Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

[P]
The Technology Behind The Emergent Music Web Site

By garyrob in Technology
Thu Apr 18, 2002 at 05:20:54 AM EST
Tags: Technology (all tags)
Technology

Emergent Music is a Web site for the purpose of identifying the best undiscovered music. A key problem it faces is motivating people to do the work involved in finding that music, such as inputting ACCURATE ratings, reliable recommendations of music, etc.

[editor's note, by rusty] Full disclosure: The author is CEO of Emergent Music, and has bought many textads here in the past.


ADVERTISEMENT
Sponsor: rusty
This space intentionally left blank
...because it's waiting for your ad. So why are you still reading this? Come on, get going. Read the story, and then get an ad. Alright stop it. I'm not going to say anything else. Now you're just being silly. STOP LOOKING AT ME! I'm done!
comments (24)
active | buy ad
ADVERTISEMENT
Some sites don't need to provide extra motivation for these kinds of things because people do them for their own reasons, and do them appropriately and meaningfully. Kuro5hin is an example of this.

But there is so much music that has to be waded through to find the good stuff, and it is so much work to do it, that we felt from the outset that it's really too much work for people to do without some kind of benefit to them. In fact, that's why it really isn't being done yet in a scalable way on the Internet (IMHO).

Now, some ratings sites provide rewards to all raters. There was a site a few years ago, for example, that sought to provide motivation for rating discussion comments by offering a tiny number of frequent flyer miles in return for each rating.

The problem there is that the motivational system is all wrong. It doesn't motivate people to issue meaningful ratings. In such systems people get the most reward per unit of effort by paying no attention to the content to be rated at all, and just entering as many ratings as possible in the shortest possible time.

Obviously, that would not be a very good way to motivate people to do the work of determining what songs are better than others (and what recommendations of those songs are better-written than others). For example, doing that work well may require multiple listenings to the same song in many cases (since many of the best songs do not reveal their greatness in one listening).

So we needed a way to determine who is doing a good job and who isn't.

There are two parts. First we need to determine whose ratings are meaningful and whose aren't. Then we can use those meaningful ratings to determine which recommendation authors are good at that and which aren't.

The way we determine whose ratings are meaningful is to measure each rater's tendency to predict future ratings for the rated items. That is, for each rating, we ask the question "Did it meaningfully predict future ratings?"

That involves some calculations based on Bayesian statistics. The exact mathematical equations are, for now, are a trade secret. We haven't decided how public we should be with respect to those details. But essentially, the way to get a reputation for being a good rater is to rate early and predictively. That is, once we have enough ratings for an item that the community opinion for the item is pretty well established, you have lost most of your opportunity to enhance your reputation for predictivity by rating that item.

Note that this technique wouldn't work if we were asking people to rate mainstream music, because the community opinion for a work of mainstream, mass-market music is usually pretty well-known by interested community members quite quickly through the mass media and word-of-mouth.

But Emergent Music is focused solely on undiscovered work. So early raters don't have an external way of knowing if the rest of the community will agree with them. They have to rate based on their own opinions. Which means they have to actually make the effort to listen.

Hopefully, I have explained things well enough that you can see that we identify people who are good indicators of what music the larger community will enjoy. So we can now reward them for that good work, while not rewarding those who aren't good at it or serious about it. We award them with "EM Points" which can be used to publicize artists that the raters care about and to bid for free review copies of CD's. In the future other benefits may emerge.

It is my personal belief that this provides a practical mechanism on the Internet for separating the wheat from the chaff on a very, very large scale because the work is done in a very distributed way, and I am aware of no other existing and viable solution for that. Since it is also my belief that such a mechanism is the final step needed to enable a shift in the music industry away from least-common-denominator, mass-market musicc and toward music that addresses people's individual tastes, I am very excited about it. So, I am very excited about EM.

Of course there is the question, "How does predicting community opinion enable serving individual tastes??" Ah! Glad you asked. ;)

The current version of EM is geared toward finding undiscovered artists with appeal to most people who care about a given genre but have been unable to break through (maybe, for instance, they aren't photogenic enough to make it through the screening folks at the major labels). This obviously has its own value -- even the Beatles were unable to break into the recording industry for two years and were just at the point of giving up when George Martin signed them.

But in the version we plan to release in a few weeks, genre creation will be by community input according to an evolutionary technique that, in the current EM, is used to enable many people to contribute improvements to a single music recommendation. We expect the genres that evolve to be very dynamic and truly grow out of the needs of the community, and to be, in fact, micro-genres representing the tastes of very small, highly-focused communities of people with a great deal in common musically.

So, these micro-genres will comprise a new form of collaborative filtering. Note that we also have a more traditional collaborative filtering engine wating in the wings (already coded) which is based on ratings, as is done on Amazon.com. Since we have very reliable ratings, we expect to also be able to generate very reliable individualized recommendations. It will be interesting to see whether the micro-genre approach completely obviates this more traditional approach or compliments it. We don't know yet.

Sponsors

Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure

Login

Related Links
o Kuro5hin
o Emergent Music
o Also by garyrob


Display: Sort:
The Technology Behind The Emergent Music Web Site | 73 comments (53 topical, 20 editorial, 0 hidden)
best way to measure music (4.50 / 6) (#6)
by speek on Wed Apr 17, 2002 at 03:36:57 PM EST

Why not make a plugin to audio players that signals your site that song X is being listened to? That way, your site can simply track how many times people choose to listen to a certain song. That's got to be the best way to determine how well liked a song is.

Anyone willing to go to your site and rate songs should be willing to go to your site and download the plugin and configure it. From that point on, they're done. It's not so different from that thing that goes on the internet to figure out the names of the songs of you current cd, even though you never told it what cd you put in? Ought to be fairly straightforward to do....

--
al queda is kicking themsleves for not knowing about the levees

Definitely (none / 0) (#11)
by garyrob on Wed Apr 17, 2002 at 04:00:52 PM EST

yes, early on we talked about that a lot. Soon may be time to do it. Obviously there's been a lot on our plate... at this point we're focusing on simplifying the UI -- something that people are making it clear we REALLY need to do -- and some other things like adding an RSS feed containing the best songs as "payloads"...

[ Parent ]
Ad (2.00 / 5) (#7)
by krek on Wed Apr 17, 2002 at 03:51:29 PM EST

Are you not supposed to pay a couple of dollars to get this sort of thing into the corner of the front page?

We did (4.50 / 2) (#8)
by mgoyer on Wed Apr 17, 2002 at 03:54:43 PM EST

We've bought a bunch of ads and in the discussion of our most recent one someone suggested posting something as an article..

Matt

[ Parent ]

ads shouldn't influence content (4.66 / 3) (#10)
by Delirium on Wed Apr 17, 2002 at 03:57:46 PM EST

If this would've been a legitimate article before ads existed, it should still be a legitimate article now. The mere fact that ads exist should have no bearing at all on the stories, or else they're failing half their purpose (which is to provide k5 a revenue source without damaging the community).

