Slashdot Mirror
AI Bots Pick The Hits of Tomorrow
Wolverine Inspector writes "The Music Industry uses a product called HSS (Hit Song Science) made by Spain's Polyphonic HMI. According to The Guardian "while no one's talking about it, it seems that the whole record industry is already using AI to choose hits. From unsigned acts dreaming in their garage, to multinationals such as Sony and Universal, everyone is clandestinely using a new and controversial technology to gain an edge on their competitors."
Even though it costs about $5,200 US/$6,500, many artists are starting to buy it to help them write succesfull songs."
It would appear that the music industry is not ailing as much as they would like us to believe.
So instead of people blaming the HUGE record industry that produces crap, they can blame a machine! Sounds like a scapegoat to me. Either that, or the record execs are SO STUPID when it comes to music, that they have to get a machine to help them out.
Is it just me or did the article quote music industry folks as saying the software must work becuase 95% of the hits of the last decade scored highly. The software is a mathmatical model based on the hits of the last century.. so of course it scores them highly.
But in time, these predictions have always proved wrong.
The music industry has proven again and again that "time" no longer matters. Bands like The Stones, Aerosmith, etc, are all a thing of the past. They don't need them. They want acts like Spears, Maroon5, etc who rise to the top of the charts quickly through marketing, consolidation, and payoffs, and who are only there for a short time before the next big thing hits.
Touring, actual music playing, and actual singing are overrated. The HSS printout says so.
Just tweak this, this, and this. Add a synth here, here, and here. We have a hit. Two hits, maybe three, and we can continue to whine that we don't make any money because we spent it all marketing something that died after 3 years.
A subsection of the Music Department composes popular songs especially for the proles, using a special machine, rather than human artistic talent. An especially popular number at this time is "It was only a hopeless fancy," sung by the woman who hangs her washing outside Wilson and Julia's room
I remember a story from my BSc AI course, the US army wanted an AI program to detect tanks in satellite photos. Spent ages training it with countless pictures (with and without tanks).
In the end they had a system which had no idea what a tank was; their training photos with tanks were all taken in summer, those without in winter.
They _did_ have a kick-ass "have the trees got leaves on" detecting program, though.
What betting this system has been trained on "derivative shit", and will detect your quotient of derivation, rather than actually if the world will like your song?
I can't help but think this (and similar issues) is in some small way a lost opportunity for academia, which likes to pretend that popular music (i.e. music that people like to listen to) is somehow less valuable than "serious art music." Here, we have thousands of people who could be leading intelligent discourse on music, many of whom like popular music but won't dare say it because of an unwritten stigma that popular music is "low brow". Because of this, a potentially vocal, educated population that could be smacking RIAA execs upside the head now and again, or at least crying foul, instead relegates itself to the "classical" niche, often the "new music" sub-niche. Said people actually do speak out from time to time, but are so isolated by genre that they seem rarely to be noticed.
As one of said people, please excuse me while I return to my clarinet practice and writing my string quartet.
-- I prefer the term "karma escort."
Living near New York City, I consider myself lucky to have access to the New York Times Classical Music station. I am so sick of the garbage that has been produced in the last 10 years (Except Eminem, for some retarded reason I like his crap) that I barely ever change my radio tuner off 96.3
As for what I listen to at home and work... Ironically it's all old school stuff from Black Sabbath and Beasty Boys earlier music, plus.. more classical
I wonder if I'm the only guy who's so totally jaded to new music that I touch nothing new, period.
"It's not stealing if you don't get caught!"
I seem to remember a Piers Anthony series called the "Apprentice Adept" in which there was a game where you could play music. It was judged by a computer, so while your music may have sounded like utter crap to humans, it would be given a high score because it was "technically" right. This reminds me of that.
The music industry might turn itself around; they go ahead and do something this stupid. Music is art. It is not objective. It is not rational. It is not definable. It is not quantifiable.
This system will destroy popular music. It will define the elements of a "hit" song, then it will only determine that songs with those elements could possibly be hits. That ignores the history of music where what's a hit changes from year to year.
I listened to punk rock for decades. In the 80s songs by the bands All and 7 Seconds would never have been recognized by any system as being hits. But fast-forward a decade and suddenly artists like Blink 182 and Greenday ARE having hits using the same formula.
Basically, this system will stagnate the music industry as it will lock it into a very narrow form of music and it will not be allowed to grow. People will get even more bored which will lead to decreases sales.
If someone says he and his monkey have nothing to hide, they almost certainly do.
" Remember the good old days when the listeners picked the hits?"
When was this?
Music has been marketed to death for a very long time. Heck even in the 60s do you think the Beetles landing in New York to all that press coverage and screaming fans just happened?
