I've had this dream of my ideal music player for a while, and one of the key features is it's "eclectic divergence" recommendations algorithm.
The idea is that it doesn't do what spotify and pandora do where they try to give you the same-y-est music they can find, or whatever most people who like it also like, but instead finds and studies more eclectic users.
So, for instance, if there is a user who generally has good taste regardless of genre, and they happen to like music that's all over the place in terms of genre, they would be marked internally as a "pioneer" or something. Their music tastes would then be used to create eclectic divergence.
But I don't think an eclectic divergence algorithm needs to stop there. It could be used for pretty much everything: movies, social media posts, memes, even (potentially) leadership roles.
What do you guys think of such an algorithm? Can you imagine making it? What kind of problems do you think would arise?
I just want to talk about this with someone because none of my friends RL are programmers.
I should also say, I think an algorithm like this might do wonders for the bubbles we've been putting ourselves in. I think one of the healthiest experiences a person can have is traveling because you can't help but see just how much every little culture is stuck in it's own bubble, and how much everyone is really the same in the end.
Evan Mitchell
Adding more trippy stuff in respect of the post that died for this one.
Isaac Perez
How do you determine good taste when you can't use popularity as a metric?
Cooper Reed
I dunno. My thought was good taste could be determined by non eclectic people per genre, but the genre branching would be based off of people who crossover that. Average users could still define what the best of some genre is, but their opinions on the broader topic of what genres complement each other wouldn't be as useful.
Connor Ross
Maybe I'm not understanding but it seems to me that if the normal popularity rating is correct and reliable and the purpose of this algorithm is showing people music (media,etc.) in different genres which may also be popular then all they really have to do is play a good rating song in this other genre.
Maybe you are talking about a discovery system where you like a song, the algorithm finds other users who like the song and then finds a song out of what they like regardless of rating.
Colton Thomas
short term user satisfaction is achieved with music already familiar what you suggest is useful occasionally, and thats sort of what spotify discover weekly does
David Thompson
While we're on this topic I would love for recommendation algorithms in general to have a "I like variety" option. I hate when youtube just recommends me more of the same. why would I watch what I just watched? Same with google supposedly tailoring results based on how you search. Stop putting me in a bubble!
Liam Morgan
I'm not really sure what you mean OP, could you explain it with an example please?
Aiden Brown
>the algorithm finds other users who like the song and then finds a song out of what they like regardless of rating.
This could definitely be one approach. I'm not particularly married to any one idea on how it's done, because I think there's more to it than what I'm imagining.
At the core, though, is the knowledge that current systems of scoring seem to really only work for people who all think the same. Not much weight is given to people who think differently because they are a minority.
Lucas Martin
how do you even measure if a suggestion was good, if the user comes back to a song, that's a metric, but if you are aiming for diversity of listens for users they should not really go back to a song i have worked on a music recommendation, we recommend based on repeated listen of a song in a given time period after the first listen of the user
Kayden Scott
Right!? I subscribe to one Republican-ish youtuber and suddenly my recommendations list is all politics from that side. I feel like there's got to be someone out there who's subscribed to both some really good republican channels, and some really good democrat ones, as well as a bunch of other stuff, and I want to see those things.
Btw, if anyone knows a really solid SJW or Feminist channel, I would appreciate seeing it. Right now all I have for them is hate, but I can't help but think it's because I'm not getting the whole truth.
Camden Taylor
that is a totally reasonable response of the system, most people subbing to that will click those recommendations, and thats the goal, if you are not found to be a target, you will be recommended something else that is basically their best bet for you showing you random vids is literally less sensible
Blake Parker
Yes that is accurate. Most systems don't want to offend or give the user a bad experience so they just show more of the same. I don't think it's really the systems fault Afterall if the user wants other stuff they should just type that in.