Not to mention that this isn't a topical comment...

[ Parent ]

The point... (none / 0) (#46)
by brainrain on Thu Apr 18, 2002 at 08:30:17 AM EST

I think you're missing the point that Rusty was trying to make "[editor's note, by rusty] Full disclosure: The author is CEO of Emergent Music, and has bought many textads here in the past." He is simply stating that this user, while we may have no idea who he is, is, in fact, a participating, contributing member of the Kuro5hin community. I found this statement to be very important, and crucial, to the validity of the argument that was presented in the article. If Rusty had not put that statement in then I would have had to assume that it was simply someone reviewing the web site, or someone who was close to the operations of it. However, seeing the fact that this technical article was written by the CEO of the company described brings that much more weight to the strength of its argument.

--
Kleptotherapy - Helping those who help themselves
[ Parent ]
I saw the ad (4.00 / 2) (#23)
by wiredog on Wed Apr 17, 2002 at 04:25:45 PM EST

And suggested he submit an article about what he was doing, because I thought it interesting. Not everybody is going to see the ad, after all. Or check out the conversation attached to the ad.

Peoples Front To Reunite Gondwanaland: "Stop the Laurasian Separatist Movement!"
[ Parent ]
Better living through technology (4.33 / 3) (#9)
by mogador on Wed Apr 17, 2002 at 03:56:10 PM EST

I really enjoy the two threads running though this article. One highlighting the need for a reliable source of music distribution that exists outside of traditional structures and another, the harnessing of both technological and human strengths to achieve the desired effect. It seems that some comments are more concerned with grammar than the wide ranging implications of the scenario this article describes, a scenario that has some very exciting implications for the future.

What is motivating you to create this site? (3.00 / 4) (#18)
by yankeehack on Wed Apr 17, 2002 at 04:16:57 PM EST

I'm wondering about this, only because the type of "community" you develop is in fact dependent upon your interests.

I assume that besides the quest for better music, you probably want to make a little money off of this. Are you looking to go to "industry" events? Or to be written up in the next issue of Spin/Rolling Stone/Vibe/etc? Who are you going to charge--the submitters or the subscribers?

Fuck this; I'm goin' drinkin'. Which, in the US, is a "right-wing" activity, unless what you're drinking is more than $20 a bottle--then it's "left

Who we're going to charge (4.00 / 1) (#20)
by garyrob on Wed Apr 17, 2002 at 04:21:56 PM EST

As the site exists today people can go to Amazon and buy an album of an artist you see on the site and we get 5%. (We need to see if we can work that out with CDBaby too.)

We also allow people to "sponsor" artists, so that will hopefully be a revenue stream as traffic on the site builds. BTW, by "sponsoring" we don't mean a tipjar for artists but a notice that makes it more likely that they'll get reviewed, rated, etc. It's akin to a text ad on Kuro5hin except people bid for the top spots.

Basic use of the site is free for everyone and we expect it to remain so, although at some point we may separate some special features into a members-only, paid-for area.



[ Parent ]

facilitating subgroups of listeners (4.57 / 7) (#19)
by Delirium on Wed Apr 17, 2002 at 04:19:37 PM EST

I'm not entirely sure from your article whether you already have this implemented or are planning to, but it seems indispensable for a site like this to facilitate getting recommendations from subgroups of listeners with similar tastes. I'm not sure of the specifics, but some sort of system where people whose ratings I've liked in the past have more influence on the scores I see in the future. That way if I absolutely can't stand indie rock, I never see highly-rated indie rock. And if someone with whom I agree 90% of the time rates a song highly, I should see it near the top of the list. I think this has a better chance for being an accurate predictor of what I'm going to like than a general average of all users does, since subgroups can often corrolate their tastes with much better accuracy than the overall group can.

It's there (3.00 / 1) (#24)
by garyrob on Wed Apr 17, 2002 at 04:39:02 PM EST

Yes we have implemented the site, click here if you want to visit it.

The article I've posted tries to answer exactly the issues you're raising in your comment... right now ratings are localized to standard genres but soon they will be localized to highly specific and evolved micro-genres addressing particular tastes. We also have a more traditional collaborative filtering engine waiting in the wings that does more exactly what you suggest, but we're not sure it will add anything useful over the micro-genres....

[ Parent ]

yeah (none / 0) (#30)
by Delirium on Wed Apr 17, 2002 at 07:07:22 PM EST

I visited the site before posting; I suppose the collaborative filtering is what I was asking about. But you may be right, the micro-genres (especially if they evolve well) may make collaborative filtering unnecessary. Depending on implementation they could conceivably be better also, as it would be possible for me to be a fan of several micro-genres simultaneously (while with collaborative filtering you'd end up with noise if for example I love 100% of what someone recommends from genre A but they unfortunately also happen to be a big fan of genre B which I can't stand, making their overall corrolation to my tastes closer to 50%).

Anyway seems interesting, I'll follow it and see how it works before making additional suggestions.

[ Parent ]

variations (4.00 / 5) (#25)
by dr k on Wed Apr 17, 2002 at 05:25:32 PM EST

First, I hope/assume that your rating system is mathematically sound, you aren't just taking an averge, the sum of all ratings divided by the number of ratings. Hm, after looking at the site I'm not too hopeful.

Second, mass-market music is about more than just popularity. There's a lot of money going around, payola is alive and well, distributors regularly put the squeeze on independant record stores, &c. Popular music is popular because 90% of the audience doesn't know any better, or they really don't care what they listen to. Education would be the most effective way to get these folks listening to something new.

Third, I personally cultivate friends who have an interest in music, who keep me informed of the latest Death Chant releases, and I caution them not to buy the new Crystal Method. I don't really have a point here, I just wanted to plug Death Chant.


Destroy all trusted users!

Mathematically sound? (4.00 / 2) (#28)
by garyrob on Wed Apr 17, 2002 at 06:07:07 PM EST

Yes, it's mathematically sound. It's based upon Bayesian statistics and was created with the help of this expert in the field at Wharton.

CAVEAT... we're just starting up, there aren't all that many ratings yet, so the engine doesn't have much to work with yet.

Re education... that's why written recommendations play a big role on the site. The best ones give a context to help people know where the artist is coming from or provide other keys to listening and the recommendations themselves are rated partly on that basis.



[ Parent ]

statistician (4.00 / 1) (#31)
by dr k on Wed Apr 17, 2002 at 07:58:31 PM EST

Well, then I'm going to assume that your closed 6-point rating system [poor/average/good/v.good/excellent/"supreme"] is just a user-friendly way of presenting the underlying open-ended survey data.

So tell me this... can a record - say, Farmer Jim's Musical Saw Greatest Hits - score poorly in comparison to all music, yet be ranked as "supreme" within the small-but-relevant genre of musical saw performances?


