> This is why I introduced the concept at the beginning of royalties for AI work. Ultimately AI is not an alien brain generating texts or images from its own imagination. Neural networks and large language learning models input real people’s intellectual property and output that intellectual property when prompted by someone using the program. The novelness of the output is merely how the IP is reformatted to fit the parameters of the user’s prompts.
The crux of the argument is this paragraph. But swap out AI in that paragraph for "writers who have read the classical western canon" and suddenly you realize that no one has original ideas, and that the idea that somehow humans are uniquely capable of being the primary causes of their own thoughts and owe transitive credit (and royalties!) to the great authors of all time, or their estates, is madness.
The threat is the even further enclosure of the commons and the expropriation of culture by corporate interests who actively seek to destroy common culture by preventing the very people whose participation in a now privatized "culture" enables corporate profits, from ever earning a cent because they do not "own" the "IP" to the characters that they love and that are only common because they are a shared element of a common culture. It is absurd. If media corporations had to compensate whenever someone mentioned a trade marked character because it counted as advertising then let them keep their "culture" private. Until then, they are the beneficiaries of billions of dollars of stolen advertising and stolen culture.
> The crux of the argument is this paragraph. But swap out AI in that paragraph for "writers who have read the classical western canon" and suddenly you realize that no one has original ideas, and that the idea that somehow humans are uniquely capable of being the primary causes of their own thoughts and owe transitive credit (and royalties!) to the great authors of all time, or their estates, is madness.
I disagree. As a cinephile and bibliophile, I'm often astounded by the creativity of writers. I don't know where they get their inspiration. Sadly, I don't possess such creative inspiration myself, despite having read "the classical western canon". Much of the great work is actually autobiographical, taken from the writer's own life rather than derivative from previous work.
I would argue that autobiographical content is no less derivative than reading a book or watching a film, then retelling it. Any creativity is in the difference, and only the difference.
Learned to read from the Commodore 64 user manual, Spock and Wesley Crusher as a role models. Summer holidays in Mousehole, first trip abroad was to meet my Uncle in Canberra, amongst other parts of Australia and Asia, where I totally failed to see Uluru or get beaten up by a kangaroo (though in fairness I was like nine and would have been mild and gross irresponsibility respectively on the part of my parents if they had in fact happened). Raised Catholic by an atheist father and a mother who filled the house with Hindu idols, Orthodox icons, healing crystals, and runic horoscope equipment. Saw my father suffer an epileptic fit at the kitchen table — what's now called 'tonic-clonic' but at the time we called it 'grand mal'.
You know, the usual."
(Can you guess what, if anything, is embellished?)
Um... I think you're missing the point? Almost anyone could create autobiographical material that sucks. I'm sure LLMs could cobble together autobiographical material from sources that sucks too.
Truly creative writers are capable of producing material that's good, entertaining, interesting, moving. That's a rare talent, and it's not derivative.
Ok, but nothing in your comment helps me change my mind about what you original point was, as that still looks like you asking "where do they get their inspiration? Autobiographical experience!" and me saying "being inspired by reality isn't any less copying than being inspired by existing canon".
> Truly creative writers are capable of producing material that's good, entertaining, interesting, moving. That's a rare talent, and it's not derivative.
I think we must be using the word "derivative" differently.
To me, Strata by Terry Pratchett is clearly derivative of Ringworld by Larry Niven. I'd say it's not just good, entertaining, interesting, but also that it's superior to the original… and yet, still derivative.
From what I've seen from ChatGPT-3.5, it… writes fiction like someone born in 1950 who was always disgusted by sex and violence and who now has mild Alzheimer's and gets confused what they're supposed to be writing about — that it's as good as it is, is of course miraculous, but even in this state it can be put in front of an editor and turned into something fun far faster than getting a really good human to do it nearly right the first time.
(Some people don't realise it's absolutely bad enough to need an editor, just search for reviews containing the string "As a large language model" for examples).
As an example of getting it right (IMO), the following link was mostly generated by AI, minimal editing by me for formatting, repeated prompting and plenty of cherry picking to make sure it kept to the fictional reality and didn't start veering into "this hypothetical creature" or "look what nonsense ye olde people used to believe in":
> I'm sure LLMs could cobble together autobiographical material from sources that sucks too.