This is good but I am guessing you are assuming that the pool of music to choose from is already been curated and accepted by the user. I always thought it would make for a good shuffle algorithm which would look at total plays for a song and skip those that have a high number and play those that have a low number however this approach itself has its own issues with music played on shuffle and it's implications on the total play count.
Ryder Wood
I don't know for sure, but song rating systems can help. "More like this," "less like this," "more like this but not this same damn song." Beyond that, though, there are plenty of other metrics. I don't think the way it's done currently is bad by most users means, but it also feels like there are so many gems that go unnoticed.
Ryder Davis
John Oliver is well spoken and funny on that side of things. If anything it's entertainment.
But yeah, in terms of variety, I've just had to subscribe to a whole host of things depending on what I feel like listening to. They'll all sadly recommend me more of the same, but by subbing to tons of stuff, I force my own variety.
I do think, like someone else insinuated above, an algorithm could see how 'variety users' work and try to analyze those types of jumps in behavior. Or, idk, it would recommend something completely different, but 'known to be of decent quality', whatever that means.
Chase Thompson
It is very reasonable. I mean, it's lame that it overrode the things that had been there before (exurb1a, vsuace, smarter every day, real engineering, AvE, ...), but I get how it applies to most people. That being said, I still think it's a little overboard/heavy handed, but that's my subjective opinion.
In short, I get that money is king, I'm just dreaming user. Please don't wake me up yet.
Justin Harris
well, most people don't benefit from listening to tings that have low numbers thats just your hipster factor, i bet you would enjoy a song/video better just because you see lower numbers on the listen/view count
Matthew Flores
>John Oliver
Yeah, he's pretty good. Enjoy a lot of his bits. I don't feel like he's given me much insight into what legitimacy of SJWs exists, but it's better than a complete echo chamber.
>see how 'variety users' work and try to analyze those types of jumps
it seems really interesting to me. I would love to be able to turn on my music players recommendations list and get something that somehow meanders all over time and culture, but remains consistent.
Maybe it could cross analyze these jumps with how much non variety users like them to really hone it in.
Ryder Robinson
It's not so much the low numbers. But just in general, I do like finding things that are really good but overlooked, like you user. I'm sure you're great, but under rated.
Juan Johnson
having a little background in machine learning, i think youtube recommendation will not get much better, i think it is basically perfect personalization is already very high, and it is not easy to see the big picture when you just want to pick your next watch and you dislike all of your recommendations the system is maximizing watch time very well
Julian Lopez
this may be true, but judging by the bad taste in my mouth, I get the feeling it will eventually become something society no longer wants. Not for a very, very long time though. And, once again, just my opinion.
John Nelson
i'm just saying, you have a positive value on low popularity, many people have that, but all in all you mostly listen to popular stuff too, so your money is not where your mouth is
Justin Nguyen
Well it's with the assumption that the library is curated so it won't be as much "ugh why this" and more " gee I haven't listened to this in a while"
Angel Miller
fun thing is, the system will learn if its too elaborate recommendations are driving people away from watching, it will dumb the recommendation down to compensate for that personalization is here to stay
Xavier Ross
How about an algorithm that puts random songs from a playlist into your playlist you are currently listening to
Gabriel Adams
i don't think i understand you, what do you mean curated?
Levi Hughes
Like an old school music library not Spotify et.al You downloaded over the years heaps of music you like. You don't mind any of it but when you play on shuffle some songs are overrepresented and other songs are underrepresented
Jack Allen
ye, most people don't do that anymore you will meg much further if you have many users listening to many songs and trying to figure out what users are similar and what songs are similar, if you wanna talk recommendation just you listening to you hand picked albums does not have much room for imporvement
not sure how you would do that, or what the point is, whats on that other list, is it personal?
Eli Morales
Most of the time people want same-y music to what they're listening to because of mood. If you're relaxing listening to some lounge or jazz, you don't want to have death metal interrupt. It's perfectly reasonable that someone can have a good taste in both genres but what they listen to at any given time is mood driven.
