Or, sure, you can also terminate your record deal. Hope you have 500 grands around just for that.
Frankly I don't see it ending much better for visual artists.
I think the modern world has become too complacent in terms of labor organization - the time of plenty left a lot of people content to take whatever was given to them because there was such a glut of excess that it was freely shared. That sharing is coming to an end and we're returning to a time when we need to demand fair and equitable treatment.
It’s even worse than you say — it was murder on digital retail too, right at the time when it was on track to compete with or exceed old physical sales.
Spotify adopted the economics of piracy and stamped them with the false veneer of legitimacy.
> Spotify adopted the economics of piracy and stamped them with the false veneer of legitimacy.
As a side note, in the beginning Spotify used pirated music off The Piratebay without asking for permission from the copyright holders.
There are lots of other examples of this happening too... I believe some of the early nintendo retro releases were emulators running pirated roms
If anything, that feels even worse.
> I believe some of the early nintendo retro releases were emulators running pirated roms
If Nintendo has a licence for the game that the ROM was an unlicensed pirate of, while that's weird, it doesn't seem fishy in the same way.
Like, every nerd from the era has an mp3 collection. Mine is literally the only data that I have that's been around since I was 15 and survived multiple HD crashes.
How are you going to get the streaming business up and running without some seed data?
Are you also mad at Uber and Lyft?
Yes, and much of it is pirated. We all know that.
> How are you going to get the streaming business up and running without some seed data?
Pay money for streaming rights, probably. You're suggesting the only way to start a streaming service is to do so illegally?
> Are you also mad at Uber and Lyft?
Yes
The music industry was not looking to break the stranglehold they had on CD sales. Someone had to come in with a 'shoot first ask questions later' attitude to get to where we are today.
Uber and Lyft did what they did for the same reason -- the (oftentimes mafia-backed) taxi cartels had a monopoly on pricing and taxi medallions and the only realistic way to break that was to operate illegally.
I think you will find that the law (and copyright) to be extremely overrated. Copyright, in particular, should not exist in its current form, especially with digital data that is not bound by the laws of physics or physicality, to say nothing of the various entities which have carved out most of the royalties that an artist can make for themselves.
And in doing so, each has failed to realize better solutions for people. Uber & Lyft have increased our dependence on wasteful, dangerous, expensive personal vehicles. Spotify is a worse model than Bandcamp for indie artists. It's also a worse model than pretty much every other streaming service in that it pays so little to artists. So, yeah I don't like it when rich people break the law to get more rich at society's expense. We should be breaking the law to make society better.
And I would say Spotify is doing a social good. As a nobody artist, being able to point people to stuff on Spotify is really compelling. Any other way I'd have to force people into either buying or using an app they are unfamiliar with just to listen to it. That said I am not trying to make money from music, that is just foolish.
Well, I understand where you're coming from here, but that's not really Spotify doing a social good, it's just the inevitable effect of their cultural dominance. Everybody[1] has Spotify. But they've taken that position due to their unethical growth plan.
Anyway, I've realized that we're discussing multiple things, not one thing. Spotify can be bad for artists who want to make money off their art, while being good for those who don't want/need to make money from music.
[1] I don't have Spotify.
What Spotify actually did was cannibalize digital retail. Because of course it did: it used massive capitalization AND ignoring pesky laws/policies about actually compensating people creating the work Spotify's service depends on to give that work away to consumers for essentially free, until they'd created that expectation in the market and had enough pull as a channel to buddy up with labels (and the ironic thing is it's not even clear how profitable they can be, maybe leading the charge to the bottom has downsides).
> Copyright, in particular, should not exist in its current form, especially with digital data that is not bound by the laws of physics or physicality
Copyright conventions become more important in the face of falling barriers to reproduction and transmission, not less. That's how they got created in the first place: physical copying got industrialized and could take place with scale and ease that was unheard of before then. And the rationales behind them are still solid today, because they're about the economics of creation, not distribution.
If copyright (and DRM) was enforced the way some wanted, we wouldn't have a rich history of remixes, recontextualizations, or bootlegs that allow a work to live on far past its shelf-life. Memes would be a shallow husk of what they are today. We wouldn't have many, many genres of music, and things like the Amen break wouldn't exist and we'd all be worse off for it.
For that reason copyright needs to be abolished. There are other ways of monetizing intellectual property. Once a given work is released to the public, its creators have literally zero control over it, despite the technological artifices we construct otherwise.
As if nobody has to actually create the first instance of a given work, it's just copies all the way down.
That's why the economics of distribution/reproduction are distinct from the economics of creation. Maybe it's clearer if we use the term invention. The major input is time, time to develop the faculties needed to invent/create -- often years if not decades -- and then the time needed to put into a specific work.
The legal claims/controls on distribution give inventors leverage they can use to get better returns on that time, which provide better incentives.