Destroy all trusted users!
[ Parent ]

Yeah (4.00 / 2) (#33)
by garyrob on Wed Apr 17, 2002 at 08:40:45 PM EST

"Well, then I'm going to assume that your closed 6-point rating system [poor/average/good/v.good/excellent/"supreme"] is just a user-friendly way of presenting the underlying open-ended survey data."

Yeah, under the covers those ratings are translated into another kind of scale before being fed into the main processing. It's an attempt to be user friendly. I have seen collaborative filtering sites that tried things like continuous real-valued "rating" input but in the end they have changed to discrete levels of rating input because users find it easier.

"So tell me this... can a record - say, Farmer Jim's Musical Saw Greatest Hits - score poorly in comparison to all music, yet be ranked as "supreme" within the small-but-relevant genre of musical saw performances?"

There are detailed rating criteria. One of the reasons we go to so much trouble to determine who is good at rating, thereby enabling us to reward the right people, is so that they can go the extra step of rating according to criteria.

The rule is not to rate things that are not in your sphere of interest. We aren't separating ratings according to different detailed categories as you suggest. We are waiting to see if it would be helpful to do so and will if it is.

If people rate things that are not in their sphere of interest, such as somebody who doesn't care about musical saw performances rating such a performance, they will be at variance with the people who do care about that music. That is, their rating will not be predictive. So they will have a poor reputation for rating, and their ratings will be given a very low weight, and also will lose the ability to get free review CD's and publicize performers they care about.

Now, this mechanism alone may or may not be enough. We'll see. We certainly have the power to break the ratings into totally separate spheres as you suggest and have gone to great lengths in the math processing to do everything right. So we'll do that if it turns out there's a reason to; if not we won't. There's nothing much else to it.

You're asking the right questions, but I think you've been jaded by products that are not seriously put together by people who care, and therefore assume the worst. But we do care and are doing things right and will do them righter as time goes on. I think that the more you know about what we're doing the more you will be convinced that we're doing things as you would personally like. I agree with your cynicism in general about the way people often do things. You might want to check out my bio if you're curious about my personal background for this kind of project. Or you may not. ;)

[ Parent ]

music of interest (5.00 / 1) (#41)
by Delirium on Thu Apr 18, 2002 at 04:19:08 AM EST

One problem I see is how to define music I'm interested in. Since the idea here is to find relatively unknown music, one might assume that a lot of it doesn't fit well into traditional genres. So it's difficult to tell if I don't like some particular bit of music because I don't like the genre, or because I don't like the song. Or perhaps I don't like the mixture of genres, or think it's poorly done, etc. I suppose one could stick to only rating genres one is a big fan of, but that would seem to defeat the purpose...

[ Parent ]
RIGHT! (none / 0) (#59)
by garyrob on Thu Apr 18, 2002 at 03:37:52 PM EST

Exactly. If you're interested, please see my comment "Response to questions raised here" on this page for much more discussion of this issue!

[ Parent ]
perils of userfriendliness (none / 0) (#64)
by martingale on Fri Apr 19, 2002 at 06:52:44 AM EST

You're obviously purposefully vague, but in translating the ratings into another scale, you're not going to change the information content of the user's rating. Basically what I'm saying is that if you request users to rate on a scale from 1 to 6, then whatever data transformations you do on the rating, you'll still have a rating from 1 to 6. You'd have to randomize the user's rating if you wanted better resolution for the data, but that would be done algorithmically by the back end, so wouldn't add any useful information.

By the way, kudos for displaying the ratings also in the ALT tag, so I could see them in my text browser. However, the ALT tag actually repeats the textual description below it, so perhaps you might put a number of hashes there instead.



[ Parent ]
MOTR (4.66 / 6) (#35)
by TON on Wed Apr 17, 2002 at 09:57:26 PM EST

Let me see if I have this right.

Ratings are rated up in confidence based upon the number of subsequent ratings that agree with the first. So, ratings that fall outside the mainstream will probably get low confidence, but ratings that match the general public's consensus will gain in confidence. Won't this tend to simply promote fairly middle-of-the-road ratings of middle-of-the-road music? True, this music may emerge from unknown to known, but where is the benefit? There is already more MOTR music out there than most anybody can keep up with. Or care to, for that matter.

Even if genres, and micro-genres are implemented, this will just perpetuate genre conformity. Where does genre breaking/crossing music emerge in this scheme? It is precisely this kind of music that tends to have difficulty breaking into the commercial marketplace and getting people's attention. If the music doesn't fit a neat pigeonhole, a wide range of comments with indeterminate confidence may be the result.

I'm all for a way for independent musicians to get a wider audience, but I'm not sure that this is it. If I've got it wrong, let me know. I guess I'm looking for more detail about the technology.

"First, I am born. Then, the trouble begins." -- Schizopolis

Ted


The journey is more rewarding than the destination (4.25 / 4) (#36)
by IoaPetraka on Wed Apr 17, 2002 at 10:05:06 PM EST

I have a great joy for listening to music. I have an even greater joy for finding music.

Perhaps I fit into a rather small demographic, but part of what draws me to music is the chase. I pretty much listen exclusively to underground, mostly unheard of music acts, in genres that practically do not exist according to the charts. I don't belive that I've taken this route because I feel the music is any greater, but because that route is more interesting for me.

I usually allocate a good two hours a week to researching and discovering new forms of music, and new musical acts to buy in to. Nothing is more rewarding that suddenly hitting pay dirt, and finding that perfect sound for the moment. If I find a group that can provide that, I'll make haste to purchase their music.

While on the surface a tool like this technology would seem to be something great for a person like me, in a way I feel it would subtract from the experience of finding music. If I had my tastes were quantified into a metric, and mathematically cross-indexed with like-minded individuals (no matter how elaborate and "accurate" that model may be), producing a Dream Wish List, I would lose part of what makes music so enjoyable to me.

Hey, like I said, I am probably in a minority wedge of the chart, most people would probably enjoy a tool like this. I just feel that listening to music, and finding music, is a personal art that only you yourself can perfect.

.:.
Ioa Aqualine Petra'ka

But we need you (3.00 / 1) (#37)
by mgoyer on Wed Apr 17, 2002 at 10:58:45 PM EST

Ah, you may not need Emergent Music or similar services but the rest of us who don't have the patience to spend hours uncovering those undiscovered gems need you.

We need you to break our Clear Channel dependency.

Matt

[ Parent ]

Tell you what... (none / 0) (#61)
by IoaPetraka on Thu Apr 18, 2002 at 08:26:59 PM EST

I'll give it a try, if I like the system I'll hang around. I'm certainly not opposed to helping others find good music, just because I prefer the 'hunt.' I do admire your intentions, and it is always refreshing to see an organization that acknowledges the weaknesses of the normal methods of music aquisition, such as Clear Channel, Spin, Entertainment Weekly, and the many other "services" that are essentially just paid programming.

I do have a few recommendations. Much of the more obscure, and new music cannot be found at the standard online vendor sites. A lot can, but there are other vendors that are more active in promoting fringe genres and new groups. OtherMusic, and Darla are just a few that come to mind. Both of which are a great asset for those looking for alternative musics, as they provide write-ups and mp3 samples of nearly everything they sell. These types of vendors should be given representation in your choice list if possible.