And for the real eclectic users there's probably not a great deal of useful data you can extract from them. They'll give anything a try before dismissing it, so you'll end up with a very large number of tracks with few plays each. It's only this kind of user that might actually want such an algorithm in the first place. A normie wouldn't want something that plays random things that are unfamiliar to them.
Jackson Reyes
>you have a positive value on low popularity >but all in all you mostly listen to popular stuff too
These two can coexist. There is a middle ground, with plenty of stuff that would have been popular if it weren't for some form of chaos or another.
> your money is not where your mouth is
doubly so with the pirating.
Austin Long
>if you're relaxing listening to some lounge or jazz, you don't want to have death metal interrupt
totally true. but perhaps there are some death metals that are more jazzy or vice versa? I think that user might really like those.
Jacob Bailey
You could say it comes from a huge amount of different factors. Some that come into my mind:
-How rich the letter is. For example, comparing "time" from PF versus your typical electronic song. - The use of original melodies or reefs. - The use of typical rhythms (mainly in pop and edm songs). - Complexity of the song: number of instruments, number of people singing, different instruments, difficulty of what they are playing.
Of course in 2017 no algorithm would be able to judge that, but most stuff could be voted by users.
Jose Bailey
Here's my take on their situation, if you want to escape the echo chamber.
All sides just want society to be better. They have different approaches to this, and different ideas for how to do that. in the case of "SJWs", they are overly-sensitive towards other people's feelings, particularly groups of people who haven't been regarded highly in the past. In and of itself, it's not bad to desire empathy for other groups of people in society.
it becomes bad when - the problem is exaggerated - they expect everybody to be empathetic towards everything (unrealistic, not everyone has exposure to everything). Worse still, they try to be empathetic towards every situation without actually knowing it sometimes (hence accusations of appropriating culture, when the original culture doesn't care) - they refuse to listen to people who think differently, or think that those who do not understand them have no empathy/should hold their tongue.
It's noble to want everyone to feel accepted, often I feel these are people who've been mistreated or bullied in the past. However, the way a lot of these people go about doing this is not good at all. You want to bring people together? Then don't get angry, don't divide them, don't get outraged over everything. Instead, discuss, love, accept and live your life.
I don't know where you can find an outlet or source of SJW info, but that's my approach towards their behavior.
Josiah Myers
this is a pretty cool way to categorize music.
Hunter Wilson
So, from what I'm hearing from people, there's a lot of interesting factors and break cases. What seems like it might work so far is somewhere along these lines:
Sameness listeners can give a solid idea of what is a good center point for any given genre, I'll call it the "pop core." Pretty much everyone listens to multiple genres, and could create connections between various pop cores. But also, every user has some amount of divergence from the core, and that is probably related to what other genres they like. For instance, a person who likes both jazz and death metal probably likes a lot of different jazz from someone who listens to jazz and classical. They probably both enjoy a good amount of the pop core jazz classics though. In any case, they are divergent from it in different ways.
I think that using these divergent cases, an algorithm could kind of sniff in all sorts of directions from a pop core, and if one of them hits, follow it for a bit. If someone appears to like progressively more metal like jazz, at some point it could take a chance and leap into jazzy metal or something.
In any case, this has been fun to talk about. Thank you anons.
Kayden Morris
Definitely think this would be cool. And there are totally things you can easily pick out with an FFT, such as tonal complexity, rhythm, amount of distortion or bass (super easy), originality (expensive but also fairly easy), synthetic vs acoustic (subset of complexity focused on higher frequencies), and dancyness (usually a product of bass frequency rhythm).
My personal preference is towards soft, melodic, dance able music. It's been really good to me when I'm programming. I also like chillout.
by the way, does anybody know the best way to save a thread for oneself? I'm thinking either `File > Save Page As` or `File > Print`, but I've never done it before. I would like to keep this thread for myself though.