That would be a biography not an autobiography. And AI writing in general will remain so until some AI gets a long term memory and also embodied… although it might be the long term memories are autobiographical "writing" created as we sleep from our waking memories.
> Ok, but nothing in your comment helps me change my mind about what you original point was, as that still looks like you asking "where do they get their inspiration? Autobiographical experience!" and me saying "being inspired by reality isn't any less copying than being inspired by existing canon".
The question is about royalties for writing. You seem to argue that writers don't deserve royalties because all writing is derivative, "a series of footnotes to Plato" as A. N. Whitehead put it. My argument is that good writing, indeed great writing, is not merely derivative, otherwise it wouldn't be interesting. Great writing is a rare talent, I might say divinely inspired if I believed in the divine, so I'll say genetically inspired instead. I think that writers need to be supported, protected, encouraged financially so that they can continue to exist and write great material for our benefit and enjoyment. The alternative is bleak: if writers are not supported via royalties, then the only financially viable "art" will inevitably be mass-produced, mass-marketed pablum. That's a depressing prospect.
Of course super-famous writers make a ton in royalties, but it's actually the other writers for whom royalties are crucial. In a lot of cases, royalties make the difference between survival and destitution.
> To me, Strata by Terry Pratchett is clearly derivative of Ringworld by Larry Niven. I'd say it's not just good, entertaining, interesting, but also that it's superior to the original… and yet, still derivative.
I haven't read Stata, but I thought Ringworld was overrated, so I'm inclined to believe that another work could be superior.
> That would be a biography not an autobiography.
I meant "faked" autobiographical writing. Writing in the autobiographical style. I'm sure you can ask OpenAI or whatever to write an autobiography. Hell, some actual "nonfiction" autobiographies are quite fake (and frequently ghostwritten).
> You seem to argue that writers don't deserve royalties because all writing is derivative
You're interpreting my words rather creatively there.
And by creatively, I mean "wrong".
I'm being more inclusive, not exclusive, for where the work is done.
To take a different tack: it makes sense for many authors to go to a publisher and have a back-and-forth with an editor and end up with only 10% of the gross sales, and this implies that 90% of the value isn't really coming from the author (in most cases and almost all the famous authors are exceptional).
> The alternative is bleak: if writers are not supported via royalties, then the only financially viable "art" will inevitably be mass-produced, mass-marketed pablum.
I think we already live in such a world: as you agree, most authors cannot sustain themselves from just the writing, only rare exceptions like JK Rowling and Stephen King. As the rest only barely make do, I assert that this demonstrates they are forced to do exactly what you decry.
(There's a lot of free stories distributed on reddit etc., some even have patreon set up. I don't think that makes a difference either way).
> it makes sense for many authors to go to a publisher and have a back-and-forth with an editor and end up with only 10% of the gross sales, and this implies that 90% of the value isn't really coming from the author
I'm not arguing for any kind of philosophical "labor theory of value". My point is very practical: writers need royalties. Yes it sucks for the writers that they only get 10-20% of the sales, but it's infinitely better than 0%!
> as you agree, most authors cannot sustain themselves from just the writing, only rare exceptions like JK Rowling and Stephen King.
I don't agree with that and didn't say it. In fact it's empirically false. I said that in the hypothetical scenario with no royalties, most authors wouldn't be able to sustain themselves. In the real world, there are a lot of professional writers other than Rowling and King who make a decent, good, or even great living.
The reason that the film writers and actors are going on strike is that streaming and AI are an existential threat to them. But it can't be an existential threat if they don't exist in the first place. Professional writers and actors do exist aside from the super rich and famous ones (who are striking more in solidarity than from personal fear). Of course show business is very difficult to get into; nonetheless, it's a lucrative industry supporting countless professionals. And most of them need their royalties, because they don't get the giant unfront checks that the big stars do.
To expound on your point, corporations will be able to use AI - giant conglomerates such as Disney, that own enough media to train AI models on only works they own copyright on (and own enough senators to make sure those copyrights last), will happily use AI.
