Why do we need copyright laws, anyway?

The second in a series of articles about the collision between copyright and generative artificial intelligence

Copyright’s real job is different from what most people think it ought to be.

In the USA, the primary purpose of copyright is to foster innovation and spark novel ideas amongst the public for the betterment of society.  

Some people think that the primary purpose of copyright law is to protect the authors who create original work, but that’s not quite accurate.  The US Constitution explicitly refers to the purpose “to promote progress” in society.

Article 1, Section 8 of the US Constitution empowers Congress “to promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries.”

In practice, copyright law confers upon the author a time-limited right to control the use of a fixed original work of authorship. During this period, the owner has the exclusive right to publish, distribute, transmit, or perform the work and make derivative works.   

Clearly, copyright laws do confer valuable exclusive rights on the owners of the work, but according to the US Constitution, these rights are intended to promote the further creation of novel ideas that will serve society by spurring progress. This is not idle speculation. Thomas Jefferson and James Madison conferred about this exact point while drafting the US Constitution.

The upshot? The public benefit is supposed to outweigh the time-limited monopoly rights of the individual owner.

In practice, however, that’s not always how it works. Later in a subsequent article, we will explore some examples of regulatory capture that distort copyright law and divert it away from the public benefit mission. I’ll show you this sort of regulatory capture has recurred for centuries

Copyright is structured as an economic incentive for creative people. The reason why the government safeguards an owner’s control over the work is to encourage further innovation and creative output; that progress serves the interest of the public.

Key point: the only way copyright works is to have the government as a backstop. In the next three articles, I’ll explain why this is so important and how this dependency was abused by tyrants.

It’s complicated. Copyright laws include certain carveouts that enable the public to interact and engage freely with the ideas, even quoting them and incorporating those citations into new works.

One consequence of this carveout is that the US government is never far removed from debates about copyright. The rules and regulations that govern these rights are continuously contested; they are a never-ending work-in-progress in both the legislature and the courts.

The dynamic tension in a nutshell:  copyright law confers a quasi-monopoly on an individual owner for the exclusive right to exploit a piece of intellectual property for a limited period; but it also provides a way for the public to make use of that work under certain conditions.

This strikes some people as a contradiction. It seems confusing: you may own a property… but a total stranger also enjoys the right to make use of your property under certain circumstances. Without permission or even the courtesy of notifying you.

Why?

A tricky balancing act

Copyright law is intended to strike a balance between two conflicting interests:

1. The interests of private individuals and companies who wish to express novel ideas and profit from publishing original works that contain such ideas; and

2. The interests of the public, which needs a way to debate, discuss, process, and respond to the new ideas that enable society to flourish and evolve.

To engage with an author’s ideas in a meaningful way, members of the public must have the ability to make use of the author’s work by citing it, quoting from it, and critiquing it.

If the author’s work were treated like private real estate, such as a fenced-off physical property, then members of the public would not be easily able to make use of it, just as you cannot make use of your neighbor’s walled garden without permission.

Copyright laws seek to provide the public with sufficient freedom to engage with the author’s intellectual property in a constructive way without diminishing the author’s opportunity to profit from her work.

That’s easier said than done.

It demands a fine balance between opposing forces. Maintaining this balance requires continuous adjustment, because technology evolves ceaselessly; and media constantly flows into new containers; and society adopts new habits of media consumption and public discourse.

Fine balance demands a light touch. Unfortunately, government is a blunt instrument, incapable of fine tuning at the rapid rate that technology evolves. Our current copyright laws are the result of centuries of slow tweaks and minor adjustments with occasional patches and updates. They do not change at the pace of technology.

All of which explains why copyright laws are confusing and sometimes dissatisfying.

If you’ve been trying to make sense of the many complaints about artificial intelligence and the unpermissioned use of copyrighted works, you’ve probably been exposed to some of the frustration. Perhaps you’ve seen it expressed on a sign carried by a striking screenwriter on a picket line in front of a movie studio.

The US copyright system seems like the worst in the world… with the exception of everything that preceded it. Allow me to explain.

What happens when ideas and social stability collide

A society defined by ideas will be shaped by dialog between parties that may not always agree.

Such a society is likely to be highly dynamic, not static.  When that dynamism collides with a slow-moving government apparatus, things sometimes can explosive.

As the next three articles in this series will demonstrate, the history of striking a balance between truly disruptive ideas and social stability is marked by dark episodes that occur whenever the balance cannot be maintained. When that happens, conflict follows:  social order begins to unravel as society splinters into factions.

The faster ideas circulate, the more likely that volatility and instability will follow.

As this series of articles will show, the consequences of volatility include a surprisingly bloody history of intense warfare.

This is no exaggeration.

Before we had stable copyright laws, society suffered from decades of bitter conflict over ideas, beliefs, and ideology. Millions died, cities were burned, and empires crumbled. I’ll write about that in the next three articles in this series.

After a century of horrific violence, around 1700, societies in Europe began to realize this arrangement was sub-optimal. That’s when European societies, beginning with England, decided to accommodate the novel proposition that anyone, not just a select few, should have the right to express new ideas and profit from them.

Nowadays we take this blood-soaked history for granted. That’s a big blind spot.

Are we heading for another cultural conflict? Or just another cycle of creative destruction?

