The Long-term Worries of Editors
Compared to the diminishing bulging of their wallets, the complete loss of hope truly terrifies and fuels the determination of writers to resist the end.
In late November 2022, OpenAI unveiled the groundbreaking generative AI service, ChatGPT. The strike, in response to ChatGPT, began five months later. ChatGPT engages in conversations, writes articles, and even crafts various stories under specific clever prompts in collaboration with humans.
The arrival of ChatGPT, akin to a bolt from the blue, foretells the fate of all creative industries being disrupted by artificial intelligence.
Capitalists, naturally, won't miss out on this "low-cost, high-output" production tool. According to insiders, Warner Bros., one of the traditional "Big Five," has entered discussions with OpenAI on leveraging ChatGPT.
For instance, Warner Bros. has already started using ChatGPT-generated program summaries on its Max streaming service. In the future, AI might generate script overviews/summaries for film and television projects, enhancing executives' efficiency in project understanding.
Internet companies like Netflix, purely based on the web, won't miss the opportunity for "cost reduction and efficiency improvement." According to leaked traces on a headhunting website, Netflix is recruiting product managers in artificial intelligence, a position expected to impact the streaming platform's future strategic decisions significantly.
While these attempts haven't yet reached the core of creation, writers know that the day when AI challenges core creative work is inevitable. To resist the "erosion of AI into the realm of human creation," they must take action.
The construction of artificial intelligence systems relies on three principal foundations: computing power, algorithms, and data.
Computing power is abundant and can be acquired with money. Algorithms, the mathematical models behind generative AI, have seen significant breakthroughs, with ready-made pre-trained models quickly adapted to specific creative applications. The only limiting factor for the industry is data. Without high-quality training data, even the most sophisticated AI systems remain "empty-headed idiots."
For the content, especially in the film and television industry, high-quality data translates to existing scripts—scripts of various types and styles.
The current capabilities of AI systems still need to be improved by an experienced writer. However, if AI collaborates with writers, allowing them to modify AI-generated scripts and optimize the AI system in return, it could significantly accelerate the "intelligence" of AI systems in scriptwriting.
Hollywood writers are mainly unwilling to contribute their work as "raw data" for AI model training. They fear becoming obsolete and discarded by the media industry like "waste."
Therefore, one of the main demands of the writers' strike is that "AI-generated storylines will not be considered 'literary material,'" and "production companies must disclose whether AI entirely or partially generates the provided original material." Writers acknowledge their inability to stop the trend of media giants using AI to generate content. They hope they can refuse or file lawsuits if their works are used for AI training.
AI companies argue that the ultimate goal of these generative AIs is not to replace creators but to liberate them from non-creative work, allowing them more time for creative aspects. Cristóbal Valenzuela, CEO of Runway, an AI video startup that received a $140 million investment from Google and NVIDIA, states:
"When it comes to AI creating videos, we often hear terrifying or fairy-tale-like rumors... But the actual applications of AI are much more nuanced. You can't just input 'make me a movie,' it will create a whole movie."
Studios claim that AI tools can speed up the production of storyboards, essential tools in series creation that use graphics to showcase how the story unfolds and the corresponding scene designs. After completing the basic storyboard, AI systems can quickly fill in the required details based on the story and scene designs, accelerating the creative process. This work required a significant amount of time from junior production staff.
Besides expediting the primary production process, AI systems can also assist in fine-tuning the final product. For example, producers can adjust scenes using digital models of actors without the need for reshooting. This includes turning someone's head or slightly altering their expression. Convincing actors to leave behind their digital models poses a new challenge.
Another application type is dubbing actors in different languages. AI can make audio sound like it has an Italian or German accent or style, even adjusting the lip movement of actors accordingly.
With enough input data, can AI models produce entirely new movies, or even just a scene or setting, without the involvement of real actors? This thought is both exciting and chilling, depending on your perspective. AI companies disagree with the opposing views of some Hollywood practitioners, as their purpose is to replace others, not be returned.
The Compromise
Fortunately, after nearly 150 days of the strike, labor and management finally reached a consensus and announced a temporary agreement on September 25, 2023.
In October 2023, the WGA approved this agreement with an overwhelming majority, officially concluding Hollywood's most extended labor dispute and ensuring both parties could safely collaborate over the next three years. Based on the currently disclosed content, both parties made some concessions, with the management making slightly more significant compromises.
As AI technology is still in its early stages, its application scope and depth are unpredictable. According to the agreement reached so far,
Both parties acknowledge the uncertain and rapidly developing legal environment surrounding the use of generative AI.
The management and WGA agree to hold at least two meetings annually within the contract's three-year term to discuss new issues related to AI applications.
If generative AI technology makes significant breakthroughs in the next three years, management and WGA will likely engage in a new struggle over AI applications. Three years later, the scriptwriters' strike may reappear. Based on my understanding of the development of generative AI, the likelihood of such breakthroughs is very high.
We have entered an era where AI disrupts everything. Please buckle up and hold on tight.
Share your thoughts!
Be the first to start the conversation.