Another quibble, you need "Ambient" and "Experimental" and/or "Post Rock" added to your genre list. Those are the three I frequent most often, and I really cannot choose between the supplied genres.

.:.
Ioa Aqualine Petra'ka
[ Parent ]

OK (none / 0) (#69)
by garyrob on Fri Apr 19, 2002 at 04:11:20 PM EST

We'll implement your suggestions. Thanks. Please give us more feedback on our discussion board at EM as you explore it. I caught the messages you've posted so far.

[ Parent ]
YEAH! (3.00 / 1) (#58)
by garyrob on Thu Apr 18, 2002 at 02:59:46 PM EST

As Matt said, we NEED you. We exist for people who feel like you do, and who would also enjoy having the power to help artists you love become introduced to their natural audiences. You have the chance to help artists and music lovers each have better lives. EM provides an outlet where you can have that effect.

I really hope you join us.

[ Parent ]

The fundamental problem with your method (4.50 / 4) (#38)
by dukethug on Thu Apr 18, 2002 at 12:17:05 AM EST

Is that, on the whole, members of your community will trend towards the middle-of-the-road with "Good" or "Very Good" ratings. You don't give them any incentive not to.

I, for example, am a huge fan of the Olive Group, an Austin area band. I think they're stuff rocks. I recognize, however, that the vast majority of people will just appreciate their stuff as "good," and not have the same religious experience while listening to it that I do. So I would get more points by rating them simply as "Good", a middle-of-the-road rating. The same thing is true for all of the music on your site, to the point where everything is rated roughly the same, and there's no more distinguishing the good from the bad than there was before your site came along. If you can tell me why I'm wrong, I'd be more than happy to hear it, because the need for what you're doing is there.

Either way, I'm a statistician by trade, so I'll be hanging out on your site just so I can figure out how your variant of Bayesian statistics works, even though I think that the problem you are trying to solve could be better approached using non-parametric methods.

Think about it. (none / 0) (#54)
by garyrob on Thu Apr 18, 2002 at 01:51:54 PM EST

[Note, I posted a reply to this comment before, and then the kuro5hin server seemed to hiccup and I can't find my reply, so I'm going to try to recreate the same reply. Forgive me if the earlier one pops up somewhere and there are two versions.]

Since you are a statistician by trade, I believe that if you think about it long enough, you will probably be able to come up with solutions to the problem you point to in your comment.

We have our own solutions, which are trade secrets at this point and will be part of our intellectual property portfolio. I can only tell you about them under NDA and will if you are interested enough and we see some benefit from our side -- if for some reason you are seriously interested, email me and let me know more about your background (grobinson@transpose.com).

I think the real problem here is the one I brought up elsewhere, that you're assuming that we are stupid marketers who can't think of these problems for ourselves and certainly can't solve them. It's an assumption that is hard to avoid given the low quality of most Internet-bubble startups. But it is not an accurate assumption in our case. We are engineers, serious about actually providing a system that does what it is supposed to do. For more on this aspect of things, see my comment "Response to questions raised here" elsewhere on this page.

[ Parent ]

Um (4.50 / 2) (#40)
by ghjm on Thu Apr 18, 2002 at 04:05:54 AM EST

Doesn't your system depend on the notion that two songs can be 'better' or 'worse' than each other in absolute terms? Is it meaningful to average one person's opinions with another? Once you discover some music that is 'better' what do you do with it then?

No... (none / 0) (#52)
by garyrob on Thu Apr 18, 2002 at 01:01:19 PM EST

...not in absolute terms, but to people who are interested in the TYPE of music that thesong represents.

We posit that if we take a group of people who all like backwards accordian music, then to that group there is a KIND of absolute goodness for each song. VERY roughly, a way of approximating the idea is as the proportion of people IN THAT GROUP who like each of that kind of song. (That is a rough approximation in the way that a 2- or 3-dimensional image is sometimes used to represent concepts in n-dimensional space!)

If NOBODY likes the song, not even people who like that type of music, then it may have some kind of abstract goodness to it, but it isn't a kind of goodness I'm interested in. It's not a kind of goodness that is useful to people trying to build methods to help people discover high-quality undiscovered music.

So, there is a kind of objective goodness which has to do with only looking at the opinions of people who care about that kind of music. It's a KIND of goodness... I can imagine the existence of valid philosophical arguments that there are other kinds of goodness, but this is a practical kind of goodness for real-world uses.

[ Parent ]

Problem solved (4.50 / 4) (#42)
by gnovos on Thu Apr 18, 2002 at 04:23:08 AM EST

Now, some ratings sites provide rewards to all raters.

That, right there, is the problem. Providing rewards will consistantly prove to fail. Truct me. Want to know the best way?

Easy, ALL songs get rated as "extra bad", not neutral, I mean the worst of the worst rating. Only songs that have a following, any sized following, will get rated up. Yes, you will miss out on the 0.001% of Mozart-skilled virtuoso solo artists that keep thier music secret from thier friends and family, but for the most part, you will see that the fans that really really love a song will go out of thier way to get thier favorite band out of the gutter. From that point on new people will find the music and start rating it more realisticly. You also put a *fade* counter on the ratings, so that every week or so, it loses X points. The idea being that it takes a real struggle to keep it afloat. Bands that consistantly stay above the waterline, despite the downward pressure, are guaranteed to be the cream of the crop.

A Haiku: "fuck you fuck you fuck/you fuck you fuck you fuck you/fuck you fuck you snow" - JChen

Exceptions (3.00 / 1) (#45)
by tekue on Thu Apr 18, 2002 at 07:35:57 AM EST

Well, I've found that music made by humble, quiet, sensitive people is the one that suits me the best. Those people are not likely to break out in your scheme, which I consider a flaw in it.

Remember, all things considered, the point is to find music which would not surface any other way -- i.e. music made by people who are not that much outfront and loud. It's to find those people creating beautiful music within the confinments of their own bedroom/basement/garage, not those who actively take part in their own promotion.

Those who fight for it strong enough will surely gain some attention anyhow.
--
Humanity has advanced, when it has advanced, not because it has been sober, responsible, and cautious, but because it has been playful, rebellious, and immature. --Tom Robbins
[ Parent ]

I have to disagree (none / 0) (#55)
by gnovos on Thu Apr 18, 2002 at 02:02:33 PM EST

When I think of humble and quiet people, I think of people like Moby, who I've never seen say a harsh word, or even raise his voice, but he's got a huge following.

A Haiku: "fuck you fuck you fuck/you fuck you fuck you fuck you/fuck you fuck you snow" - JChen
[ Parent ]
"If you don't have anything nice to say... (4.00 / 1) (#47)
by buffy fan on Thu Apr 18, 2002 at 08:43:53 AM EST

.. don't say anything at all" (as I was told when I was young)

I think you are on to something here - Because music is so subjective, it tends to evoke polar (and often quite snobby/bitchy) responses. Having a system that enables this is asking for trouble.