> Copyright (and DRM) was enforced the way some wanted, we wouldn't have a rich history of [blah blah blah]
Copyright was enforced for literally centuries and for much of that period we got a rich history of works which borrowed in a dozen ways from other works, because actual copyright law has both boundaries and blessed borrowing.
I swear, so many tech folks got half a narrative in their heads about draconian DRM as digital gulags and lost their minds to a manichaean all-or-nothing view on the topic.
Yes yes the evil suits from 1999 probably wanted to super-glue your 1/8" output and lock you out of control of your machine. Again, almost 20 years ago the basic truce on that battle was defined. There's always going to be some activity that can't be controlled, but that doesn't mean you can't define and encourage legitimate activity, so we keep the basic bargain because giving inventors/creators a say in how/where/pricing for their work is both helpful and decent, but we don't try suing individuals over their uh "freely-sourced" media collection or install policing malware.
This isn't digital hitler on every last device vs total free-for-all. It's saying Amazon can't sell your ebook or music without compensating you. And maybe even that Spotify doesn't have the right to give away tracks for free or a pittance, that buffet streaming is close enough to ownership that it calls for artist payouts that are closer to the scale of retail than broadcast. You don't need total control to get there, and we have the levels of control to make this happen.
> Memes would be a shallow husk of what they are today.
Oh no not the memes, known for their fullness and depth.
But also no, not the memes, since it'd be vanishingly unusual that any of them would end up in court let alone to be found to violate copyright law.
> There are other ways of monetizing
It's a well-known fact about creative paths that they're not littered with the luxury of money-making opportunities. There are a few ways of doing it, but they're legs of a stool, and the number of inventors/creators who have the stool up to even 3 legs is not large. Suggesting they give just one little leg knocks over a lot of stools, and for what?
A great filter, for one, that eliminates those that do art for money in favor of those that do art for the sake of doing art. Personally I don't want output from the money-motivated. I want art for art's sake. The made-for-money stuff is bland and lowest-common-denominator. It is essentially trash.
Second, it's pretty iffy that only low-quality succeeds economically. Sure, everyone can think of examples that somehow succeed with limited merit, but you can't sustain the thesis that well-rewarded work is mediocre without ignoring a lot of strong yet widely appreciated and profitable material.
But even those two big points are minor compared to the most important one:
Everyone needs money. Even artists who do what they do for love. It's the legs of their stools you're suggesting "filtering"/kicking out too.
When someone can't earn money doing what they love, they have to spend time doing other things in order to get the money. And that's time they're not creating art and time they are not refining their craft.
What you're "filtering" out is the peak of the skill they could have developed with more time as well as the art they could have created with it. Maybe even the attention and focus they have to doing it at all, hijacked by all the ways necessity can preclude even love.
And again, for what?
Anecdotally, I had a friend at Microsoft hook me up with discount Windows OEM licenses for my PC builds and it seemed similarly easy to get licenses.
As more people are able to produce music (due to cheaper tools like DAWs, more accessible music theory education, etc etc), if the demand of music doesn't grow proportionally, the average income of musicians/songwriters would decline.
The above will happen regardless of Spotify's existence. Thus, Spotify doesn't matter (much).
You're just going to end up with a bunch of sloppy tables.
People still want to listen to quality music from artists who have years of practice and experience. You can't reliably get years of experience unless you're getting paid to do it.
Sure, there are exceptions, but it's not the rule. Michael Jackson would not have existed if there was no money in the career. The money is why his father pushed so (insanely) hard.
The counter argument is trash music will just be the norm. And maybe for a while that would happen, but eventually we'll see someone (similar to the private search engines we see today) come out with a new platform with the selling point that artists get a living wage -- as long as the people demand it, and I believe they will.
Uh... and it's true? If the price of circular saws drop in price, and the demand for hand-made furniture doesn't change, then they'll become cheaper. How much cheaper is another question, as circular saws are already very cheap today, compared to hand-made furniture.
So yeah, you're right, it's just like saying that.
> if there was no money in the career
It's unlikely to decline indefinitely. Piracy, Spotify, more youtube channel teaching how to make music... all these didn't prevent Billie Eilish from becoming a star.
I don't know whether circular saws are likely to be the biggest input into a piece of handmade furniture, but my guess is they're not and it's time invested in the specific work and practicing the art rather than industrial capital.
Similarly, while DAWs while can (though not necessarily do) reduce capital necessary to do certain aspects of production, they don't represent the most significant investment into writing music. Also time, both in the creation of the specific work AND in terms of time practicing the art.
> It's unlikely to decline indefinitely. Piracy, Spotify, more youtube channel teaching how to make music... all these didn't prevent Billie Eilish from becoming a star.
Survivorship bias. Billie Eilish or any other individual success are no more an indication that all is well with the status quo than blue zone anecdotes are promises anyone who chooses can be a centenarian.
>You're just going to end up with a bunch of sloppy tables.
Well, yes, and that's how IKEA and mass production in general made many people that would be making furniture out of the job.