But you, peasant, can only train it on century+ old works. Because "to promote the progress of science and useful arts", nothing post-WWII may become public domain, or artists will starve and culture wither.
If nobody has original ideas then where did the "classical western canon" come from? And what about all those people in the Far East, who never read Homer? Where did the Romance of the Three Kingdoms come from?
I don't understand how we got to this point, where a machine trained to reproduce the results of human creativity is taken as evidence that there exists no such thing as human creativity. What are we training LLMs with, if not the results of human creativity?
I think maybe that's all the result of growing up in an era were there is so much repetitive, trope-filled entertainment that people assume nobody could ever create an original work of art that wasn't just a stitching together of other peoples' ideas. But that doesn't make sense for the same reason I give above: those ideas must have been original at some point.
> But if I did edit a feature length film that became a major box-office hit, I should get some residuals to help tide me over.
This is the line that made me sit and think about this whole construct for a bit.
I've worked as an engineer for multi-billion-dollar companies. I've designed components and whole systems that then get put into mass production and stay in production for years. Safety-critical in some cases. Some of those components are still being made today and still generating revenue for those companies, nearly a decade after I moved on.
Should my work back then be continuously rewarded if that company still uses that design, unchanged, and still generates revenue from it? I've since gone into consulting, and as a private contractor I work similar stints as a hollywood editor- sometimes I do not have a steady client for many months at a time. It would be great if my engineering work for companies past could 'hold me over' via residuals. But that isn't how getting paid for my time works. I just bid my hourly rates accordingly, of course.
You could say stock options in companies are similarly equivalent to residual checks for hollywood, but... it's not. Stock prices aren't always tied to anything rational (although... I don't know how hollywood accounting is either). And they aren't something that are guaranteed to be negotiable or granted post-facto.
What I'm trying to point out here is... a lot of people in Hollywood seem to think they have something special somehow and aren't at the will of a market to set prices. If they want better profit sharing, there should be ways to do that, but its ultimately down to setting your prices and being paid for your time and thats is how it works for nearly everyone else in the world who 'creates' things.
The difference is 1) the power law distribution of revenue and 2) the finality of a movie.
1: A few movies (e.g. Barbie, Spiderman, Oppenheimer this year) make the majority of the money. Most movies make little. So if you pay based on expected return you probably don't pay your actors at all, or pay them peanuts. That's inequitable. But you can't pay them lots because you don't know how much money the movie is going to make. The solution is profit sharing. Most software business have much more stable income.
2: A movie, when finished, is done. There may be a director's cut in a decade or two if it's a super popular movie. There are occasionally different edits for different markets (alternate endings etc.) But basically it's done. The movie that is shown on release is the movie you watch 20 years later. So you can easily apportion credit to the actors, writers, etc. This isn't the case for software. It continually evolves. You made the initial design, but 64 other developers have worked on it since. Who gets credit?
Good points, helps solidify differences that I had trouble thinking about.
But to add, for me I wasn't talking about software, I actually engineered hardware- in my case, vehicle and airframe structures. Way more direct and less abstract than software. Large software systems may be some of the most difficult examples in the credit tracking problem, for sure.
point 1 to me is an argument against profit sharing. direct profit sharing would mean that only those who contribute to a profitable movie earn any of the profit. but we also need to pay people working on movies that don't make a profit. the only way to make that work is to use the profitable movies to finance the unprofitable ones. if there is profit to be shared then it would be the profit of the studio as a whole after all movies are paid for, and that profit should be shared with all employees and contractors of the studio.
FWIW residuals are all a result of union effort, so the market IS actually what’s allowing for/encouraging the residual structure. If the studios felt they could create similar products without the benefit of union workers (eg SAG, WGA, ACE, etc), they could easily break contract and start hiring from wherever they please, under whatever rates they agreed upon. Perhaps this current strike will end in just such a scenario, which would support what you’ve outlined in your comment here.
>Should my work back then be continuously rewarded if that company still uses that design, unchanged, and still generates revenue from it?