Also, the idea of micro-genres is interesting as a grouping strategy, but the subjectivity of music means that this may be hard to implement consistently (Insert all the standard Usability Arguments about Defining Searches here)

Who actually has such well defined musical tastes? Surely people tend to like a variety of styles? With this in mind, would it not be more logical to emphasise peer behaviour, like Amazon does? i.e. People who downloaded this and rated it favourably also downloaded ....
This could potentially be extended to allow you to team up with other like minded individuals (come on, are you *really* as eclectic as you think? ;)

Basically, don't give people the chance to denigrate something - let positive factors e.g. Number of downloads, number of positive reviews, newness of file, peer similarity etc drive the system.


[ Parent ]
Actually... (none / 0) (#51)
by garyrob on Thu Apr 18, 2002 at 12:53:51 PM EST

...it does just happen that there is an element sort of like what you are suggesting in what we are doing, which is a direct result of our Bayesian approach.

[ Parent ]
my day with emergent music dot com (4.66 / 3) (#44)
by dr k on Thu Apr 18, 2002 at 05:31:57 AM EST

I've spent a little time with the site now (hey, shameless self promotion actually works sometimes), and I've got a few observations:

First, the structure and ultimate purpose of the site are really confusing. When you rate a review, you're rating the quality of the review itself, not the artist it refers to. This leads to some confusion about the ultimate scope of the site, as people tend to rate down reviews of non-emergent - i.e. "established" - artists.

Secondly, the rating system has already fallen victim to the same sort of "ratings game" that plagues the web everywhere you go. This person has a Guide for Ratings, that person has a Recommended Rating Chart, &c, &c. That means that anyone with a bias against, say, the band U2 will rate down all reviews of that band. And while it may be the case that reviews of U2 fall outside the site's charter, there should be a legitimate mechanism for dealing with this issue outside of person A's arbitrary rating system.

Finally, I don't see where "emerging [sub]genres" get put into the mix. Each recommendation is limited to one genre, and genres are - in general - poorly defined ways to classify the divesity of artistic endeavors. Genres are, from a database perspective, a many-to-many relationship, and if your programmers balk at this notion, fire them and give me the fucking job.


Destroy all trusted users!

Nuts (none / 0) (#56)
by garyrob on Thu Apr 18, 2002 at 02:05:42 PM EST

I replied to this comment already and now my reply seems to be gone. Maybe it was deleted by a moderator for some reason? If so I wish kuro5hin provided me with feedback so I'd know whether I just had some technical problem in posting or whether my reply was deemed unacceptible (in which case it would be helpful to somehow send me a reason...). Or maybe you do but I just haven't figured the site out well enough yet.

Anyway I'll try to answer it again...

1) The site is not as easy to understand as it ideally should be. We're working on it, it's a tough problem.

2) I don't understand your comments about U2... recommendations of old-hat stuff like U2 IS outside the charter of the site, and that is legitimate and people will rate them down accordingly.

3) Emerging subgenres... that is something that is coming in a few weeks, not there yet. There would be no point in having it now because we don't have enough users yet to "power" it. It would only make things seem more confusing.



[ Parent ]

problems with U2 (5.00 / 1) (#60)
by dr k on Thu Apr 18, 2002 at 04:27:31 PM EST

Over here you have the site charter, and over there you have someone "recommending" U2 on the site. Now it might be an extremely well written review, and people who aren't familar with the charter, or simply don't care, will vote it up, while others scream "Stop voting this up!" and vote it down. So the ratings have at least 2 meanings: quality and appropriateness. You have no way to determine how much these are influencing the voting. And yet, you give special status to those who vote more in line with the majority opinion.

It is very similar to the broken trusted user model here on k5. Everyone starts [started] out equal, but as a status quo is established, people start to write comments that they know will simply appease the trusted elite. In a site where you are looking for the new and radical (emergent music is, after all, not something people are going to lack opinions on), it is unfortunate that you would want to establish this same kind of stodgy community.


Destroy all trusted users!
[ Parent ]

trust system broken? I don't think so (n/t) (none / 0) (#65)
by martingale on Fri Apr 19, 2002 at 07:20:49 AM EST

Do you have a longer rant on the "broken" trusted user model you can point me to read? I don't think your objections are meaningful.

The trusted elites AFAIK differ from the others by being able to rate to zero. Given that a comment isn't rated to zero, there is no difference in behaviour between the elites and the others. My experience is that interesting and well crafted comments tend to be modded up (ie not around 1,2) regardless of their content, but crudely stated and obviously unfinished thoughts vary widely in rating. Just because you think it's an interesting comment, that doesn't mean others do.



[ Parent ]
Recommendations? (5.00 / 2) (#48)
by afree87 on Thu Apr 18, 2002 at 10:23:10 AM EST

Recommendations based on what? When I visit Amazon.com, I get recommendations based on what I've told them I've bought and how much I liked it. When I visit your website, I get recommendations based on what everyone else likes. Guess who I'm going to keep on visiting?
--
Ha... yeah.
Er... (3.50 / 2) (#50)
by garyrob on Thu Apr 18, 2002 at 12:51:09 PM EST

Frankly Sir, you don't have a clue how Amazon makes those recommendations, which is by comparing your expressed likes to other people, finding those who are similar to you, and then seeing what THEY like, and finally recommending that to you. (At this point they are not be doing this as simply and directly as the technology produced by Net Perceptions, the supplier of Amazon's technology, did originally, but ultimately it is the same principle done more efficiently.) As someone who is personally responsible for creating that kind of technology in the past, which performed well relative to real world competitors (including against Net Perceptions technology) in testing with real-world data, I can tell you that we are rejecting it only because we are doing something we believe to be more powerful, and we DO have that older type of technology wating in the wings in case any need appears for it. If you are interested in understanding more, you might want to see my comment on this page entitled "Response to questions raised here".

[ Parent ]
Huh? (4.00 / 1) (#62)
by afree87 on Thu Apr 18, 2002 at 09:53:38 PM EST

Frankly Sir, you don't have a clue how Amazon makes those recommendations, which is by comparing your expressed likes to other people, finding those who are similar to you, and then seeing what THEY like, and finally recommending that to you.

I know that perfectly well. I'm not assuming Amazon has a team of people making sure I'm getting recommended the right things; I'm simply saying that it seems to me that their system works much better than yours. If a person who doesn't like the sort of music your community likes enters your site, he'll soon be leaving without buying anything.

Your "micro-communities" idea seems like it might work, but I'll still take the "old" technology over that any day.
--
Ha... yeah.
[ Parent ]

Er... (none / 0) (#67)
by garyrob on Fri Apr 19, 2002 at 02:41:29 PM EST

Your "micro-communities" idea seems like it might work, but I'll still take the "old" technology over that any day.

