Meta Would Like to Copy Your Work, Not Pay You

The advent of applied computer-based statistical analysis that we mistakenly call AI represents the future of tech in that it seeks to automate for us every artistic or creative project that may bring us joy while leaving things like labor for us to muddle through. In this way, “art” is a product, one to be thoughtlessly constructed and just as thoughtlessly enjoyed. The presence of the product called “art” is the goal, irrespective of quality or coherence. This can lead to the overriding feeling that all art and creative pursuits are interchangeable and meaningless; recognizable brands and franchises laser-printed onto hollow shells offered to us not because there was a story to tell, but money to be made. It can create a sort of contempt, one in which audiences are seen as baby birds in a nest screeching for food and large corporations are our mother birds vomiting an indistinct gruel of content into our mouths.

Meta, a company that continues to thrive as many point out its role in eroding our democracy and poisoning the public information well, has admitted to feeding its AI a large library of copyrighted books without the permission of authors or payment for them. Perhaps “admitted” is too accusatory word, however, because Meta plans to defend this choice by making the argument that they’re acting under fair use. The library in question is called Books3 and it’s a dataset created by researcher Shawn Presser in 2020 to provide a greater wealth of information to train machines on. Books3 totals 37GB and includes over 195,000 different books and has been used by several AI companies including OpenAI. Meta’s proud announcement comes amidst a lawsuit filed by a group of authors, including Sarah Silverman, whose work they allege has been infringed upon and the ongoing suit the New York Times filed against Meta and OpenAI.

It’s unclear as of yet what shape this fair-use argument will take. There are four factors to fair use:

1. the purpose and character of the use, including whether such use is commercial or is for nonprofit educational purposes


2. the nature of the copyrighted work


3. the amount and substantiality of the portion used in relation to the copyrighted work as a whole


4. the effect of the use upon the potential market for or value of the copyrighted work.

Dustin pointed out that the fourth point might be the avenue Meta chooses as they could argue that there was no money made for Meta from the use of the copyrighted work in their AI, but Seth explained that fair use is far from a simple call. According to Seth, no one factor is decisive and each can be used to sway for or against the others. We may be sitting on our hands waiting for the answer on this one, especially with an issue as complicated as this. It can be very easy to feel the answer is obvious, but our legal system is still catching up with the leaps and bounds of tech.

Speaking as a writer and one who’s been working on a novel for quite some years, I find the idea of anyone’s work being taken without permission or pay to be pretty dispiriting. It makes me sad that books, along with all other forms of art, are valued so little by us as a people that we treat these cultural artifacts as nothing more than fodder for a machine. It makes me sad that instead of digging into the beauty and flaws of human-written stories, we take their quality so for granted that a machine without emotion should be trusted to create them.

My sentimentality aside, it’s an unfortunate reality of capitalism that the meaningful is reduced to the meaningless to make more art faster for lower costs and that a robot need not be hired or paid. As consumers in this environment, it’s on us to keep our eyes open and stay conscious of what’s happening. Quite literally, buyer beware! If books and the writers who create them mean something to you, then this issue is something you could make a lot of noise about. Be aware that Meta is trying to have their cake and eat it too with the “AI” they wish to feed with copyrighted material they didn’t pay for or license.

Instead of perhaps investing in humans with brilliant brains, Meta is trying to take advantage of novel legal territory to make the unpaid use of copyrighted material they don’t own legal. We can watch this happen, anxiously worrying, or we can make a lot of noise about it. I accept that the system we live in is adversarial to us, but as consumers and humans with voices, making a huge deal of this has value. Meta could get away with this, they could rewrite the laws so that their learning machine can write all the books we’ll ever need and no human need ever write again, but we cannot be silent on the way to that. Regardless of the outcome, resistance has value and in my view, defending the work of artists is always worth it. The Summer of Strikes was historical, but the struggle isn’t even close to over.