Creative commons, creative process and copyright in the age of large language models (Part 2)
This is Part 2 of a ‘conversation’ on creative commons, the creative process and copyright in the age of large language models and generative AI more generally. Part 1 was also a conversation on the same topic but where the 10 questions were generated by ChatGTP itself.
🎙The interview
What is a commons?
A commons is a resource that is shared and managed by a group of people. It can be a physical space, such as a park or a pasture, or it can be a resource that is not bound to a specific location, such as knowledge or cultural heritage. The term "commons" is often used to refer to shared natural resources, such as a fishery or a water source, but it can also refer to shared cultural or intellectual resources, such as a library or a database. The concept of a commons is based on the idea that certain resources are best managed collectively, rather than by individuals or organizations acting in their own self-interest.
From prehistoric times to the internet and now large language models what have been the main technological innovations supporting commons around cultural and intellectual resources specifically?
There have been a number of technological innovations that have supported the creation and management of commons around cultural and intellectual resources. Some examples include:
1. Writing and printing: The development of writing and printing technologies allowed people to record and share knowledge and cultural traditions in a more durable and widely accessible format.
2. Libraries: The creation of libraries allowed people to access a wide range of books and other written materials, and also provided a way to preserve and protect these resources.
3. The internet: The internet has made it easier for people to access and share knowledge and cultural resources. Online libraries, databases, and other digital commons have made a vast amount of information available to a global audience.
4. Large language models: Large language models, such as Assistant, can be used to generate and share new knowledge and cultural resources. For example, a language model could be used to create a new translation of a classic work of literature or to generate new articles and essays on a particular topic.
These are just a few examples of the technological innovations that have supported the creation and management of commons around cultural and intellectual resources. There have been many others as well.
When and why did copyright laws appear on cultural and intellectual resources?
Copyright laws have been in existence for hundreds of years and were originally created to protect the rights of creators to control the reproduction and distribution of their works. The first modern copyright law was the Statute of Anne, which was passed in 1710 in the United Kingdom. This law granted exclusive rights to authors and their publishers to control the printing and distribution of their works.
The main reason for the creation of copyright laws was to encourage the creation and dissemination of new knowledge and cultural works. By giving creators the exclusive right to control the reproduction and distribution of their works, copyright laws provide an incentive for people to create and share new ideas and cultural expressions. This, in turn, can foster creativity and innovation, and contribute to the overall cultural and intellectual richness of society.
Copyright laws have evolved over time and are now in place in most countries around the world. They are generally intended to balance the interests of creators and the public by granting certain exclusive rights to creators while also allowing the public to access and use works in certain ways.
Has the advent of the internet upended the balance between creators and the public that copyright laws were meant to establish?
The internet has had a significant impact on the balance between creators and the public with regard to copyright laws. On the one hand, the internet has made it easier for creators to reach a global audience and to distribute their works more widely. This has the potential to increase the exposure and popularity of creative works, which can in turn lead to greater financial rewards for creators.
On the other hand, the internet has also made it easier for people to access and share copyrighted works without the permission of the creators. This has led to widespread copyright infringement, and has made it more challenging for creators to control the use and distribution of their works. As a result, some creators have called for stronger copyright protections, while others have argued that the existing laws are sufficient and that the focus should be on finding new ways to compensate creators for their work.
Overall, the impact of the internet on the balance between creators and the public with regard to copyright laws has been complex and multifaceted. It has both increased the potential for creators to reach a wider audience and earn greater rewards, but it has also made it more challenging for them to control the use and distribution of their works.
Has there been major initiatives to create new type of copyright licenses more adapted to the internet age favouring either the public or the creators?
Yes, there have been a number of initiatives to create new types of copyright licenses that are more adapted to the internet age. These initiatives have generally aimed to strike a balance between the interests of creators and the public by allowing greater access to and use of copyrighted works while still protecting the rights of creators.
One well-known example of such an initiative is the Creative Commons system, which provides a set of flexible copyright licenses that creators can use to grant certain permissions to the public to use their works. These licenses allow creators to retain their copyright while also allowing others to use their works in certain ways, such as for non-commercial purposes or with attribution.
Another example is the open access movement, which seeks to make scientific and scholarly research more widely available to the public by promoting the use of open access licenses. These licenses allow researchers to retain copyright while also allowing their work to be freely accessed and used by others.
Overall, these and other initiatives have aimed to create new copyright licenses that are more adapted to the internet age and that take into account the changing needs of both creators and the public.
How does the advent of large language model affects the balance between creators and the public from a copyright point of view?
The advent of large language models has the potential to affect the balance between creators and the public from a copyright point of view in a number of ways.