At a pure level, the "old" approach would build a community around you that was the closest people to you, and see what they like, and recommend it to you.

I have built that kind of technology. In fact I was the first to build that kind of technology, around 17 years ago.

The problem is, that it isn't scalable. In order to scale it, tricks are used that APPROXIMATE the effect of the pure approach.

The micro-genre approach can be viewed as a way of doing that approximation, but with other ancillary advantages.

With any collaborative recommendation technology like Amazon's or ours, you're getting an approximation of your individual ideal cluster of recommenders. Really, you think it's obvious that the old way will perform better, but the fact is that it just isn't true, and your assumption is only built on lack of first-hand knowledge of what the "old way" really is under the covers. Of course, many people are only interested in hearing that their initial unfounded assumptions are correct, so if you are in that group, it won't matter what my background is or what I say, which is fine, but if you are not in that group, and are interested in the subject, the following may be information of value to you.

Any given recommendation system may work a bit better than another, but there is nothing from first principles that says that a system's like Amazon's can perform better. (In fact, as I mentioned elsewhere on this page, they bought their technology from Net Perceptions, and my technology of that time (circa 1995) beat theirs in actual head-to-head testing for recommendation accuracy.) If there were reasons from first principles why their way was better, we wouldn't be doing it the way we are.

Remember the key is that for scalability reasons, because it is too computationally expensive, an age-of-the-universe CPU time expenditure, Amazon's way only approximates what you think it does.

The approximation technique we are using through micro-genres has ancillary advantages; for instance it provides clear places for people to put new music where it will get exposure from people who like that kind of music; it provides hierarchical navigation so you can quickly find things of interest to you without entering a lot of ratings, etc.

As an experienced professional in this area, my guess is that if EM scales up to getting a large number of users, the two systems are likely to perform indistinguishably similarly purely from a standpoint of making reliable recommendations based on the data in each respective system. But the key is that the micro-genre approach is far more supportive of other techniques to enable undiscovered artists to emerge without mass-marketing. That's why we're doing it this way. Amazon, unless it eventually does something like what we are doing, or finds another way to achieve those ends, will always have a mass-markiet bias, and that will LOWER THE CONCISENESS WITH WHICH THEIR MUSIC IS ATTUNED TO YOUR INDIVIDUAL TASTES, unless you really only want mass-market music. There are people like that, and they will always be happier with Amazon than with us. People who are true music lovers, who aren't just into current or past mass-market trends but actually want great music, however, are likely to be happier with us as we scale up.



[ Parent ]

'course (none / 0) (#68)
by garyrob on Fri Apr 19, 2002 at 02:45:11 PM EST

you might also just be troll, which is why I probably won't pursue this thread further. ;)

But I thought the real reply I posted might be of interest to other people.

[ Parent ]

Response to questions raised here (5.00 / 2) (#49)
by garyrob on Thu Apr 18, 2002 at 12:19:38 PM EST

Folks, [Note, this comment is long. Matt Goyer just told me my comments should be short. But there are intelligent people here asking intelligent questions that I don't think can be meaningfully answered in just a few words. So I am going to post this ONE long comment for those who are interested in a meaningful, detailed answer, and try to pare it down when I can in other replies I make.]

I am going to respond to "buffy fan's" questions in a new comment rather than burying it deep in the thread where her or his questions occur, because the questions resonate with things a number of people seem to be curious about.

With respect to our micro-genres, buffy fan says, "would it not be more logical to emphasize peer behavior, like Amazon does?"

I think this is a great question and for those people who are interested in such things, I am going to address it in this comment.

First, I think it might be useful to give some history, so that readers may be a bit less likely to assume that if we aren't doing things the way that appears most logical upon a quick examination of the problem, it's not because of ignorance, laziness, etc. There are OTHER reasons.

For those who aren't interested in that aspect and are only interested in a direct response to buffy fan's questions, please scroll down to "DIRECT RESPONSE TO BUFFY FAN'S QUESTIONS".

I have noticed that in the discussions here, there seems to be a fundamental assumption on the part of some intelligent people that we are being... how do I say it... stupid in our approach. That we haven't even considered certain obvious issues and problems and strategies. That the first thing an intelligent reader of these threads might think of is obviously a better strategy than ours.

I think that such assumptions are made because intelligent people can get jaded by the amount of crap out there. There ARE many crappy web sites out there, created by people who think that good marketing, rather than good technology, is the way to create a successful company.

But this company is honestly technology-driven. We don't even have a marketing person aboard at this point. It happens that I was the first person to create a commercial system that does what buffy fan suggests, in the mid-80's. It was a voice-mail-based dating service that monitored people's keystrokes in response to voice ads of the opposite sex... you could fast-forward over them, reply, store them away for later listening etc. These were considered to be "passive ratings" and we formed a cluster of similarly-minded peers around every user; these peers were used to recommend people of the opposite sex.

I know that it was the first such system because I've talked to people like the CEO and top tech guys at Net Perceptions (NP), which supplied Amazon.com's technology, and everyone so far seems to agree that mine was the first commercial application (or maybe the first application period; that's harder to establish) of the principles that a decade later came to be known as "Collaborative Filtering" (CF).

It also happened that later CF technology produced by my company beat the published technology of NP's people (they were at the U of Minnesota or something like that, I'm not sure I'm remembering the exact school as I write this) before they started NP. At the time they were academics and published their work in academic papers.

We tested ours against theirs on a database of actual movie ratings and ours more reliably found peers that made recommendations that actually turned out to be liked by the person receiving the recommendations. Again, this was the core task to the exact solution that buffy fan proposes (if I understand him or her correctly).

You may have heard of another company, Firefly, that came out of the MIT Media Lab with similar technology. We also tested ours against their published academic work and ours again proved superior at finding peers who could make reliable movie recommendations. They were subsequently purchased by Microsoft.

And I know exactly why ours beat theirs in both cases (and in fact in the case of every peer-finding technology we tried). It was because they didn't put much effort into theirs. They used the most obvious things as their solutions. That was enough to get a paper published or to get funding (at least if you're a personal friend of a top-5 software industry VC, as one of the founders of NP was ). It was shocking to me. But that's how it was. Whereas we DID seek, from the beginning, to do things as well as we could from a technological perspective, not because we are morally superior or geniuses, but because we are engineers instead of marketers.

In fact, when I criticized Firefly's technology to their CEO, his response was not, as I expected, "We're working on it, here's what we're doing..." Rather it was, and these are his EXACT words, "You can make up for that with marketing."

So I don't blame people here for assuming the worst about us -- that we aren't serious about technology, that any smart person who thinks about it for 10 minutes can come up with a better solution or figure out why our site will never work, etc. I do think that is often the case with new companies, particularly in the Internet space, so it is rational to initially assume we would be in that group. But there are other kinds of companies too, driven by quality of engineering, which we seek to emulate: 3M and Google are among them.