Even in tailor-made stuff good cheap tools does make work of skilled maker far quicker. And you can get more people trying to get into that if the tools are cheap.
Hardware is cheap, software is free/near free so there is far more people trying, when you no longer need to spend small car worth of money just to say play electronic music
> People still want to listen to quality music from artists who have years of practice and experience. You can't reliably get years of experience unless you're getting paid to do it.
Most musicians got that by playing in garage bands and doing concerts.
And many of them did it entirely for free, out of passion, till they were good enough, far before fancy computers were in everyone's pockets.
> The counter argument is trash music will just be the norm.
It is the norm far before Spotify happened I'm afraid
Actually I think something has happened to the textiles industry whereby demand must have driven a certain band of suppliers out of business, and now try as I could I can't get polo shirts in the same think quality cotton weave I could 30 years ago. There is probably some niche source possibly online but I don't know how to discover it; the standard "throw money at luxury mall brand" route seems to not work any longer as the brick and mortars have watered down their materials as well. Sic transit gloria mundi
I like wearing industrial clothing (like red kap cotton work shirts) and to my eye seem like they're made about the same quality they always were.
I think the existence of Micheal Jackson is quite tragic, so I don't think it's a good thing that a system tortured him into being a famous musician
Of course, this isn't new. Tim O'Reilly said something similar in the context of book publishing probably getting on to 20 years ago at this point.
What happened instead was that Spotify led the pricing change by taking capital, cheating policy, and producing a consumption avenue that cut the price by orders of magnitude.
And meanwhile:
> cheaper tools like DAWs, more accessible music theory education
The gains in education are fractional. The library or a neighborhood piano teacher were good enough resource wise. YouTube eliminates the trip (and the funny thing is that we're iffy on even rewarding those people proportionally), but isn't a new opportunity.
And even for materials that are better in the way that 3Blue1Brown is for math... just like you're going to have to sit down and spend a lot of time actually doing problems rather than just watching the videos if you have any hope of really getting it, the constraint when it comes to producing music is still sitting down and putting in the time, not only on the specific problem/work in front of you but in the background to do it elegantly.
DAWs are great and can make up for some margin of missing virtuosity, but you have to put in the time practicing using them too -- they become their own instrument.
The constraint on making music has always been time. And what gets you more time to do something? Either having another source of wealth, or getting economically rewarded for doing that thing.
Spotify and the damage it's done the market absolutely matters. Just because music is getting through the damage doesn't mean there wasn't some lost, and not just quantity, level that could have been leveraged to through the magic of compounding focus. Anybody who's read Graham's "maker schedule/manager scheduler" should already know this.
Why do I care if spotify's investors have replace the record label investors?
But let's say for the sake of argument that's how it worked before the internet. If so, why did we let that level of disruption just replace one set of bad guys with another?
What we should have had instead was what we were on track to have: multiple-scale digital recording retail, Apple to Bandcamp to individual artist site to local indie collective as point of sale. Charge whatever artist and buyer choose and can clear a transaction at, they keep 70-90%. True streaming, some with a format, some algorithmically customized for both audio wallpaper and music discovery... but in every case not user-programmed because that's how you justify the difference between the ridiculous fractional cent payouts and recordings.
That's the world we could have tomorrow if the policy was there.
More likely, everybody's too used to a decade of having the privilege of an unlimited basically free recording buffet that Spotify used to cannibalize the industry and we won't do it no matter how it erodes the economics of creation. But we could.
Also, like everything in the universe, they are subject to supply vs demand.
And fundamentally the supply exceeds the demand.
The problem now is that we have so much content (music, books, movies, short vids, long vids, etc...), and not enough aggregate time to consume it all.
If you can't hear the difference, see a doctor.
You can apply this on professional filmmaking or vlogging. I guess amount of time consumed audiovisual production today is much higher on amateurish production thanks to antisocial networks.
Nowadays most valuable is attention. Cheap stimuli is easier to consume. That's what technology teach us.
But a 2k$ guitar is certainly better than a 50$ guitar. Not only in how it sounds but in how easy it is to play it.
My 1st guitar was bad so I couldn't do barrè chords. I thought I was bad and pros could do it. Turned out pros just had better guitars.
Better guitars also have less noise, better cables are shielded.
Yes, more expensive is better (up to a point).
It's not just about the sound.
Even when you go into hardware you can still get plenty for cheap.
Good that you changed your mind.
You could replace most of this category with a Markov chain bouncing up and down a simple key without most people even thinking about it, and I know because this is exactly how I made music for my shareware video games a decade ago.
Did your video game sell as much as outcast? A game with a proper music score.
https://en.wikipedia.org/wiki/Outcast_(video_game)
Does your game have a wikipedia entry?
Could I assume that people enjoyed outcast more than your hobby game?
Sure, but they'll also watch daily soap operas, and the meme "I showed my favourite film to a loved one, but they paid no attention" predates multi-screening.
> Did your video game sell as much as outcast? A game with a proper music score.