Lincoln Young
well, you can always just bookmark a link to the archive of this thread (i.e rbt.asia)
Austin Parker
thanks, stealing this idea
Levi Thomas
save page as is what i do
Oliver Perry
I grew up in a sort of hick town that grew a lot of wine, not far from San Francisco. It was a strange mix of hick, yuppie, and hippie. There were kids with confederate flag flasks, taking dips of chew at lunch, and hanging out with gay friends.
Point being, I definitely was raised around a good chunk of the points and understand them, but having moved across the country to Texas just moved me from one echo chamber to another. It used to be very easy to get into a good debate where someone proved to me a thing about Social Justice or the like. Nowadays I get really good arguments on why Trump that I had never imagined, but am missing the former thing.
It's really unfortunate that so few things are capable of tanking over such synthetic divisions. I wish an algorithm would.
Austin Sullivan
The idea is given freely.
Kevin Torres
Also, you're a selfish fag and I doubt you have the social intelligence to grasp the concepts required to actually make the thing. But who knows. If you can pull it off, I'd probably like you, so no worries. It solves itself.
Christopher Bennett
Not gonna happen, (((them))) are in charge of what people like, what's popular, what people should like, how they should evolve, etc. They do this by using psychology and media.
Not a chance they are going to let this be replaced by a "recommendations algorithm".
Example:
Men are followed by algorithm, men like this and that piece of media. Men have similar political beliefs as "eclectic" as their taste in music (great deviation from mainstream thought and taste). Goyim gets what Men like as a recommendation, this influences goyim and they end up watching videos or music, or even political podcasts that are not (((approved))) by (((them))). If this keeps on and spreads among the goyim, it's bad for (((business))).
tl;dr: jew wants you in the bubble they made for you, they'd shut this down
Jackson Brown
I think that current state of technology is caused by clever people that don't care about privacy and corporations still their best business and marketing ideas. They can provoke discussions on needed topics and gather ideas. Maybe even here.
Alexander Carter
I definitely see where you're coming from with this. One thing though, I think there's a big difference between rich zionists and jews.
Rich zionists were the very middleman minority which created the validity behind Hitlers convictions against them, but they ditched Germany far before the war ran Germany thin and the deportation of Jews became the extermination of them. In the end, the only Jews that Germany did exterminate were middle class and poor ones who had really didn't have much relation to the middleman minority.
Point being, Jew != Rich Zionist. Most of x are y != most of y are x.
Nathaniel Brooks
I'm thinking something similar. In my imagination, the finding of pop culture cores is only the first step of deep learning. It's next could very well be trying to escape from them, thereby threading a cohesion that is simultaneously globalist and less easy to control. Perhaps.
Jack Sanders
I like all sorts of music, from dubstep to 50's rock and roll. but when I'm listening to dubstep I don't want to hear any 50's shit, and vica versa your idea sucks and you should feel bad and kys
also pandora has the best suggestion algorithm, spotify sucks
Charles Adams
Clueless.
Aiden Lewis
>kys
no. I have a wonderful life with good friend, projects I take pride in, and lover after my own heart. Killing myself would be, unquestionably, a mistake at this moment.
Also, you missed the point. You probably like different forms of both of those genres than other people, but such nuisances are IMO under represented. Even in Pandoras algorithm. In any case, I recommend you read the entirety of this thread before your next point, because this first one has already been covered easily.
Eli Scott
>Rich zionists were the very (jews) minority which created the validity behind Hitlers convictions against them, >Clueless.
If you don't believe any validity to Hitlers convictions existed, I recommend you study some History. He didn't pick Jews out of a hat. But he also failed to recognize the distinction between Zionists and the far more common and agreeable person who just so happens to have some form of Jewish heritage or another.