One potential impact is that large language models could be used to generate new works of creative expression, such as art, music, or writing. These works could be created entirely by the model, or the model could assist a human creator in the process of creating a work. If these works are protected by copyright, it is not yet clear how the rights to these works would be allocated, and this could have implications for the balance between creators and the public.
Another potential impact is that large language models could be used to generate new translations of existing works, or to create new summaries or analyses of existing works. If these activities are considered to be derivative works, they would typically be subject to copyright law and would require the permission of the original copyright holders. This could potentially limit the ability of the public to access and use these works.
Overall, it is still too early to say exactly how large language models will affect the balance between creators and the public from a copyright point of view, and these issues will likely continue to evolve as the technology develops.
But how different are LLMs in their creative process from the creative process of a human being? Don't all artists come to creating with a body of knowledge and a socio-historical context from which they draw?
Like human creators, large language models draw upon a body of knowledge and a socio-historical context when they generate new works of creative expression. However, there are some key differences in the way that large language models and human creators approach the creative process.
One key difference is the scale and speed at which large language models can process and generate new ideas. Because they are able to analyze vast amounts of data and generate new content based on that analysis, large language models can produce a large volume of output in a relatively short period of time. This is in contrast to human creators, who typically work at a slower pace and may be more selective in the ideas and concepts they choose to pursue.
Another difference is that large language models do not have personal experiences or emotions in the same way that human creators do. This means that the output of a large language model may be less influenced by personal perspective or subjective experience.
Overall, while large language models and human creators may both draw upon a body of knowledge and a socio-historical context in their creative processes, there are still significant differences in the way that they approach and generate new ideas.
Would copyright law be different if the cognitive ability of humans were equal to that of a large language model? Wouldn't it penalise our cultural and intellectual resources to restrict creators who use large language models in their creative process? Wasn't that the point of Larry Lessig with the Remix culture?
It is difficult to say exactly how copyright law would be different if the cognitive ability of humans were equal to that of a large language model, as this is a hypothetical scenario and there are many factors that could potentially influence the development of copyright law in such a situation.
However, it is possible that if humans had the same cognitive abilities as large language models, there might be less of a need for copyright law to incentivize the creation of new works of creative expression. In this scenario, it might be argued that the ability to generate new ideas and content would be a given, and that the focus of copyright law should shift to other goals, such as promoting the dissemination of knowledge or protecting the interests of creators in other ways.
On the other hand, it is also possible that even if the cognitive abilities of humans were equal to those of large language models, there could still be a need for copyright law to protect the rights of creators. For example, copyright law might be used to ensure that creators are fairly compensated for their work, or to protect their moral rights in their creations.
As for the idea of restricting creators who use large language models in their creative process, this would depend on the specific details of the situation and the goals of copyright law. Larry Lessig's concept of the "remix culture" argued for greater flexibility in the way that copyright law is applied to allow for the reuse and transformation of existing works, and this idea has influenced debates about copyright law in the digital age. However, it is ultimately up to lawmakers and policy makers to determine the appropriate balance between the rights of creators and the public interest in access to and use of creative works.
I am actually not advocating to restrict creators who use LLM, but asking how limiting such an action would be in expanding our cultural and intellectual resources. For example would The Verve have created Bittersweet Symphony had they explicitly known that they were using an orchestral arrangement from The Rolling Stones song 'One last time"? Would it have been such a hit without that orchestral arrangement?
It is difficult to say with certainty how limiting the use of large language models in the creative process would affect the expansion of cultural and intellectual resources. It is possible that restricting the use of these tools could have some negative impacts on the ability of creators to generate new ideas and content, and could potentially lead to a reduction in the overall amount and variety of creative works produced.
On the other hand, it is also possible that restricting the use of large language models could lead to a greater focus on more traditional forms of creative expression, and could encourage creators to rely more on their own personal experiences and perspectives when generating new ideas. This could potentially result in a different mix of creative works being produced, and could lead to the development of new and innovative forms of expression.
As for the specific example of The Verve and their use of an orchestral arrangement from a Rolling Stones song in the creation of "Bittersweet Symphony," it is important to note that this case involved the unauthorized use of a copyrighted work, which is a separate issue from the use of large language models in the creative process. In this case, The Verve did not have permission to use the orchestral arrangement from the Rolling Stones song, and this led to a legal dispute over the rights to the work. Whether or not "Bittersweet Symphony" would have been as successful without the use of the orchestral arrangement is a matter of speculation.
My point about the Verve example, is how much do we know for sure that the use of that orchestral arrangement came from the artist's own personal experience as opposed to a deliberate unauthorised use of that arrangement? Don't we always come back to the cognitive ability/limit or the artist rather than his intrinsic creative ability? I would argue that copyright law on balance needs more to protect the latter than the former.