Even in aggregate over all the games: I wish :P
But the real question is: how much of that was the music?
Even now, were I to redo that period of my life (and so not need to caveat markets changing), the music isn't what I'd focus on changing — shareware was already a bad idea though I didn't realise it, MacOS shareware written in Java just as MacOS got its own (ObjC only) app store moreso.
That actually makes Spotify worse, because they could have offered that product instead of using huge sums of capital to reshape the expectation anybody with a device is entitled to listen to any work on demand for free.
I guess the good news is that it wouldn't cause a fuss if someone were to change policy so that you can't pay out buffet streaming like it's digital radio and people ended up having to buy songs or at least do the honest work of piracy if they won't accept the app directing programming. After all, most people are just as happy listening to a Markov chain generate bloops for free.
It's about keeping an entire industry, live or recorded, and their milieu alive.
The truth is that what happened wasn't a liberation. It was a methodical purge of the medium-sized side of the music industry. Now we're reaching the point of having 5-6 industry giants taking all the money plus...yes, an inordinate amount of people making mostly self-referential music in their own bedroom on weekends, music that will reach no-one outside whatever local scene they hang around. But most of them were making music even before, and were by their own choice irrelevant to the industry. (True, now they can also become influencers on Twitch and maybe one out of thousands can make a living by streaming their life 24h/day. One ticket for the lottery, please). Whoever was between them and the majors is being squeezed out of the game.
Humanity has always been constrained by the 24 hour a day thing, but the economy has grown nonetheless.
That last little bit is interesting. Back in the physical media era, if Artist B fell out of your rotation, you could sell your record/tape/CD and decrease the size of their new market a little bit. Then we went to DRM, and every song you bought was a sunk cost; if you didn't listen to it, you still payed. Now with streaming, it's back to the downsides of physical media; if you stop listening to Artist B, they stop getting paid.
There was already "enough" recorded music decades ago that someone could fill their living ours with constant listening and not exhaust it. If that's really all there is to it, I'm sure you'll have no problem committing to never listen to anything after around 1970. If you have any hesitation about that then you might start to see serious shortcomings in this conception of supply.
"there are more people able to make and publish music then ever" also papers over nearly everything that matters about the statement. "There are more people" is the defensible part. There's half an argument we have greater access to affordable digital tools for production than ever -- but I'm not even sure it's half. The constraining factor on music composition, performance/recording, production is always time. Even where the tools themselves save time that's a series of converging terms that stops at a limit because to make good music you have to practice using those tools plus others. A lot.
Set up a system which rewards those people in proportion to the audience they find and those people are both equipped and incentivized to spend more of their time into making not only more but making better, because they aren't required to spend their time doing other things.
Set up a system which says "Oh yeah, we shouldn't value any of this, people should just do it in their spare time" and sure, some will do it in their spare time. But they'll miss out on the compounding effects of focus and its power laws because they're occupied with whatever other stuff policy+markets have been set up to value instead. And their audience and the rest of the world will miss out on their power law peaks.
Which is why I was too generous with my earlier "never listen to anything after around 1970" thought experiment. Really, don't listen to anything except debut releases before 1970. Some of the debuts are really good, of course, and labors of love (or capital-backed love) as you say. But the post-debut work is what's enabled by the economic feedback.
There will, of course, likely often be survivors to bias ourselves to the status quo with. And perhaps that's good enough for some. Hell, maybe we're even rapidly getting to a point where we don't even need most artists at all, we can simply have software trained on all the work of all the artists that have ever recorded produce it for you, and be done with not only the pesky idea of rewarding musicians whose work we appreciate but having a pesky human being involved in direct production in the first place.
Well, I'd say yes because that's what the conversation is about.
Seems like a natural iteration in the ordering of complex systems. Beyond legal regulations it would be great to start to think about new solutions, if they ever exist.
AI will also have an additional effect: it will be isolating in the sense that the need for other humans will decrease.
These two points alone, strengthened by many others, have led me to conclude that the world is MUCH better off with AI and that tech companies are ruining the world with their abominations.
There are many machines replacing hundreds or even thousands of people- farm equipment, trains, tunnel boring machines etc.
With software you could say chip makers, developers, and energy companies will get stronger but I don't think there's a comparison. The keyholders will be a much smaller group with a greater power if we stay onboard the AI train.
Sure, and even taking that into account we produce far more food with far less human labor than we did 100 years ago, and that's a good thing.
If somebody with zero skill in the arts can produce output of similar quality as a craftsman and about a thousand times faster, what is the point of art anymore? Sure, one can enjoy the very act of creating art, but we can't deny that art has value in relation to an audience, and is also a display of skill and a source of pride.
What if AI generates the perfect music just for you, based on your taste? Here we lose any and all social/cultural aspects of music. There's no point in discussing music as we have no shared experience. There's no point in emphasizing your favorite song because everybody exclusively listens to favorite songs.