Brandon Scott
Looks like a new Sup Forums project
Hudson Brown
>I have a wonderful life with good friend, projects I take pride in, and lover after my own heart.
i'm jelly
Jaxson Moore
Again, clueless. You should read actual history and stop spouting stupid bullshit like "rich zionists". Jews started a communist revolution in Germany which fucked up everything. That's all I'll say. Your turn to research more on the subject.
Jace Moore
what if we make it and do it in such way that only we can control this and change behavior of all normies to what we consider being ok
Mason Lee
The problem with "samey" recommendation algorithms is the classic exploration vs exploitation problem. Exploitative algorithms just pick the stuff that it most predicts you will like. So if you've liked a bunch of rock songs, rock songs are the safest bet, and it won't show anything else. Explorative algorithms try new stuff constantly. They value getting more information. This will initially seem to give worse and more random recommendations. But over time it will be exponentially better.
How to do this requires some understanding of bayesian statistics. But it's called Thompson sampling. First you build a bayesian model of the user and have a probability distribution of the uncertainty of each parameter. Let's say the "likes_rock" parameter is 20% likely to be true based on your previous data. Then, with 20% probablity it will show you a rock song and see if you like it or not. Over time it will learn with more certainty what your preferences are, and stop showing you genres you are very unlikely to like.
Matthew Collins
Can't you always look at it with hindsight? Say a certain song gets popular; just look at the early adopters.
Gabriel Rodriguez
When I search for "psytrance" (that's a music genre of which I own quite a few tracks of) in iTunes 12, nothing shows up anymore. This is the case since Apple Music exists and was never a problem before the "smart" (dumb) search field existed.
Another thing that triggers me, that I have to press enter when searching for tracks in iTunes 12.
Easton Gray
Well, if it's any consolation, I had a very, very bad life for a long time prior. Carry a coin.
Jayden Cruz
I think it would make much more sense to let them find themselves than to try to force ourselves in. Normies are simply interesting people who've failed to gestate.
>divergence Good idea, but you really should study behaviour of the 80% retards out there >the patrician taste - all over the place Good idea >the absolute dogshit taste and people who dont give a fuck - also all ovet the place, even more so. How would you factor them out?
If anything you should introduce a way to factor the plebeian taste out of predictions
Matthew Russell
This is incredibly interesting. I think the potential for improvement exists in the zone of showing songs one doesn't like. I labeled it as "sniffing" above, but if my assumption that there exists a continuous (though likely not fully populated) spectrum between any two genres, it could choose to sniff in that direction rather than go all the way. In doing this it could (potentially) keep close enough to whatever pop core to be passable, while only really moving down the roads that get an observable positive response.
Parker Miller
At the moment, the concept I'm describing a concept I think might be effective as "sniffing"
See posts:
Jordan Sullivan
Very cool, I tuned in. I'm more of a stream guy as well, but "the bug" is still annoying.
Adrian Ortiz
It's a pretty alright couple of stations. I can put them on and not feel the need to change them all night sometimes.
As for that "bug," the link following does cost money, but it was some of the best money I've spent on an app in a long time:
An algorithm that picks songs that 1. you haven't heard of. 2. everyone but you likes.
Trivially easy and no one would want it, that's what you want, right?
Aiden Thompson
Do it and include it on newpipe
Lincoln Reed
That is very nice of you, thanks.
Swinsian looks nice, but I'm more of a VLC guy now. Let's me even change the playback speed, a feature every vinyl deck has. It's still bugged in the library section at the moment (what a surprise...), but I'm sure it's going to be fixed sometime.
I would love to see a music player with youtube-dl built in. Click a button next to e.g. the artists name and get some kind of popup (or whatever) showing you the stuff available with the option to download that instantly and have it automatically added to your music library.
Yeah, I'm aware the dev would receive death threats, kek.