DIRECT RESPONSE TO BUFFY FAN'S QUESTIONS:

Going back to buffy fan's question: I agree, You are right, it DOES make sense to "to emphasize peer behavior" as you suggest which is why I have been responsible for the creation of a fair amount of technology based on that principle.

However, I believe that the evolutionary micro-genre principle has the potential to be more exciting, which is why we are going to bring it onto the site before the more traditional peer technology you suggest. (BTW, we have already created such technology, based on information theoretic principles, and written the code for it, and it's waiting in the wings.)

buffy fan: "Who actually has such well defined musical tastes?" I agree that that question makes a lot of sense with respect to the kinds of rigid, superficial genres that the music industry has used so far in defining genres. Personally, I have never had much use for those kinds of genres. For instance, if I get on ANY Internet radio station, no matter how many HUNDREDS of streams it may have each depicting a specific small genre, I have NEVER consistently liked the music.

So, if we were talking about those kinds of genres, I would agree with you, they would have no use to me either. I consider it proven that those kinds of genres do not work.

But we are envisioning genres that evolve directly out of community need. The genres that best serve the actual community's need for focused taste-centers will emerge, we posit. This is totally different than the top-down approach where some industry analyst decides how to group music and what to call the groups. This is a bottom-up approach. I believe the results will be very different.

In fact, if I called them emergent-evolving-micro-taste-centers rather than micro-genres, I suspect that this whole question might not have even emerged. But I just think that micro-genres is a more succinct term.

The taste-centers that emerge should be clusters where people with ACTUAL similar tastes want to congregate in order to most effectively share music with like-minded people... that is, peers.

I posit that these taste-centers will not be far from the clusters that are created by more traditional nearest-neighbor (peer) based clustering techniques, but will have some advantages, like having distinct identities, the ability to be reached quickly through navigating a hierarchy rather than by inputting hundreds of ratings and mathematically comparing yours to those of other people, and others.

buffy fan: " Surely people tend to like a variety of styles?" Of course. I, for one, like Beethoven, jazz guitarist Ralph Towner, folk songwriter Leonard Cohen, and music of other types. I'd say my tastes are quite varied.

I see this as a nice thing about the micro-genre approach. When I am in the mood for songwriting with mystical, poetic lyrics, I will go to one micro-genre. When I'm in the mood for lively jazz sax playing, I will go to another. Having this kind of micro-genre will make it easy for me to do that.

Whereas, if I am grouped with peers who happen to also like Beethoven, John Coltrane, and Leonard Cohen, as our earlier CF technology, and Net Perceptions, and Firefly's all did, I will get finely tuned suggestions, but they won't necessarily be what I am looking for at the time, and I will have had to enter hundreds of ratings (or made many purchases) to get those recommendations.

But you're right... that peer approach does capture something special. I currently think that the main common thread in music that I like is that it is honest. So, if peer-based recommendation technology considers all the artists I like, it will find peers who also like honest music instead of manipulative, saccharine music, EVEN THOUGH THERE ARE NO TRADITIONAL GENRES FOR "HONEST" MUSIC. IMHO that's the great strength of the peer approach, and that's why I spent years focused on developing a sound theoretical framework for doing that.

However... I posit that the evolutionary, emergent micro-genre approach WILL DISCOVER THAT SAME INFORMATION. I think there will be micro-genres that capture honest folk songwriting as opposed to manipulative, saccharine folk songwriting, which would be captured in other micro-genres. The names of those genres will come from the fans of artists in those genres, so I assume that the "manipulative, saccharine" folk songwriting genre will be called "Romantic Folk Music" or something like that. ;)

I think that such genres will emerge because they are being created to suit the needs of individual music lovers rather than a corporation's need to categorize for mass-marketing purposes. They are a totally different animal.

NOW, that being said, I am quite aware that I may be wrong about what will happen with our evolving, emergent micro-genres. I don't think I am, but I may be. It is, after all, an experiment -- one that can only be meaningfully conducted in public on a large scale with a many, many users. That's why it's going to go directly into our Web site rather than proven in a lab somewhere first.

SO, we also have a more traditional recommendation technology waiting in the wings. It finds peers by means of information-theoretic calculations. It is, I believe, a step up from our earlier peer-finding technology. It finds peers who have similar tastes without restriction to traditional genres and groups them together for purposes of recommendation. We have already written the code for this. We just aren't integrating it into the site now because we think the emergent micro-genre technology is a more interesting and exciting. Unless the emergent micro-genre (I guess I'll call it EMG) technology works out incredibly will, we plan on introducing the peer-finding technology subsequently. We don't know how much value it will add over EMG because EMG hasn't been tested yet.

I hope this answer to buffy fan's question gives interested readers of these threads a better idea of what we're about and what we're trying to do.

EmergentMusic is good, but MovieLens is better (none / 0) (#73)
by robertb on Fri Jan 24, 2003 at 08:21:15 PM EST

I do like your ideas. I assume that when you say "peers" you are doing some sort of clustering. And I assume that that's in addition to the Bayesian statistics. Since computers are getting faster and faster, you might consider something deeper -- you know, decision trees/hyperplanes. I'm sure you've looked into that.

In any case, I really like MovieLens. Perhaps you were implying that you've already compared your technology when you mentioned Net Perceptions, but I find MovieLens to be quite good at predicting enjoyable movies so far. And, despite it being based on "marketed mainstream" movies, MovieLens still manages to bubble up the weird alternative movies which I like up to the top.

I tried EmergentMusic. I prefer punk rock and its selection was quite tiny. And many of the links were either dead or I was unable to listen to the music. Perhaps the service is just not popular enough. It's really a shame 'cause there's just so much music to hear and it's hard to be selective.

Anyway, good luck!

[ Parent ]

marketing music (3.00 / 1) (#53)
by clark9000 on Thu Apr 18, 2002 at 01:05:00 PM EST

garyrob--

I'm an avid music addict, I checked out your site, your comments, the open letter to the EM community in your blog, and I think you have a great idea here. I have a friend who works for Sony music and I was just discussing with him last night how it's too bad that only the mega-hit stars make it, and so many good but unknown artists are ignored.

My question is about where you feel the importance of the non-auditory falls into this discussion. As you say, the reason so many people listen to the big stars is not always that they are musically better, but that they have billion-dollar music companies behind them, that run huge marketing campaigns to get their artists heard. But the thing is, the framework in which the sound is presented makes a difference. Britney Spears music (we'll pick on her since she's the icon of the phenomenon) is not just a collection of sound waves. She is the color pink, lip gloss, tight clothing. Often, people who like Britney not only like the sound files she produces, but they also like pink, lip gloss, and tight clothing. (Conversely people who dislike her tend also to dislike all those things.) It's all part of a comprehensive music experience. The fact that Louis Armstrong grins ear-to-ear, while Miles Davis turns his back to the audience, affects how we experience their music.

So, since record companies traditionally provide this framework, unsigned, unknown artists have to find some other means of providing it. Web sites, for instance, help fill this role--pictures, bios of the artists, manifestos (depending on who it is), go alongside the mp3 files.