What if you need to write a long essay and use AI to help write it. I receive it and use AI to summarize it. Other than this interaction being supremely depressing, what is the point of it at all? Just submit it to the big machine and perhaps some of it will show up in my use of ChatGPT-17.
So you were faster to write something whilst I was faster to consume it. This allows the both of us to do more in a single day. This "big win" won't gives us back free time nor raise our wages though. I just means that the nature of the work is for us to take the job of being guard rails for AI, a soul crushing job in itself but also temporary, until the rails are no longer needed.
I like to think "AI" will make art better reflect its real value, devoid of the tangential flat costs associated with housing, clothing, and feeding humans in the process of producing art.
The consumers at large demand driving down the cost for consuming and enjoying art, and raise hellfire if there is so much as a suggestion of raising that cost. Remember how much controversy there was and still is about raising the standard price of video games from $60 USD to $70 USD? And that $60 USD today is pennies compared to $60 back in, say, 1995.
If the consumers at large demand the cost of art to go down and "AI" will make the process of producing that art better reflect that real value, isn't this overall a good thing insofar as making the price tag more clear and agreeable and closing down sweatshops?
I do agree with you in large part. I think I’m just slightly more optimistic that people who are driven to create will continue to do so and that people who really want real and human experiences and interactions will be able to find them with effort. Probably not anywhere on the mainstream internet though. Maybe even only in person.
AI is ok, but it will never ....
until it does, and then the goalposts will move.
the only logical endgame that I can see is AI replacing all human endeavour (creative, technical, physical, mental).
There will eventually be philosophers trying to make sense of the profound understandings coming from machines.
That consideration seems more along the lines of worrying about the eventual need to escape earth than a future on a closer horizon worth worrying about.
I’m more concerned about how every facet of our children’s lives will become inundated with shoddy ai being used to extract maximum profits at the cost of any humanness, and the death of all genuine communication on the internet.
Thinking that the progression is going to slow down is just wishful thinking.
But there still is no real creativity. No emerging concepts aside from complete accidents that cannot be replicated again. The same is true for the other direction with models like Clip could create an interpretation of generated images. It is impressive, but there are still clear limitations. You cannot expect linear growth here, it could be that the current AI approaches are wrong, we hit a plateau and need fully new approaches for significant improvements. What we now have is insane amount of data and more powerful hardware, it could be that we have years of iterative and slow improvement while people fine tune their models.
I think LLM have the same problem overall, it is just more difficult to notice.
Do you mean "world is MUCH better off without AI."
What you wrote doesn't make much sense withing the context of your comment, but I have to ask because there are some software engineers that find abominations appealing for some reason, or just lack the ability to tell the difference between desirable technology and a technological abomination. I think a big component of the latter is many software engineers' overconfidence in their abilities that makes them easy marks, and the willingness of many kinds of hype men to exploit that to con them with propaganda.
Taxes are competitive. If you have extremely high taxes, then the businesses that can move out will move out, leaving a smaller tax base and requiring you to raise taxes even more in a death spiral.
It's worth repeating that doing nothing increases overall societal costs if large parts of the population become unemployed, whether due to lower statewide income taxes, increased welfare costs, people resorting to grey collar work, and other secondary effects from increased displacement (lower consumer spending, rentals sitting empty, health epidemics, increased crime, etc.). UBI does not have to be a blanket instrument either: rather than income, we can focus on making certain goods and services universally available, such as access to food surpluses that would otherwise be overturned or basic internet access to enable people to remain connected without expensive contracts.
A solution may even exist beyond taxation: making reschooling and job pivoting more accepted within industry, lowering admission costs to tertiary education, or guaranteeing placement of employees when let go on account of automation or cost-cutting. What way the pendulum will swing remains to be seen.
They don't get to keep the tax revenue you know.
Fractional reserve banking is pretty much an urban myth. Banks create money when they make commercial loans. The Bank of England explains it quite nicely here:
https://www.bankofengland.co.uk/-/media/boe/files/quarterly-...
Socialize the essentials, let people work for the non-essentials.
If there isn't enough work to go around for people who want more than a substistence living, start reducing the definition of "full-time" until there is. If only 50% of working aged people can find work, redefine full-time as 24 hours/week
There won't be UBI, period. Though I could see a future where obsolete people are warehoused in sex-segregated poor houses until they die out, if it's determined that their freedom is threat to stability.
I just don't see it. They will need a job, they just won't be able to find one. It is not a utopia, it is a disaster.
BUT, sometimes I want something that will automate the fudge out of my PC (imagine command prompt on overdrive). I usually DDG for the solution and end up in some 10yo solution in StackExchange, which doesn't do the thing.
My friends have all forgotten their DOS skills.. so I turn to ChatGPT and boom! I get me 2 paragraphs script in 30secs.
Do I hire devs? Hell yeah and we pay well, and we will continue to do so for many years. Do I use ChatGPT for the small (personal) stuff? Hell yeah too.