Once a something crosses the line into big money it is all over. Look at Comic books, sf, Anime is on it's way to being market driven.
Even computer software is now more driven by fluff than substance. BSD is every bit as good as Linux but it does not get the "mindshare" that Linux does. The Amiga, Atari ST, and Apple Mac where much better systems than MS-DOS and the 68k was a much better chip than the 8088 and 80286 but PCs won because of marketing
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
In many areas of the US, we're seeing a rise in the demand for organic, non-trans-fatty, less-processed foods (e.g., Whole Foods). Actually, it's more acurate to say we're seeing a rise in the supply. The rise in demand necessarily preceded this rise in supply.
Similarly, if too many musicians over-process their music, we will see an increased demand for more "organic" music that will evenutally lead to an increased supply. The end result might even be better music.
Ben Hocking
Need a professional organizer?
That argument has merit until you follow it out to the extremes. Is a pen cap art? What about each individual Cheerio? How about a pile of dog crap? It's all been produced, but the quality, however, is, at best, mundane, and at worse...feces.
There comes a point where you have to draw a line - and I admit that it is a grey area, and up for interpretation - as to what art may be. Saying everything is art completely removes whatever meaning the word may have.
The article says: ". . . [an] A&R director at EMI believes that HSS as a hit predictor merely reinforces decisions taken by A&Rs, those record company employees given the job of discovering new songs and artists. "A good A&R has a very accurate instinct for what the market needs," he says - and the fact that 95% of hit songs in the past 50 years are high scorers seems to back him up."
Um, HSS is using past hit songs to define high scores, so the fact that past hits have high scores is not some sort of vindication of the job these mysterious A&R guys have been doing. The real question is why that figure isn't 100% - I'm guessing this is probably because the clusters are fairly wide, so some songs manage to be far enough from the algorithm's definition of the cluster to be classed as non-hits, despite being part of the training set?
If??? IF??? My god, have you heard pop music nowadays? I can't tell them apart because they've become so formulaic. A lot of modern groups sound like they must be a dime a dozen; or at least carefully compiled to match some already known formula.
As to wether an increase in demand for innovators happens, it's happening on a small scale. It's just only a small percentage of people who go looking outside the mainstream.
I personally don't listen to music most people have even heard off -- I just don't expect that I make a dent in the total sales.
Lost at C:>. Found at C.
Like LAUNCHCast before the RIAA leaned on them, and then Yahoo! acquired it and ripped out everything that made it innovative and cool?
You'd rate songs on a scale from 'never play this again' to 10, and the system would select new songs based on what you'd already rated and insert them into your personal "station" rotation.
But the real killer feature was that you could search for other users whose tastes were statistically similar to your own, subscribe to their stations, and learn about new and different music and artists as some of their favorites were added to your rotation. Want to buy a song? Click on it.
Absolutely cool collaborative software. Unfortunately, if you wanted to expend the effort, you could abuse the system to constuct a station that could (gasp!) play a specific song at a specific time for free, and the RIAA wouldn't allow that.
So the only thing that had gotten me to purchase any new music in years was eviscerated, stuffed full of ads, and then sold to Yahoo! as a 'service' with all the collaboration gone. You could pay money to lose the ads. Whee.
Bitter? Me?
They can cluster Backstreet Boys and Britney Spears all they want. Just please, pretty please, don't kill any newborn Pink Floyds or Deep Purples with some junky software. This would make me fold little paper boats from my IT degree, that's for sure.
I am putting myself to the fullest possible use, which is all I can think that any conscious entity can ever hope to do.
Just out of curiosity, I'd like to see how some of my favourite songs score on such a system. I have a hard time believing that 'Echoes', 'Shine on..', or other great music fits their calcuations very well.
TZ
I don't think DJ's at radio stations are free to play anything that isn't on the approved play list. I submit the following anectdote as my only evidence. Here in Phoenix there was (and still is, though in a new form) a modern rock radio station called The Edge 106.3. The station was changed to a Spanish language format, but it was known that this was going to happen long before it actually did. The existing DJ's knew they weren't going to be around after the change. The afternoon drive DJ (I believe called Dead Air Dave) started playing the "What are they going to do, fire me?" song of the day every day at 5:15. It was basically just some song he likes that wasn't on the approved play list. Sometimes it would be a deep cut from an album they already played, sometimes a new local band he wanted to give some pub to, or something he heard on the internet and liked, or some old school band that didn't get played any more. Now I'm not one who usually listens to the radio much, I've got too much money in cds and mp3s. However, I always tuned in to hear the "What are they going to do, fire me" song of the day. Why? Because it was different! I was guaranteed to hear a decent song, that I probably hadn't heard before (or at least in a long time). I wouldn't always like the songs he picked, but at least it was something new. Kinda like eating a new dish at a favorite resturant. And I always *learned* something, becuase he'd explain why it was today's song. The ironic part is that after the station closed most of the staff reformed The Edge as a truly independant radio station on another channel. That DJ continued the "What are they going to do" bit for a few months after. They don't do that any longer, but the station itself is pretty decent. The even stream their broadcast on the Web, which few stations seems to do these days. You can read about the station and find the stream at: www.theedge1039.com
The boomers are all aware of this situation where the music from 40 years ago is being constantly pumped into public spaces and how much it annoys people (like younger people of so-called 'Gen X') who don't share a cultural identification with these recordings. Mostly though, they don't care if only because they happen to like this music.
