In Absalom, Absalom, William Faulkner writes “as though houses actually possess a sentience, a personality and character acquired not from the people who breathe or who have breathed in them so much as rather inherent in the wood and brick or begotten upon the wood and brick by the man or men who conceived and built them.”
Those words were written almost a century ago, yet we cannot deny their applicability to the current phase in the technological revolution. Now, no one in Faulkner’s day (nor any of us less than a decade ago) could have foreseen the advancements artificial intelligence has brought to every facet of media, much less every other industry.
But doubtless, we should still bear in mind the principles undergirding Faulkner’s words: how something is built reflects not just the character of who built it, but the character of the thing itself. Alongside the many enhancements AI continues to bring to society, its rapid adoption has not been without controversy.
Beneath the amazing power of AI’s capabilities lie widespread concerns about plagiarism, authenticity, and intellectual property. Using narrative intelligence, we can delve into the resistance AI faces in the media landscape, particularly from creators like newspaper publishers, authors, artists, and more.
One of the largest stories within the AI vs. publishing narrative is the national discussion about whether training large language models (LLMs) on copyrighted material constitutes fair use, or if it could be deemed copyright infringement.
The New York Times and UMG, who have both filed suit against OpenAI, claim that since LLMs like ChatGPT memorize and reproduce exact quotations from copyrighted material when prompted, this constitutes copyright infringement.
If a court agrees, this could spell disaster for generative AI. Proponents and stakeholders of generative AI claim that training an LLM on copyrighted material is fair use, as it is transformative and is akin to parody or quotation for the sake of discussion.
Another branch of this narrative covers AI art. Several companies are either utilizing AI-generated artwork or developing guardrails to protect themselves from having their work stolen. Adobe “had a record number of patents filed and issued last year” according to their general counsel Dana Rao.
On the flip side, entertainment companies like Wizards of the Coast and Square Enix have implemented AI-generated material in their promotions and, in the case of Square Enix, into their product directly. The video game developer has openly used AI for some of the visual elements of its upcoming title Foamstars. Specifically, Square Enix used the generative artificial intelligence program Midjourney to bring in ready-made visual assets to the game. Midjourney itself has been embroiled in lawsuits over claims that it is stealing IP from artists.
Additionally, the growing use of AI voice replication in music and voice acting remains a fear for many in the creative industries. One of the issues in last year’s SAG-AFTRA strike was the replacement of real voice actors in media with AI-generated voiceovers.
Every corner of the entertainment industry is grappling with the effects of AI. With any sort of published content, the focus is always on copyright. Irrespective of AI’s inevitable adoption by the public, the question of what constitutes fair use by developers and their programs will be of great consequence not just from a legal standpoint, but from the standpoint of the artists whose works are being used.
Is it despicable that an LLM uses content from a creator without consent or payment, or is it simply fair use? Where is the line? Those are some of the questions decision-makers around the world are contemplating, both in and out of the courtroom. Whether we reign AI in or let it roam free in the name of innovation, AI policy is being decided right in front of us.
Like the house in Faulkner’s novel, the builders of AI are responsible for the personality and character of their creation. As those who are living through the origins of AI’s mass adoption, we also have a say. How will we build this house? The answer is crucial; regardless of how the house is built, we have to live in it.
Want to gain a better grasp on narratives, like AI, that are shaping your world? EdgeTheory is dedicated to providing clarity in an overloaded information environment using AI-driven narrative analysis. From executive-level briefs to exhaustive data modules for analysts, EdgeTheory provides solutions that help you make the right decisions faster in the information environment.