I believe so but for a different reason. I think a model where you are paid very little but remain the owner of your code, a hit and miss where you gradually accumulating income over many years, could work wonderfully if you also remain partially responsible for maintenance. The revenue shared should go up for every hour of effort but be a fixed part of earnings a bit like stock options but with various safeguards build in. If the company goes bankrupt at 1% of the [say] 30 year period they should pay 99% of the agreed upon salary as if a normal job.
Over the 30 years one could agree to pay many times that amount.
I've previously called it a priest hood where each monk maintains a specific module or library. The less work they do the better the results.
I had a similar discussion on Tiktok of all places regarding the Rachel Zellger comments on her playing Snow White.
“I had to sit 17 hours in a Disney Princess dress! I want money for every hour it is watched” and I had a similar thought to you.
I feel like its a good idea but I’ve definitely missed something with capitalism: I thought the notion of being paid for your time leaves the fruits of that labour in the rightful hands of those who “invested” the capital up front to pay you.
If thats the case then getting residuals is directly against that notion at least, but hey, its what they negotiated so good on them.
However, where would that end hypothetically- what makes artistic works more inherently valuable in the long tail? I mean- aside from the fact that we all agree.
Why is design work, or even craftsmanship not rewarded the same? Its not like there is additional work to be done for artistic works.
It it because of likeness? I don’t fully understand.
That's just it- at least with a likeness- your face, your voice... I can see a bit more of a justification. It's you, and it means 'you' are a continuous part of the product directly.
But an editor? A composer? A VFX artist? Why not the light guy or the craft services people?
How about this thought experiment:
My tile guy did a bang up job on my shower when I rebuilt this house. I mean... literally this rock wall in the bathroom is a work of art. It definitely increased the value of the home. I paid about $1600 to the guy for that work, it took him about 3 days. (I actually can't remember what it cost right now and I'm not up on masonry work prices so if that number seems way off don't read into it.)
Now, about 8 years later, and the market value of this house is about double what it was before I renovated. If I sold the house now... that nice shower is definitely adding to the fact the house is worth what it is, along with the overall market increase. If I made a 180% return on this home when I sell, should I fire off a $1280 check to my tile guy?
You could replace that rock wall with a turd wall and the house would be worth exactly the same and it's being valued that ONLY and ONLY because other houses in that location ask the same prices.
I lived to hear everything. "My property has doubled in value because of a turd wall".
After finishing reading the article, it didn't mention one anecdotal cultural trend a friend who's teaching film class at his local community college noticed.
Every semester at the beginning of class, he'd quiz his students what they've watched. In the past few years, nearly everyone has limited knowledge of current films, let alone old films. He realized young people do not have the attention/habit of watching long-form films.
Social media/TikTok videos own the top of the attention economy ladder now, not movies. This trend is not reversing anytime soon.
Here's an alternate interpretation: they're watching television now rather than movies. Because the kind of character development and depth of storytelling that can happen in 20 or 50 or 80 hours goes far beyond what you can do in 2 or 3.
And the idea that young people can't watch long things due to a lack of attention goes directly against the evidence of the increasing habit of binge watching.
Forget about watching a mere movie for just 2 hours at a time. Try 4 or 5 hours instead, where you can cover almost half a season of a drama. You immerse yourself in the world for an entire evening, the way people used to do with novels.
But I wonder if their knowledge of TV is also as deep.
When I was a kid the TV was filled up with a lot of repeats of TV shows. Things that had been create 10,20 or even 40 years before. I'd watch them because there were not many other options (probably because new shows were too expensive to waste on offpeak timeslots)
I wonder what percentage of today's kids would recognize any of the below shows beyond the name (or more-recent revivals) let alone have watched a large number of episodes.
As someone born at the end of the 70s, and not anywhere near the US, I was kind of surprised by how many of those 60s shows I knew. (which also says I never realized they were that old...)
Among younger generations, binge-watching is very frequently done by having the video going in one window, while doing something else like Discord in another window. Or having the video playing on a laptop while browsing TikTok on one’s phone. It is no counterevidence for declining attention spans.