In our present moment, things seem to be speeding up again. It’s not clear that our current arrangement is durable. Generative AI has already altered the balance of power between content owners and technology companies; ultimately GAI could very well upend much more than the content industries. That makes people uneasy.

And this is occurring at a much faster pace than the technological changes of the past, like file sharing, home video recording, radio and even player pianos.

Generative AI threatens to devalue the work of human artists. Enthusiasts claim that AI can potentially replace human artists. That’s bound to ruffle some feathers, too.

The unprecedented speed of this transformation is likely to increase the volatility and intensify the conflicts that characterize our current moment.

Why do governments allow this turmoil? Why allow technology firms to destabilize the marketplace of ideas. Why do politicians seek to strike an elusive balance instead of simply forbidding the unlicensed use of intellectual property by generative AI systems?

The simple answer is that we tried that before. It did not work. As the next three reports will demonstrate, attempts to regulate discourse and to suppress dangerous ideas with censorship always fail. Worse, these episodes of top-down control often end in violent conflict and, sometimes, in wholesale slaughter. I know that sounds like an exaggeration. In the next article, I’ll share a vivid example with you.

There is another, more hopeful, answer: within chaos lies opportunity. A steady process of disruption is necessary to keep things flowing and growing. Without the free flow of disruptive ideas, societies tend to stagnate.

Creative chaos is the alternative that the US government is aiming for.

But chaos is…well,… chaotic. Careers and businesses are destroyed when things change quickly. Those who desire progress may be doomed to live in interesting times.

You might say it’s a tradeoff between tyranny, control and censorship on one side, versus chaos, free expression and creative destruction on the other.

Our current approach to copyright law favors the latter. It does this by balancing two forces.

First, democratic societies seek to stimulate creative innovation. Without copyright protection, individuals might have less incentive to innovate. Why do the hard work of devising novel concepts and turning them into published works if someone else can clone your work for free?  If that were the case, some inventors and artists might give up; that would lead to fewer new ideas. Fewer ideas lead to less innovation and, cumulatively over time, less progress for society.  

Copyright solves this problem by allowing the owner the exclusive right to exploit the work, commercially or otherwise… but only for a limited period.

On the other hand, without the ability to cite and quote freely from works, the public debate about the ideas would be constrained. This could limit access to the ideas and lead to a different kind of stagnation. In the extreme case, if there were no legal way for the public to cite works in the public domain and generate derivate works, then publishers would have the ability to extract monopoly rent by gating access to the concepts necessary to help society to make progress. As later articles in this report will demonstrate, monopoly can become a form of censorship.

Copyright laws solve this problem by permitting “fair use” of the work for several specific reasons, such as education, criticism, parody, news reporting and research. Certain laws even mandate compulsory licensing to ensure that valuable content is not withheld from circulation in new distribution systems.

The issue at the heart of generative AI and pre-trained AI models is whether such training on copyrighted work can be done without permission.

Is training considered fair use or not?  Are the works generated by AI considered “derivative works” or are they original works? Is generative AI a “transformative use” that turns old texts into something so new and so useful to society that it is fair to use an author’s work without notice or consent?

The question for the court to decide: Are the benefits to society of AI so great that they outweigh the rights of content owners?

We will examine these questions in the final section of this series of articles.

Spoiler alert: for some questions, the answers remain maddeningly murky. That’s because those questions will continue to be debated in the courts until a rough consensus emerges from courts. That will take time.

Today we find ourselves in the opening round. It’s likely to be an epic bout: it won’t be settled in a single round.

Debate and dialog about ideas is the best way for new information to become integrated into social norms, attitudes, academic research, criticism, the creative process, and professional disciplines. Dialog shapes society: more dialog, and more intense debate, tend to foster more progress.

For this reason, copyright laws deliberately provide a mechanism that enables people to make use of original work that is owned by someone else, including the ability to quote some portion of the work verbatim, while the owner retains control of her work.

It's not a perfect solution. Far from it. But as this report will demonstrate, it’s a big improvement over the ad hoc systems that preceded it.

The core principles of copyright are bound up with foundational concepts that pertain to innovation, creativity, and communication, such as: authorship, control of publication, the right to reproduce and distribute works, and the free speech and censorship. As the next article will show, these topics have been debated fiercely for centuries.

These factors make copyright rather complicated. Yet they also make copyright a fascinating window on the interplay of technological innovation and human creativity, and the evolution of civilization. That’s what we are focused on in this series of essays.

The tortured history of copyright regulations and laws also illuminates potential outcomes for the regulation of artificial intelligence.

To understand how copyright might pertain to the advent of GAI, let’s start with the historical context. Where did copyright come from? Why was it necessary?  How did copyright affect society and technological innovation?

In the next article, we will begin a dive deep into those questions.

Disclaimer: I am not an attorney. This article is not intended to provide legal advice. Besides, you should never take legal advice from some random stranger on the Internet. If you are dealing with copyrighted material, or if you are concerned about fair use in generative artificial intelligence, it is best to consult with a legal professional.

This is the second in a series of articles that consider copyright in the context of generative artificial intelligence. The next article in this series will examine what Europe was like before copyright laws were introduced: it was an ugly, brutal, epically violent place.