Now, if a company wants to outsource everything to an LLM/AI then I wish them the best of luck, coz when something will break (and oh IT WILL), Tthe contractor they screwed over should charge them x50!!!!
I have witnessed this firsthand when I dove into the deep end on something over my head, GPT-4 Code Interpreter went into an error loop and I had to learn all of the background knowledge I was foolishly trying to avoid.
> Tthe contractor they screwed over should charge them x50!!!!
IT already does this after they were outsourced. They build up IT companies that take at least 3 times as much for consultation and you still need to employ local IT that actually implements the solutions. And their wage also doubled as well.
more like hundreds of millions
> AI will also have an additional effect: it will be isolating in the sense that the need for other humans will decrease.
unless there's a complete restructuring of our society then a repeat of the late 18th century seems to be the likely outcome
with their stake in society gone: the peasant class get fed up of eating dirt and storm the bastille
(I really, really hope the AI revolution turns out to be just hype)
One thing I wonder about a repeat of history is if the lowest classes still get enough of a share of the increased output in income/QoL would there still be a revolt about the increasing wealth concentration?
I'm not saying that I agree with this approach.
I've been watching videos of Guy Michelmore in youtube. Not because I will ever write any orchestral music, but because I like his energy and envy his shed. Would I bother if Guy Michelmore were an AI?
It could also have some interesting avenues, like feeding some variables to the AI from say a video game (number and type of monsters on screen, mood etc.) to generate music reacting to what is happening on screen
That is not what AI would ever generate.
TBH "Fly or Die" was way more common on the US side of the industry. And even in the USA by the late '90s to the end of the 2010 it was somewhat doable if you were skilled enough to make a living solo (we're talking 60-80K/year max) as a "jobber" opening for bigger acts on local venues.
Like, the entire NYC indie scene got a start from this premise. If you get a chance, give a look to "Meet Me in the Bathroom"[1], which is a documentary specifically of this timeframe.
[1]https://www.youtube.com/watch?v=n71c1Szjv08&themeRefresh=1
I think Gen AI will commoditize the mundane and "typical", and heavily push people into creating something extraordinarily unique. I think there is the same pressure even without AI, when as a creator you have to standout amongst the sea of people vying for people's attention.
I believe GenAI can be useful in a way too. For e.g. If I'm an artist looking for inspiration, I can have a GenAI tool create some "random" works that I can get inspired from.
If the court rulings hold and AI works cannot be copyrighted then us end users do not have to pay for it either... but that seems like a race to the bottom. Like the end of a craft. Why would anyone create art if it has no/minimal downstream value?
Artists need to band together in some sort of union or not agree to do art with that AI clause or perhaps only do art with a no-AI use clause. And have an allowed AI-clause that is prohibitively expensive (like in the multi hundred millions per piece). That way 'accidents that happen' have a prescribed recovery amount plus other requirements like pulling the generated artwork. "Hey, we understand it may have been accidental, but here is the bill."
First, AI generated art is random and disposable. Yes, you'll get a great image that you can use once, but then what? You can't build a campaign on it.
Second, AI generated art can't be copyrighted, so knockoff competitors are free to use your AI-generated marketing images.
At the very least, you can seed the AI with a paid graphic artist's work (seed-based AI images can be copyrighted). But that artist will do it better than your unpaid intern.
The visual entertainment "supply" is not limited by the current state of tools. It's always limited by the skills of the top crop. Professionals are always ahead and hard to come by. The industry's self-regulating mechanism is novelty; what is abundant becomes fundamentally uninteresting and dies.
In graphic arts, the overlap between people with technical proficiency and vision/taste is probably quite high, but it's not one-to-one. There are people with excellent taste who can identify great art or design when they see it, and who can perhaps imagine incredible masterpieces in their minds, but cannot draw a convincing stick figure. On the other side, there are people who can expertly make someone else's concept real, but can't come up with a compelling concept themselves. AI will be great for the former, and bad for the latter (or at least force the latter to adapt).
Whether this will have the effect of concentrating wealth or distributing it more widely strikes me as a very difficult question. It may be devastating for certain professions, but could also enable a whole new class of entrepreneurs. I could see it going either way, or the two effects may cancel each other out and economic equality stays about where it is. We're in the realm of complex systems here, so I wouldn't put much stock in anyone's prediction.
The problem is that an artist still needs to eat in the 10-20 years it takes to develop "creativity and taste".
What AI will do/is doing is knock out the entry-level jobs. If you can't train humans on the entry-level, you will eventually have no experienced people.
The ai generation of artists will grow up with ai tools
There is no subtitute for drawing a face to be good at drawing faces. No AI will ever help that.
I think it's a different kind of talent, and not automatically a lower bar. The key to being a professional artist is being able to offer variants based on given direction. Either way, it's much much more than pushing a button or holding a lever in place for a period of time.
No. First off trademarks exist and they found that work done solely by the machine couldn't be treated as a work for hire copyrighted by the machine and assigned to the operator. There is no reason to believe that work couldn't be treated directly as copyrighted by the human operator who has creative input nor is the matter with the images used to train the model truly settled.