So my question is, do you propose other methods? How does this fit into your vision for EmergentMusic.com? Or do you think that musicians should be judged solely on their audio output?
_____
Much madness is divinest sense
To a discerning eye;
Much sense the starkest madness.

-- E. Dickinson
That's an interesting question... (5.00 / 1) (#57)
by garyrob on Thu Apr 18, 2002 at 02:56:06 PM EST

I guess if I had to give an answer, I'd say that I think those factors will come to play a different role. Today, as you point out, non-auditory factors play a huge part in marketing the artists because catchy visual images, controversial personalities, etc. are very effective ways of getting mass-market attention. I suspect that as we replace mass-marketing with micro-marketing techniques using the Internet, those factors will become less important because it will be more possible for auditory factors to take a stronger role in spreading the word.

But I can imagine something like EM where the whole artist package, non-auditory and auditory together, is considered more than it currently is with EM. You're right, even with a true artist like Miles Davis the persona of the artist is part of the experience.

Even so, I tend to think that artists will develop relationships with their audiences that are less based on persona and more based on actual relationship, as the Internet enables artists to interact directly with people with no risk of physical harm (for instance from interacting with a fan who turns out to be a murderous stalker). I suspect that some artists will add to their incomes by being "sponsored" by wealthy fans who, for a substantial amount of money, would get some kind of direct access over the Internet such as video chat...

Anyway, it's an interesting question and even brings up areas where EM might evolve in the future...

Thanks for your comment! (And for the kudos, which I appreciate.)

[ Parent ]

I'm not sure (none / 0) (#63)
by clark9000 on Fri Apr 19, 2002 at 04:14:52 AM EST

I suspect that as we replace mass-marketing with micro-marketing techniques using the Internet, those factors will become less important because it will be more possible for auditory factors to take a stronger role in spreading the word.

Well, you may very well be right, and part of me hopes that you are, but I disagree. Non-auditory factors have always played a big role in how we enjoy music. There's a continuum: for some people, image is very important, sound less so, for others, image is less important. Sometimes people are not even aware of how image fits into their enjoyment of music, but it is always there, in the background. Musicians who create a persona (real or fictitious, little matter) to go with their music will always have a leg up on musicians who don't.

Your point about how the internet enables safe fan-artist interaction is a good one. Seems a great way for unknown artists to become known is for them to answer all email, for instance. But for a musician to say, "I don't need to make a web site, I don't need to answer email, the music will stand on its own," seems foolish to me (unless that's part of the persona, of course).
_____
Much madness is divinest sense
To a discerning eye;
Much sense the starkest madness.

-- E. Dickinson
[ Parent ]
Yes (none / 0) (#66)
by garyrob on Fri Apr 19, 2002 at 10:50:49 AM EST

It's just hard to say. It seems obvious to me that the image stuff is intrinsic to how music is marketed now. So, the importance of those factors in enjoying the artist's work are a bit hard to measure, because their current prominence is caused, to a great degree, by the record company's need to market is cost-effectively as possible. So, our data for measuring the importance of those factors is skewed.

My guess is that for people who REALLY love music, it won't matter much. For instance, I agree that Miles Davis' persona is interesting and part of my experience of his music. But I would still love his music without knowing anything about him. Wouldn't you? Really? There are other musicians that I love who I know little about. I enjoy getting information when its available, but it isn't really core to my experience. The fact that I like getting extra info does not mean that it is core to the experience.

I'm equally sure there are other people for whom the image is everything and the music is less important.

I have thought of that as a mass-market phenomenon. But this discussion raisesd a whole knew point in my mind... artists may emerge who are not mass-market, who are a combination of musician and performance artist, and attain a small but loyal worldwide following over the Internet.

Currently, EM is not supporting that. But it could. New forms of performing could even emerge that are not consistent with mass-marketing but are consistent with micro-marketing.

This is very, very, interesting to me. Worthy of a discussion with more people here on kuro5hin. Not sure how to make it happen though.



[ Parent ]

This sounds like a good idea .... (none / 0) (#70)
by omegadan on Tue Apr 23, 2002 at 11:49:37 PM EST

but I have this gnashing, uneasy feeling in my stomach about it.

First of all, I really hate the way we're throwing around the term Bayesian statistics, statistics is voodoo math, I know you've got a PHD that says it isn't, but it is and we both know it.

Secondly, if you accept statistics are capable of making predictions, they're going to make shitty predictions, and I realize shitty is a fairly loose term, so allow me to clarify. The problem is that "popular music" != "good music". Popular music has been trending downwards in terms of complexity, musicality and required talent for 50 years now.

I submit you may be able to predict what might be popular, but you'll never be able to tell me what is good.

I will admit Im taking this personally, I write music professionally for my company, <shameless plug> thehumbleguys.com, and my personal project, monkelectric.com - When I write music professionally, its catchy, simple, and sweet. When I write music for my own person it's complex, it's involved, and it's dirty. All things that aren't popular :-) What I see as an "independent artist" (its funny having this title, "independent artist", just 2 years ago I was "dude with recording studio in garage") is the artists saying, "I'm making great music, I don't know how to tell people about it." Solve that problem; you'll be a millionaire.

Nevertheless, I wish you all luck - and I'm going to try and prove myself wrong by signing up for your website when I'm ready to launch my monkelectric project.

Daniel "OmegaDan" Morrione [dan@morrione.net]
thehumbleguys music group - "Everything but love songs"
Scoring / Sound Design / Audio Programming
thehumbleguys.com [omegadan@thehumbleguys.com]
-
Monkelectric - Internationally ignored musicians.
monkelectric.com

Religion is a gateway psychosis. - Dave Foley

crapfuckers anonymous (1.00 / 1) (#71)
by christfokkar on Fri Apr 26, 2002 at 11:47:10 AM EST

BLAH. Another commercial site attempts to exploit fan contributions towards its own financial goals. What do fans get for their work? "EM Points." How marketroid. And we have to spend them in auctions? What the fuck?

Site design is weak and primitive. Install slashcode, you'll probably triple your functionality (and readability).

CEO is hype-ridden. What was that about Linux? That analogy went nowhere.

Finally, why do we need some corporate entity to provide a forum for discussing music? I think the proper Linux analogy is that we don't need you. Oh, you have music licenses and a RealAudio contract? I have MP3's dickwad.


Your search - vinyl did not match any recommendations. I think that sums up the whole site right there.


Go beat up your friends and take their gold necklaces. That's how niggers like you should be spending your time.

This looks a lot like FairShare (none / 0) (#72)
by termfin on Sun Apr 28, 2002 at 06:41:30 PM EST

This looks a lot like Ian Clarke's FairShare [proposal that he and a few others developed over the Summer of 2001.

I wonder whether that inspired Emergent Music...

The Technology Behind The Emergent Music Web Site | 73 comments (53 topical, 20 editorial, 0 hidden)
Display: Sort:

kuro5hin.org

[XML]
All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!