Even though it is framed as an author's right, copyright law was originally designed to protect publishers. This is why, to this day, authors rarely make enough from their books to earn a living. What authors want, most of all, is to be read. If AI can offer that to them, shouldn't we rethink copyright?
This is a link-enhanced version of an article that first appeared in the Mint. You can read the original here. For the full archive of all the Ex Machina articles, please visit the website.
There is a widespread misconception that copyright law exists to provide authors with the incentive to write. But most writers are unable to earn a living off writing. With Generative AI challenging traditional notions of intellectual property, it is well worth asking what truly motivates writers—and what exactly it is that society should be looking to reward.
Intellectual property dates back to the mid-1400s—to a time when the importance of the Venetian Republic was on the rise. While Venetians were already wealthy from trade, they knew they would be richer still if, instead of travelling to acquire the fine silks their customers craved, they could learn how to make this fabric.
For this, they needed to entice Eastern artisans to settle in Venice for long enough to pass on their skills to local craftsmen. And so, they came up with the concept of the privilegio, a
When Johannes de Spira (master printer and contemporary of Johannes Gutenberg) decided to ply his craft in Venice, he was rewarded with a privilegio of printing books in the Republic. This transformed the Venetian Republic into a major centre for publishing. Other countries envious of Venice’s success followed suit, and privilege-based protections were issued to publishers in Germany, Portugal, Spain, France and England—to the point where the very business model of publishing came to centre around the grant of exclusive rights to print.
In time, the privilege granted to the publisher was reframed as being derived from the labour of the author. England was the first country to vest this right in the author by law, enacting the Statute of Anne in 1710, the world’s first law that elevated copyright to the status of a personal right.
In practical terms, however, this made little difference. By then, powerful supply chains were already in place, controlling every aspect of the publishing industry. This made it virtually impossible for books to reach their intended audience unless publishers allowed it to happen. Even though authors had a legal right over their literary works, they still depended on the publishing industry to monetise them.
If copyright law were truly an incentive for writing, you’d expect it to yield lucrative rewards. And yet, according to the Authors Guild, the median book income of authors in the US hovers around just $2,000 a year. If you consider the median income of just full-time, well-established writers, this improves marginally, to just over $10,000. Even among professional writers, only one in every ten can claim to make a decent living from the books they write.
Despite these abysmally low income numbers, books continue to be written around the world. Close to 4 million titles are published each year, which seems to suggest that, notwithstanding the claims of copyright purists to the contrary, most authors are motivated by things other than financial reward.
As a published author, I can confirm that the royalties I earn from the three books I have written will never be sufficient to sustain me. But that has not in any way curbed my enthusiasm. One of the reasons I, and I suspect many others like me, put in the hard work is not the money we think we are going to make, but because this is the surest way to ensure that our ideas reach a wider audience.
The rise of artificial intelligence (AI) has forced a showdown between those representing the interests of the creative industry and the companies building large language models (LLMs). The argument is that the mere act of training AI is a violation of the strict letter of copyright law and an erosion of the monetary incentive that authors depend on for their livelihood. If AI can generate stories in an author’s style, they ask, will anyone still want to buy original works?
This argument presumes that authors write to make money, which, as we have seen, is rarely their sole motive. What we do know is that one of the motivations that encourages authors to keep writing even though they may not earn a sufficient income is that it allows them to get their ideas out into the world. If that is the case, perhaps they should embrace AI rather than think of it as something that has come to eat their lunch.
According to renowned economist Tyler Cowen, the path to intellectual immortality lies in “writing for the AIs.” He points out that people already get so much of their information from LLMs, and in a few short years, it is likely that AI will become our primary source for all information on the internet. When that happens, the only way that succeeding generations will get to know what we thought about a given topic is if AI tells them.
For as long as books have been around, authors have relied on their publishers to get their ideas out into the world. AI allows them to bypass these traditional channels, giving them an opportunity to allow their ideas to shape the conversations and consciousness of the future.
Copyright has never guaranteed writers a livelihood, but AI may finally offer them a readership. If only we would let it.
Rahul Matthan
Over 2.5k subscribers
Support dialog