Well, as you are so kind and share even more stations, here's another extremely progressive one: dragonflyradio.net
The rest of them is too sacred for a shithole like this :ˆ)
Anthony Russell
No. I'm thinking more like a really good DJ. Not so much in the mixing aspect, but the finding good songs and making a cohesive order from them. I'm not sure if you've ever Dj'ed before, specifically in a lax enough environment to be experimental with it, but it's much more involved than one would think. A lot goes into trying to keep it fresh, and this is the very quality that modern algorithms seem to lack.
Anyways, modify #2 to be >many people statistically similar to you like and it's pretty close.
Jayden Sullivan
Really appreciate the link.
>I would love to see a music player with youtube-dl built in
I was thinking something similar: Built in torrent system. That, and the choice to donate a couple bucks a month and have it dived out however you like (plays x rating by default). Basically a hybrid between spotify, pandora, torrents, and patreon.
>Yeah, I'm aware the dev would receive death threats, kek.
And thus, the reason I will never attempt such a thing. But damn.. it's still nice to dream. Best enjoy this while it's still an option, I want to be able to look the next generation in the eye when I say I took what was left to be taken.
Alexander Lee
>Built in torrent system That's a neat idea actually.
Let's see what IPFS and derivates will bring. I heard there are still people using stuff like limewire, so it has a chance of existing, that's for sure.
Jaxon Adams
Okay going off of this idea, there might be a simple-ish way to get the 'pioneers' or at the very least, the people that tend to have more varied tastes.
If you perform the K-Means Clustering algorithm on a dataset with K being an approximation of how many genres you think there are (say about 7-10), and you let it run until it converges, you are left with clusters of people with very similar taste.
However, with this clustering, there will be people that are rather far from the center of each cluster, which means they are close to bordering another cluster. I'm just spitballing here, but let me know how good this idea may be
Nathan Flores
Actually, now that I think about it, it wouldn't be easy because the user would just treat the like button as a dislike button and vice versa. But okay I get what you're saying. As a side note, do you know if there's a program that allows you to pick out songs for each member of a focus group and influence what other people listen to? Something that can be gamed.
Nolan Cook
>still people using stuff like limewire Oh man. Napster, Kazzaa, Limewire. I hadn't membered in a long time.
>IPFS and derivates ipfs.io/ Cool stuff. Hadn't heard of it before. I think the people over at NetRunner might be interested in this.
I don't know much about K-Means Clustering, but if you have to guess the amount of genres up front I think it might be problematic. Then again, if the algorithm can be modified to be less dependent on that number, clustering absolutely sounds in line. It even seems like the notion of "pop cores" I was thinking of earlier
In any case, please do go on. I can't say much about the algorithm you've mentioned, but surely someone can.
Christopher Gonzalez
Surely Pandora and Spotify can both be gamed with enough effort and input. But I think Spotify might be significantly easier if it uses user made playlists as a metric for it's suggestions. I really don't know though how either works except to say "not quite well enough."
James Phillips
where can i learn to program all these visualizations?
Jackson James
This. As much as it sucks to feel stuck in a glut of similar content, this kind of focusing is really effective for driving clicks. Frankly I don't think there's a large enough interest for youtube itself to implement this. That said, I would totally get an extension that did this.
Daniel Hall
They are produced with a myriad of tools, but there's no need to get ahead of oneself.
>he thinks kys is a literal suggestion I'm glad you have a life worth living but honestly
David Howard
There are clustering algorithms that aren't k-means that don't take the number of clusters as a parameter.
I agree with the general approach though: >cluster the content, extract genres, and give a weight to each (content, genre) pair representing how likely the content belongs to that genre >From there look at each user's ranking history to determine which genres they most prefer
At this point you'd have enough to make a pretty typical recommendation engine, but if you want to go for divergence we can do more. I'll make a new post for that
Noah Roberts
i already know the canvas api though, i assume in order to make these animations i need some kind of math book or something
Jonathan Thompson
If you've got your (content, genre) and (user, genre) weights then you can go a step further and cluster the users into batches.