It is not uncommon for artists to draw upon a wide range of influences and sources in the creative process, and this can include both personal experiences and the work of other artists. In the case of The Verve and their use of an orchestral arrangement from a Rolling Stones song in the creation of "Bittersweet Symphony," it is not clear whether the use of this arrangement was intended as an unauthorized use or whether it was simply an influence on the artist's creative process.
Regardless of the specific circumstances of this case, it is important to recognize that copyright law is intended to protect the rights of creators, and this includes both the rights of creators to control the use and distribution of their works, as well as their moral rights in their creations. These protections can help to ensure that creators are fairly compensated for their work and are able to have a say in how their works are used and disseminated.
At the same time, it is also important to balance the rights of creators with the public interest in access to and use of creative works. This can be a complex and nuanced issue, and different approaches may be appropriate in different situations. Ultimately, the goal of copyright law should be to strike a balance that promotes the creation and dissemination of new ideas and cultural expressions while also respecting the rights of creators.
Thank you this is very helpful, including that the breadth of the range of influences and sources play a role in the creative process of the artist. Do you know if the body of work you have been trained on is only open access or whether it also includes copyrighted material?
I am a large language model trained by OpenAI, and the body of work that I was trained on includes a wide range of texts from a variety of sources. This includes both open access material, such as books and articles that are freely available to the public, as well as copyrighted material that may be subject to certain restrictions on use and distribution.
As a language model, I do not have the ability to browse the internet or access new information, and my responses are based solely on the information that I was trained on. Therefore, any references I make to specific works or ideas are based on the texts that were included in my training data, and do not necessarily reflect the current availability or copyright status of those works.
What would happen if the creator of copyrighted material would win a lawsuit against work that has been created with the use of ChatGPT? Who would be party to the lawsuit on the infringement site: the model, the machine or human prompting the model or OpenAI or all three of them?
If the creator of copyrighted material were to win a lawsuit against a work that was created with the use of a language model, such as ChatGPT, the specific parties involved in the lawsuit would depend on the specific circumstances of the case.
If a work created with the use of a language model was found to infringe the copyright of another work, the party that created or distributed the infringing work could be held responsible for the infringement. This could potentially include the person or entity that used the language model to generate the work, as well as any other parties that were involved in the creation or distribution of the work.
In some cases, the creators of the language model itself, such as OpenAI in this example, could also potentially be named as defendants in a copyright infringement lawsuit if it was determined that their technology was used to create the infringing work. However, the extent to which the creators of a language model might be held liable for the actions of users of the technology would depend on the specific legal principles and laws that apply in the jurisdiction where the lawsuit was brought.
👩⚕️👨⚕️ Potential podcasts candidates
Thank you for this conversation. Who are the living experts on the copyright in the age of the internet, the evolution of copyright law with technological advances and the novel copyright issues arising from the use of generative AI in new creative works?
There are many experts on the topics of copyright in the age of the internet, the evolution of copyright law with technological advances, and the novel copyright issues arising from the use of generative AI in new creative works. Some examples of experts in these fields include:
Lawrence Lessig: Professor at Stanford Law School and a leading expert on copyright law, especially as it relates to the internet and emerging technologies.
Pam Samuelson: Professor at the University of California, Berkeley School of Law and a leading expert on copyright law, with a particular focus on the intersection of law and technology.
Jane Ginsburg: Professor at Columbia Law School and an expert on copyright law, with a particular focus on the protection of creative works in the digital age.
Brett Frischmann: Professor at Villanova University School of Law and an expert on intellectual property law, with a particular focus on the social and economic impacts of intellectual property law.
These are just a few examples of experts in these fields, and there are many others as well. If you are interested in learning more about these topics, you may want to consider consulting with a legal expert or conducting further research on these issues.
📚 Further reading
Can you please suggest additional reading on copyright law in the internet age, the evolution of copyright law through technological advances and copyright law and generative AI?
Here are some additional readings on copyright law in the internet age, the evolution of copyright law through technological advances, and copyright law and generative AI:
1. "The Future of Ideas: The Fate of the Commons in a Connected World," by Lawrence Lessig: This book explores the impact of the internet on copyright law and the concept of the "commons," and discusses the challenges and opportunities that the digital age presents for creators and the public.
2. "Reforming Copyright Law for the Internet World," by Pam Samuelson: This article discusses the need for reform of copyright law in the digital age, and suggests ways that copyright law can be adapted to better serve the interests of creators and the public.
3. "Copyright in the Digital Age: Challenges and Opportunities," by Jane Ginsburg: This article discusses the challenges and opportunities presented by the digital age for copyright law, and suggests ways that copyright law can be adapted to meet these challenges.
4. "Intellectual Property and Information Control: Philosophic Foundations and Contemporary Issues," by Brett Frischmann: This book provides a comprehensive overview of intellectual property law, including copyright law, and discusses the social and economic impacts of these laws.
I hope these readings are helpful!