Because sure, if you've got Property Brothers on in the background while you do other stuff, that makes sense. You can go in and out.
But if you're watching Squid Game, or The Last of Us, or Yellowjackets, or even Wednesday -- you'd better be paying attention to every frame. Modern dramatic TV demands your full attention in a way that The Golden Girls didn't.
The idea that younger generations are binge-watching Yellowjackets while browsing TikTok doesn't make any sense.
I wonder if the shift has more to do with movie theater attendance absolutely tanking for COVID-19. Even in 2023 ticket sales have been significantly less than the relatively constant rate for the few decades prior. The advertised/direct fill in was of course streaming services but these lean away from the typical blockbuster film content or even full length films at all. I'm not ruling out new social media content trends as a factor in the trend change completely, or as an alternative fill in, but short form social media content doesn't really explain why the "why" for a recent shift. After all, it's not like TikTok invented quick social media content it's just a popular modern choice.
In my final semester I wrote/directed a short film with friends that I met in the university’s filmmaking club. Most of them were Freshman or Sophomore.
I was looking forward to discussing some deeper and more obscure films with them, and they did have more knowledge of, for example, Lynch’s work than most people I talk to. However, they overwhelmingly talked about the new Star Wars saga and Marvel movies.
It's kind of inevitable. I grew up hearing movies decried as brain-rotting, imagination-killing junk food preventing everyone from reading a good book like nature intended. Of course the medium would be dethroned by something even more accessible.
When there's something even more convenient people will latch on to that instead. Expect nostalgia for the good ol' days when social media used to instill community values and awareness into the youth.
I own a projector and good surround-sound setup, and have a circle of friends who call themselves cinephiles. Sometimes they will suggest some film for me to put on for everybody. Nevertheless, virtually the moment the film starts, they take out their phones and then keep only half an eye on the film. I don’t even bother mentioning this out loud as a problem, I feel like it would just make me look like an out-of-touch weirdo.
I started the Cinema Club at Google, which is still going on, AFAIK. You can download the list of the 300 or so movies which I put on.
I really don't care about MCU, Harry Potter, Hollywood studios, AI, or Rotten Tomatoes. It's pointless to rage about it. Just offer an alternative. I'd rather show a great movie that no one comes to than a garbage one that attracts big crowds. Once in a while we had a film that was both: Chinatown, Rashomon.
Our reaction to anyone was, "You can start your own movie club. We'll help you; here are all the secrets." The big secret was just to ask the boss for the money; 10 in a row said Yes.
The most common thing people said to me was, "You guys show some great movies. But I have to leave to catch the shuttle!"
I've wanted to do something like this, but I'm not sure of the 'public viewing' licensing / permissions / royalties required. I looked it up a while back, but life got in the way and so the idea was mostly lost (thanks for the reminder).
You can get Public Performance Rights. They ask (1) how big is the venue, (2) are you charging money, and (3) is it advertised outside your organization.
Remember: it's negotiable. Besides MPLC, there is Swank.
I can’t help but feel that fighting for residuals for digital extras is a losing cause. Regardless of what you negotiate, the studios will probably eliminate the “scanning actual humans” step sooner rather than later and replace the entire thing with something like Unreal’s Metahumans.
IP is like the car: an interesting idea that has been pushed way too far.
It's hard to even imagine the number of problems that would go away instantly if we got rid of it.
And if someone wonders if people would stop creating if there was no more IP:
1/ Creators always create; that's what they do.
2/ Arguably, the best works of art were created before IP even existed (invented at the very end of the 18th century).
3/ And most importantly, IP doesn't pay creators. That's the whole point of the article. IP pays the studios' CEO ridiculous salaries, and their lawyers' whose sole job is to create artificial scarcity.
All this actors-writers artists rights debate as of late feels like the last cries into the darkness before media, culture and the entire dialectic of creation and consumption changes forever and AI becomes THE thing behind everything. Custom generated movies, first person VR experiences, all the pleasures of a world soon to placate humans as their minds lose ascendancy on the cognitive totem pole to artificial intelligence.
>My favorite model for a distribution platform is BandCamp.