I think divergence could be achieved by interpolating between a user and the centroid of a genre they're unlikely to be in, and giving the content most liked by other users close to the first user along that direction
Or you could do as OP originally suggested and search for users with higher than average spread between genres and give more weight to content they enjoy
This seems to be a pretty standard data science problem and afaik the best way to solve it is to pull up a data set and see what works
Nolan Anderson
Not just that, I think it's often times good advice and a good decision.
Sounds good. I'm at my nights end atm, and I'll probably completely miss the next post, but this idea is "Worth Spreading." Perhaps we can do a bad TEDx talk on it. With hookers and crack. Anyways, I need sleep. Have fun.
These are some of the best of the best, and you can't hope to compete with them out the gate. But that's not the point anyways. The point is to just play around and make things you like. A tip: If you really want to get complex looking results, use recursive functions. But also, just enjoy yourself. If it looks trippy, that's good enough. If this is really something you wish to do, time will tell. Also, pic related, you could definitely do something like that.
Blake Lopez
It would be useful, I would be able to listen to music a lot easier. Can imagine how much of a nightmare it would be to implement something like that though.
Brayden Flores
Quartz Composer or Processing would be my recommendations.
Ayden Price
Thanks for sharing, much appreciated. Great thread.
I had similar ideas but a bit afraid to post them here. As I sad, companies that shill here could as well have special shills for provoking discussions on needed topics and collect ideas. We first need to figure out how to deal with them.
Brody Gutierrez
Thanks! I'll check it out.
Ryan Ross
While I find your paranoia completely understandable, even if direct at me (though it would be a pity of divisiveness), I ultimately believe there is neither a need nor a point in trying to fight (((them))). The reason I believe this is the "Horseshoe effect" or the "Paradoxical Origin of Thought."
Basically, the horse shoe effect says something along the lines of any two opposites pushed far enough become the same. Another way to think of it (that I think of it) is that a totalitarian government at it's strongest state is the same as a utopian one. The biggest pit fall of (((them))) atm, IMO, is that they spoil their own goods, and don't recognize the sort of inevitable cosmic barter between a world full of interesting people, and truly diverse and interesting life experience.
As for the paradox bit, it's basically just the notion that if you want to get something done, do the opposite. If you want to be cool, quit trying to be cool. If you want to fall in love, quit searching for it everywhere. If you want to destroy the gov, help them become their strongest selves. This is not something I've met anyone else who believes, but in my experience, our rational logical brains just aren't very good at getting things done. In fact, as far as I can tell, our brains exist about 80%-90% for making up excuses.
Anyways, that's my take. I don't believe they will be defeated head to head, but I do believe that they can be led by greed to powers far greater than fear and oppression. If you, reading this, happen to be one of (((them))), I highly recommend you take this opportunity to become even more powerful. Far greater powers exist than those created by lies. And beyond that, the only value of power is the life you live, which is just as much a product of power as it is environment. A world full of gormless nimrods is not the setting of a lesser king.
Ryan Barnes
>A world full of gormless nimrods is not the setting of a lesser king.
typo. Meant to say it is the setting of a lesser king. Easy to see too. Just imagine two kings, one has a land full of powerful, intelligent, virtuous citizens that both love and fear them for who they really are. The other has a land full of literal morons that are fat and depressing and argue about everything, and nobody really knows who the king is, and pretty much every reason they have for every opinion they have is the product of some fabrication or another.
I don't think this is a matter of good and evil, I don't believe in either. I think selfishness works great, and the only issue is when people over think it and try to define happiness by more complex rule set than pleasurable, safe (like mother and child, not walls), freely given, and diverse.
>The other has a land full of literal morons that are fat and depressing and argue about everything, and nobody really knows who the king is, and pretty much every reason they have for every opinion they have is the product of some fabrication or another. Kek. Perfect description of Sup Forums
Jace Hall
>tfw Grooveshark got killed so hard that anons are trying to reinvent it while thinking it's a completely novel idea.
