Vés enrere Epitaph of the author

Epitaph of the author

14.05.2024

Imatge inicial

In discussing the role Artificial Intelligence (AI) has had and will continue to have in our society, various points can be brought up. On the one hand, numerous successes can be attributed to the support it has provided to a whole generation. It has not only made several tasks easier and opened a wide array of new possible tasks but has also even been able to generate creative productions. However, on the other hand, AI’s development and implementation has sparked key debates and questions about human creativity, identity, and authorship.

I would like to focus in on this last part in this essay. I will debate whether perhaps part of its role in society is to liberate us from traditional understandings of identity and self-worth. Perhaps, it marks a shift away from a predominantly personal production-based conception of worth, to a society in which we value other aspects of our life and acknowledge the use of tools available to us. First though, it is worth addressing the legal implications of AI’s dissemination.

Legally, AI’s creations present some challenges worth addressing. When determining ownership of works generated using Artificial Intelligence, various perspectives emerged. Some argue that the authors of the texts or images used to train Chat hold legal rights over the outputs, while others insist that the creators who integrated AI into their final works retain copyright regardless of the initial input (D’Agostino, 2023).

There are those who contended that as generative AI software programmers cannot claim responsibility over their productions, they should not be credited as an author (Crawford et al., 2023). As a matter of fact, most academic journals firmly believed that AI should not be listed as an author in publications (Lund & Naheem, 2024). Understandably in order to preserve scientific credibility and trustworthiness, regulations and policies have been having to adapt to the increasing use of AI in publications (Lund & Naheem, 2024, p.17). 

If authorship in specific jurisdictions requires a significant scholarly contribution, generative AI productions would be exempt, as their role is so far limited to curating existing knowledge rather than generating entirely new and unique content (Crawford et al., 2023). Nonetheless if we were to consider AI in matters of authorship, to what extent should have the knowledge inputted into the software, reflected in its creations, be recognized, and protected by copyright? (D’Agostino, 2023).

In terms of the film industries, large language models which generate AI images have been accused of intellectual property theft. According to this understanding, the practice of scrapping the images which feed the software have copyrights which are not being acknowledged. However, for creators who use AI, their argument is centred on a nation of ‘fair use’. Through this they contend that as the final product is a transformation they don’t need to account for the copyrights of the original inputted images. (Cho, 2023)

This matter could and will be discussed in length over the next few years, on legal, ethical, and philosophical fronts. However, in this text I would like to question if AI’s development and popularization has challenged key notions of how we identify as humans. 

The question of authorship in its traditionally romantic sense of the word may have lost relevance in the context in which AI is developing. Copyright law, as in the legal link between a piece of work and the author, and the concept of a romantic author, as the sole creative maker of an organic production, were developed simultaneously during the 18th century (Bently, 1994).  

These ideas are inherently tied to concepts of ownership and natural right, leading to the perception of a production being the sole property of a proud author (Craig & Kerr, 2020). Apart from being an ideological tool used to established and reinforce certain inequalities, through its claims of sole right of ownership in an age of information (Craig & Kerr, 2020), it is also a conception that has underpinned most of our creations and to an extent, motivations, as humans since then. 

Previous authors have questioned this relationship. Foucault explained that through the development of copyright law, literary and consequently creative productions, entered the social order of production which govern our culture (Craig & Kerr, 2020). He envisioned it as, yet another discursive tool used to exercise power (Craig & Kerr, 2020). 

Roland Barthes, built on these ideas and questioned the potential “death” of an author. He stated that once a text is published, the focus should longer be on the sole dominion had by an author and should be understood as the produce of several other texts (Bently, 1994). This perspective places the emphasis on the content itself, and its subsequent interpretations, rather than on the intentions of an author. 

I realize that it is ironic that we still refer to these authors as the authors of these ideas when they themselves challenge this type of authorship, but regardless it is interesting to bring them up in debates of artificial intelligence. Much alike what Barthes claimed, AI’s creations were the product of other inputted texts. Should the focus be on who wrote them? Or what it says? 

Of course, legally it is important that someone’s efforts are recognized and fairly compensated, and as such the question of the copyright in AI is still relevant. We should ensure that those in power who developed and own the software, don’t use these arguments as excuses to have claims over other people’s labour. But does the romantic idea of authorship still hold value?

What if AI, marked a philosophical shift in how we identify as humans? Perhaps the notion of production and authorship through which we traditionally value ourselves and other people by can no longer be seen as a defining factor. This may also distance us from certain capitalist notions of private property, or in fact reinforce other side effects of capitalism and end up with powerful organizations exploiting hard work and labour of unrecognized producers. Nonetheless, I wonder whether with the lack of an individual author we will transition to more collective understandings of society, distancing ourselves from the traditional hyper-individualistic understandings prominent in western capitalist societies.

Overall, maybe AI’s role philosophically, frees us from notions of a restrictive and singular author. Through this we could eventually focus on more than just ensuring ownership over an idea and actually focus on the output and its contribution to the vast body of knowledge that informs our societies. 

As for these implications on the film industry and our wider contemporary creative scene, it may be the case that we progressively return to previous storytelling traditions. Myths, whose narratives have guided our societies for centuries have no strict authors and are clearly a product of countless other stories used as inspiration. Not to say that we will go back completely to a society where storytelling has no record of the production process, but we may still transition to a place in which the story itself takes a central role rather than the author behind it.

In discussing copyright law around AI, we are prompted to reconsider fundamental notions of identity, creativity, and ownership. Not only has it challenged some of the legal structures underpinning western law but has also made us question matters of identity and whether we should start to value the output over the author. At the end of the day, whether a text is created by a human or a machine, it often receives unrecognized inputs and influences. How we negotiate this legally, ethically, and philosophically is a key question for the coming years. 
 

 

References
Bently, L. (1994). Copyright and the Death of the Author in Literature and Law. Modern Law Review, 57(6), 973–986. https://doi.org/10.1111/j.1468-2230.1994.tb01989.x


Cho, W. (2023). Scraping or stealing? A legal reckoning over AI looms. The Hollywood Reporter. https://www.hollywoodreporter.com/business/business-news/ai-scraping-stealing-copyright-law-1235571501/?fbclid=IwAR0s49S9LRegkigxWC3dIDKPbsRhPwMVk8POwwXA_qUFS_9VTA9dWdd1HfY 


Craig, C., & Kerr, I. (2020). The death of the AI author. Ottawa Law Review, 52(1), 31–86.

Crawford, J., Cowling, M., Ashton-Hay, S., Kelder, J., Middleton, R., & Wilson, G. S. (2023). Artificial Intelligence and Authorship Editor Policy: ChatGPT, Bard Bing AI, and beyond. Journal of University Teaching & Learning Practice, 20(5). https://doi.org/10.53761/1.20.5.01


D’Agostino, S. (2023). Ai raises complicated questions about authorship. Inside Higher Ed | Higher Education News, Events and Jobs. https://www.insidehighered.com/news/tech-innovation/2023/08/22/ai-raises-complicated-questions-about-authorship


Lund, B. D., & Naheem, K. T. (2024). Can ChatGPT be an author? A study of artificial intelligence authorship policies in top academic journals. Learned Publishing, 37(1), 13–21. https://doi.org/10.1002/leap.1582
 


Michael Poots is a MA student in Digital Culture and Emerging Media at the Pompeu Fabra University, currently writing the thesis on the intersection between spirituality in social networks and neoliberal governmentality. Half English and half Argentine, he has previous experience in International Relations.

Multimèdia

Categories:

ODS - Objectius de desenvolupament sostenible:

Els ODS a la UPF

Contact