t have to hit the fast forward button too often?
 
Top