>First, AI generated art is random and disposable. Yes, you'll get a great image that you can use once, but then what? You can't build a campaign on it.
You can already get variations on a them and text driven modification eg make the blank a blank or make the blank blanker.
Random variations aren't interesting, they just make something abundant even more abundant and secondary. Unless you have a model with sufficient intelligence that can create something conceptually original (at which point we're all fucked, not just artists or programmers), it's not going to fly. Text driven modifications imply conceptual human input; besides, they are inherently worse than higher-order input, just like text to image alone is worthless for anything meaningful.
You are about a year behind the state of the art.
I actually did figure out what works and what doesn't in real artistic use. Which is the entire point of the article in OP - text doesn't work well beyond the basic use or amateur play, regardless of it being the initial prompt or editing; you need sketching and references (and actual skill) to do real work. I don't think anybody's using available methods of textual modifications for anything complex - they are cumbersome and unreliable, even worse than textual prompts. In fact, I haven't seen anyone using them at all.
Besides the implementation details, natural language just doesn't have enough semantic density and precision to give artistic directions, even for a human or AGI. Higher order guidance, style transfer, and compositing is how it's done.
...Other than the USPTO and the federal court system issuing multiple ruling stating the opposite, including a decision last week which specifically stated that the output of an AI model is not copyrightable, upholding an earlier decision by the USPTO... (https://www.hollywoodreporter.com/business/business-news/ai-...)
The act of prompting and customizing iteratively especially in systems which allow the user to submit a prompt that modifies the existing work for example "replace the human being with a monkey" "make the monkey pink" etc are clearly creative works that USE an AI not uncopyrightable.
If you want to argue that point you absolutely cannot do so on the basis of a case that literally never addressed that issue unless you would like to traverse the muddy ground between actuality and fiction.
The act of prompting and customizing iteratively especially in systems which allow the user to submit a prompt ...are clearly creative works that USE an AI not uncopyrightable.
If you want to argue that point you absolutely cannot do so on the basis of a case that literally never addressed that issue unless you would like to traverse the muddy ground between actuality and fiction.
The case literally deals with the output of the AI model, not the input. But on that note...under existing law, code can be copyrighted but not its output. Thus, it is logical to reason that prompts to an AI model can also be copyrighted to the extent they are not strictly functional.
But with AI models and content generally, nobody cares about the prompts/inputs. The output is what matters. (For comparison: Deep Impact and Armaggedon were both the results of the same input: disaster movie in which a team of astronaughts has to go to the asteroid to blow it up before it destroys Earth. The "models" were different screenwriters and directors. Compare the outputs: one is a blockbuster classic, and most people don't remember the other movie.)
> The headline doesn’t seem to be what actually happened. The filer was arguing that the ai created the work on its own as a work for hire and thus the ai was the author with the computer scientist merely being the owner of the copyright as it was made for hire. I don’t think the argument that ai is a tool and the human operating it is the author was considered because the filer explicitly didn’t want to consider it.
> In the review being appealed here
https://www.copyright.gov/rulings-filings/review-board/docs/...
> It makes it clear that the computer scientist doing the filing was trying to argue this was a work made for hire with the author being the computer. They wanted to argue that copyright can be assigned to non humans, but that just isn’t how the law works. The summary makes it clear early that it’s just taking their word that the work had no human input and was thus purely the creation of the computer.
This seems to be a a better article https://www.millernash.com/industry-news/paradise-denied-cop...
https://news.ycombinator.com/item?id=37189599
In short A: Computer generated efforts virtually certainly qualify for copyright.
B: Non-artists can in fact iterate and modify work not just randomly generate shit.
C: This will virtually certainly get much better over time.
You will still get much better work out of a professional who can both utilize such tools when desired, and actually create not just copy or prompt art. This thesis is supportable but we shouldn't build it on sand lest it look more vulnerable than it is.
Checkout confyui, it has an incredible amount of composability that allows you to generate new images based on others. Like image to image but on steroids.
For example, you can generate a character sheet and use it to generate the same characters on different poses using controlnet. Or you can have a base image for an object and use that to generate the same object from different angles and/or different colours etc.
Any sources for this?
I'm of the impression physical distribution is on the rise compared to the earlier days of digital music. This has nothing to do with Spotify, and all about the digitization of music itself.
Anecdotally many people I know now purchase merchandise and media as a way to support an artist they like, rather than listen to the music they make in a physical format.
I remember when I was a child, on a Sunday afternoon, my dad would put on an album and listen to it. Just listen. Very, very few people do that now.
Now we have a lot of demand for “incidental music”. Something you listen to while you do something else. Driving, reading, surfing the net, coding, cleaning…
There was a fundamental shift in how people consumed music that started around the time music became portable. Spotify won the race, but if it hadn’t been Spotify, it would have been someone else.