The blame is not on the 'boomers' themselves but instead on the music industry. These 40-year-old recordings are the cheapest and most cost-effective way to fill public space with background music. Every time one of these recordings gets played in public, someone, somewhere gets paid off. Every time.
The only way to make these recordings disappear from public space is to change the financial framework of the music 'publishing' industry, which determines who gets the money whenever these recordings are played in public space. But that's simply not going to happen without forcing the disintegration of the music industry. One more argument to 'pirate' music recordings and swap music files without money transfers.
'Blaming the boomers' is too easy because even if all the boomers were to disappear tomorrow nothing that they are assumed to control would actually change. Boomers are just filling slots in a 'system of power' that itself needs to be changed for your life to get better.
A short story by Arthur C. Clarke describes one possible consequence of this sort of thing. The scientist involved builds a computer to study the underlying theory of music, harmonic relations, wave analysis, frequency distribution, etc. and how it interacts with the brain on a physiological level. His search is related to the notion that all existing tunes are crude approximations of the fundamental melody that has eluded composers for centuries (basically a rehash of Plato's theory of ideals applied to music.) The scientist is later found in a permanent catatonic state in his lab (by his tone-deaf assistant) with the Ultimate Melody repeating over and over in an endless loop. Because the overwhelming power of the Ultimate Melody (the ideal form on which all melodies in the universe are patterned after), his mind is completely dominated by it--much the same as when a catchy tune gets stuck in your head for days, only much more powerful. The melody formed a fugue in the pathways of his brain, going round and round forever, obliterating all other thoughts.
Everyone seems to be saying (as was my first reaction when I read the story) that this will lead to everything sounding the same, being bland, etc - as if that wasn't already the case. However, I believe there is cause for optimism - because when something good comes along that really doesn't fit the "hit box" it will stand out so much above the background mush of the rest that it will be worth taking notice of. When I was growing up mainstream music seemed to be a lot more diverse, and you had to pay close attention to really keep up with what was going on. It was hard work (but usually rewarding) to sort the good from the bad. Now all you need to do is keep the radio on but turned down low so you don't actually have to listen to it, but loud enough so that when something interesting does get played, your brain suddenly wakes up and notices it. Thus it becomes much easier than it used to be to pick out interesting stuff. Thanks, lazy pigopolist music industry-type guys!
:) entering at number one was virtually unheard of - Slade's Merry Christmas Everybody did it in 1973, the next one to do so was about 5 years later! And back then you needed to sell hundreds of thousands if not millions of records to make No. 1. So basically the music industry has ruined what used to be a useful indicator of popular taste (within limits) into something that isn't even a useful indicator of how successful their marketing is, except in pure binary terms (number 1 = did OK-ish, not number 1 = flop). Basically the chart has been quantised down into fewer and fewer bits. I say it's time it was officially abandoned altogether, though those of us with any musical sensibility personally abandoned it some time in the early 1980s.
If you think I'm joking, consider this. The UK has just now "celebrated" the 1000th number one record in the charts. The track in question is Elvis Presley's tune One Night from about 2000 B.C. Last week's number one (the 999th) was Elvis Presley's Jailhouse Rock. Hrrrmm... could there be a marketing campaign around promoting Elvis records? Perhaps to help flll up the special "limited edition" (only 500,000 issues!) box sets of Elvis's Greatest Hits that were flogged off the other week, a bargain of an empty carboard box for only 10.99GBP. Marketing genius really, get the punters to stump up for an empty box, then get them to fork out 3 quid a week for fifty weeks to fill it! (Elvis fans - just say no!)
Every number one nowadays comes IN at number one, because of hyping and marketing techniques. But the 1000th number 1 needed only 29,000 sales to make it there. Of the last 530-odd number ones, all but 2 entered at number one. This makes the chart meaningless. Back in my day