> On much of the free-to-play music, the audience can listen to the album twice before the content locks and BandCamp asks you pay to continue listening to it.
easterneuropeanmovies.com[1] comes close. Obviously, watching movies twice before paying doesn't work. Instead, easterneuropeanmovies.com allows you to watch the beginning of the movies for free.
If artists want a platform like BandCamp, they should create a pool of available movies that platform creators could instantly use.
Now, it's difficult to establish a new platform because it's not enough to program the software. The movies have to be licensed and an audience has to be acquired. If artists already take care of the license side, and if they help with acquiring an audience, they could lay the foundation for their favorite distribution channel.
It is going to be weird in the future, when people realize how much use these companies get out of things like the $200 scan and perpetual license to somebody’s likeness, and as a result stop using them. All art will be presumably based on remixing a set of scans from between like 2020 and 2025.
Will that be significant 15 years from now? I imagine most work by then using 3d models that aren't based on any particular person, to avoid license fees. I can't imagine this strike ending with one-time payments for perpetual license to use actors' likenesses. However, even royalties based on screen time of AI-generated derivatives of their likenesses won't save their jobs in the longer term.
In 15 years I would expect a large percentage of stand-ins to be CG but also that there are various movies which are entirely generated bybl extensions of today's diffusion models. The actor personalities there-in will have been either created on the fly or supplied as previously created to the generator software.
Excellent long-form article. A very thoughtful look on studio/streaming economics, the vanishing long tail, and the writers strikes and AI, short-films and bandcamp and discoverability … a lot to unpack.
I didn't like the article. Very long, and very shallow. It seems to take as an axiom that the status-quo of "residuals"-based payments is worth defending at all costs.
> Scrolling through it is tough, but not because of too many options, but because too many repeats.
Repeats and already seens.
I seldom want to rewatch movies, I do so usually only to show someone some classic they missed. So I find the interface of netflix (and youtube, for that matter) infuriating in that they seem heavily centered on showing me things that I've already seen and don't have any option to suppress them.
What the heck is with that? While my preference to see something new might not be the majority of their views, I assume that most people will want to see something new at least some of the time-- so why isn't this an option? If anything it should make the service more sticky since they'll have hold of that already-seen data.
Now some better search -- year ranges, category/keyword, credits search, etc. but presumably that would take a lot of work compared to just letting me skip the stuff I've seen or seen recently.
Genuine question that I couldn't see in the article; does everyone on a film get residuals, or just the more notable roles like writer and actor? If not, did e.g. a lighting specialist not contribute at least as much as a person whose face was scanned?
AFAICT, all credited directors, writers and actors get residuals; production crew do not get residuals directly, but it's paid into health and retirement funds; uncredited actors get nada.
As a general rule (there are notable exceptions), I don't watch any movies made past 2019/20. Most of them seem to have some sort of a messaging embedded inside. Regardless of the nature of the messaging, I figure that they should pay me for watching it, not the other way around. I hope this would change soon. John Wick was awesome though...
If you don't like movies with messaging, I have bad new for you about pretty much whole movies history. Because, oh boy, old movies were quite big on messaging ... and they even had codes/guidlines to ensure messaging is right.
That's a good point. I used to watch NCIS and liked it a lot. Then I learned that certain contractors were literally financing the production. I lost interest soon thereafter.
The crux of the argument is this paragraph. But swap out AI in that paragraph for "writers who have read the classical western canon" and suddenly you realize that no one has original ideas, and that the idea that somehow humans are uniquely capable of being the primary causes of their own thoughts and owe transitive credit (and royalties!) to the great authors of all time, or their estates, is madness.
The threat is the even further enclosure of the commons and the expropriation of culture by corporate interests who actively seek to destroy common culture by preventing the very people whose participation in a now privatized "culture" enables corporate profits, from ever earning a cent because they do not "own" the "IP" to the characters that they love and that are only common because they are a shared element of a common culture. It is absurd. If media corporations had to compensate whenever someone mentioned a trade marked character because it counted as advertising then let them keep their "culture" private. Until then, they are the beneficiaries of billions of dollars of stolen advertising and stolen culture.