Listening to incidental music all the time devalues music. And we do it not because we wanted it but because Spotify, Apple music etc promote it. Until then "just play random stuff that this ML thinks is similar" was not a thing. But subscriptions make them more money than if they just let you buy albums and stream what you bought. I wish more artists didn't sign up for this but unfortunately big labels did.
But you can listen to non incidental music that you have specifically chosen while doing something. Even your dad could be doing something while listening to music (thinking).
The music had a purpose.
The transistor radio was invented in the 1950s. And quickly became used as background music as life progressed.
Also incidental music is not a new thing. Tavern musicians as background music have been around for centuries. It is hard to prove, but likely for thousands of years.
If you're already big enough that, i.e., XL Recordings can ask you to make a record without getting rights on the master, I wouldn't count it as a good example of "indie artist".
EDIT: Not making Taylor Swift money, but not many are
Also note the terms of his deal with Columbia are unlike most major deals in that he has a 50/50 profit split after his advance payment got recouped, retains either full control or 50/50 control of masters, etc.
Also, Spotify promotes my music via editorial playlists and algorithmic (eg Radio or Discover Weekly), so I’m probably making a lot more total revenue than I would have on iTunes.
then resell the output for $$$ on something like Splice
This is silly. The USPTO and Courts have repeatedly stated that AI-generated media is not subject to copyright protection, so there are no licensing revenue opportunities for the big publishers/artists/whatever. This means: AI-generated content is not protected by copyright, so anyone can use a piece of AI-generated art however they want without a license and unless the law changes AI has no value to the content industries.
EDIT: Also, the USPTO has noted that the use of AI-generated content in a work will mean that the entire work will be presumed AI-generated except for the portions the content owner can demonstrate were generated by humans. The backend costs of maintaining AI-supplemented works will almost as expensive and burdensome as the costs associated with patents.
Also, I think people on HN have a very glorified view of how much money musicians make from streaming or cd/album sales: basically zilch, unless they're popular enough to be in repeat on the radio. Most musicians made their money from performing: generally a little bit from ticket sales or venue incentives (like % of booze sales) but the real money for the performers was from the sales of band merch, which is why it gets pushed so heavily.
At the end of the day the model itself is going to be basically irrelevant, while knowing whose works were actually used to train it being the truly differentiating feature.
Yes, by lawyers, when they sue the owners of the AI model for copyright infringement, because this would not be a use protected by fair use doctrine. This will actually make human-generated works more valuable because now every work used to generate an AI work is now worth at least $75,000, even if its market value would be significantly less (or even commercially worthless) today.
Due to the costs associated with licensing of human works, if AI-content becomes a thing, it will probably be more expensive than hiring a human to do the same thing, because the model will have to account for the cost of paying a license fee for every work that was incorporated into a specific output.
Maybe I’m misunderstanding you, but how much money do you think the 7500 creators on Spotify making $100k+ [1] would be making without Spotify or other streaming platforms? My guess is closer to zero than 100k.
Also 0.09 percent of 8 million creators making 100k+ [1] sounds horrible, but my guess is that should be taken with a grain of salt. How many folks are included in that 8 million who registered, but uploaded nothing? How many uploaded once or twice? How many uploaded and did ZERO promo of themselves? How many are just plain terrible musicians?
A number of years ago when I stumbled on him, Russ was pulling in a few hundred thousand per year from streaming. Looks like he’s making 100k per week as of a couple of years ago [2]. Yes, he’s probably an outlier. But he works his butt off on his craft, handles production and writing himself, and markets himself well.
Headlines like “Big tech and AI destroying the indie music industry” get more clicks and attention than “Streaming platforms provide income where once there was none” so shrug.
[1] https://www.digitalmusicnews.com/2021/02/24/spotify-artist-e... [2] https://twitter.com/russdiemon/status/1325853093074923520
Even just training the model requires someone to copy the original work from somewhere and store it into a database to use to train the model. If you don’t have permission to make that copy then it’s commercial copyright infringement independent of anything done by the model after that point.
> A human looking at someone’s artwork is “training a model
sure, except that model often takes months or years to train (wall clock years, not 1000-core cpu-years). and the end result is not a human that can stamp out new/competing artwork every 100ms.
for any kind of creative/performance/art work, these are watershed times. us coders are not super far behind.
Spotify basically killed any money coming from the physical distribution - Worse than piracy, which was inevitable too at the time, but at least you didn't have to pay your lawyers to renegotiate with your label on top of NOT getting any money.
Adobe, OpenAI, whatever: they want artists to draw for them for peanuts to train their model, sign a waiver saying "I'm ok not getting any money from any AI art made from this", and then resell the output for $$$ on something like Splice[1], at the same time overtraining such models in ways that make extremely obvious whose artist made them in first place.
At the end of the day the model itself will be basically irrelevant, while knowing whose work were actually used to train it will be the truly differentiating feature.
But you know, "the AI did this picture, so we don't have to pay you."
[1] https://splice.com/features/sounds