AI Character Bibles - (Retrospective)
The good, the bad, and the unethical use of AI in the creation of Character Bibles
I'd like to extend my sincere appreciation to all who have supported me during this week's exploration of the AI Character Bible sprint. It has truly been an enlightening journey, eliciting a wide range of responses. Some have shared concerns or reservations (and a few potty-mouthed people have said very bad things to me), while others have expressed an appreciation for the exploration I’m attempting. Regardless, I remain undaunted by the naysayers because I believe that understanding these dynamics is essential as we shape the future together.
With this particular Character Bible sprint behind us, this episode of "The Brief" will evaluate ChatGPT's merits and limitations, especially in the context of Character Bibles. I'll also discuss the techniques that, for me, align with ethical standards, those that hover in a gray area, and practices that could potentially breach artistic, ethical, and legal guidelines.
First up - is it useful?
The utility of AI in crafting Character Bibles is multifaceted. First, AI's ability to analyze vast data sets can unearth unique character traits that might be overlooked by the human mind. Anecdotally, a writer once shared how an AI proposed a character with a mix of traits from two distinct historical figures, leading to a compelling protagonist for a historical fiction piece. Second, AI ensures consistency, automatically tracking and adhering to established character profiles throughout a script. Third, for writers seeking rapid prototyping, AI can generate multiple character arcs in moments, aiding in decision-making. Fourth, its non-emotional analysis can help in balancing character ensembles, ensuring a diverse cast without unconscious bias. Lastly, AI provides an organizational tool, streamlining the management of complex casts in intricate plots. The central truth here is that AI can be an invaluable assistant in the mechanical aspects of character creation.
A THEME TO REMEMBER. Each new era of technology initially starts as a threat to creativity, but as humans adopt that particular technology, human creativity actually expands as a result, and that new technology eventually settles into a place of utility for the creator - humans.
In my humble opinion, those writers who harness the capabilities of AI for crafting Character Bibles will likely find themselves ahead of the curve. Initially, there may be resistance. It's conceivable that while many writers might publicly criticize AI's role in creative processes, behind closed doors, they might be secretly leveraging its capabilities. However, as with many technological advancements, over time (as in 10 years), the integration of AI into this specific domain will likely be perceived not as a threat but as an obvious efficiency tool. It will streamline and enhance the process of character creation, bringing consistency and broadening horizons. Yet, it's an enduring truth that the soul, the magic, the quintessence of a character, will always be a human endeavor.
Ethically, however, it’s a bit complicated…
There’s some obvious high level ethical concerns. If not carefully trained, AI can perpetuate and even exacerbate existing biases, leading to characters that uphold harmful stereotypes. The depersonalization of the creative process also looms as a concern; while AI can generate characters, it may not replicate the deep emotional connection a human writer forms with their creations. Further complicating matters is the murky realm of intellectual property. Determining ownership rights of AI-generated characters poses a challenge that intersects legal and ethical domains. The question of whether using AI in this manner aligns with artistic integrity is also pertinent, raising the debate on the balance between technological innovation and genuine, human-crafted artistry.
But these high level analyses have limits - they are just postulations.
While high-level discussions and theoretical debates about the ethical, artistic, and legal implications of AI provide valuable insights, there's an unparalleled depth of understanding that comes from hands-on application. For me, using AI to create the Character Bible for Phi was an enlightening endeavor, tinged with a mix of empowerment and unease. At moments, the process made me feel unstoppable, as if I had tapped into a reservoir of endless creative potential. However, there were also instances when I felt as though I was trespassing on the creative territory of others. At its worst, I questioned whether I was compromising my artistic integrity by relying on the machine.
Ethically Acceptable Uses of AI for Character Bibles
Generating initial character trait suggestions for inspiration - whenever I asked ChatGPT for a singular suggestion for a trait I felt like I wasn’t being the artist. But when I asked ChatGPT to provide ten possible suggestions, I felt like I was having a conversation with a knowledge base - one where I was in control.
Creating character babies from historical figures. Amalgamating a character from one or more historical figures, in my humble opinion, is ok. In fact, I think it’s more than ok - it’s beneficial, because it helps anyone who does this to learn and share our shared history more deeply through characters in films.
Analyzing vast historical data for accurate character backgrounds - I cannot stress how much I have learned about history, culture, and people as a result of using ChatGPT in researching Phi as a character. I am absolutely addicted now to history - and that’s thanks to ChatGPT.
Offering organizational tools for complex character interactions - I love, love, love that ChatGPT can keep Phi’s Character Bible in memory and use it as an expositive, constraint, and character arc state knowledge-base. There is NO WAY I could come up with anything as effective at AI to manage this.
Providing consistency checks for characters' actions and motivations. Not only is ChatGPT a knowledge-base for state - it’s something that can be put in action, like a writers assistant - more on that later.
Murky Territory in Using AI for Character Bibles
Drawing heavily from existing fictional characters without significant modification - When I did the creating character babies exercise where I created phi as an amalgamation of three characters from existing movies - YUK! I felt like I was trodding on the work of those writers. I felt like I didn’t care about Phi, about making her unique, and most importantly make her important to me.
I felt so strongly about this that I went back and actually rebuilt Phi from scratch and removed all amalgamations from other creative works.
Over-relying on AI suggestions without personal creative input. Two things on this one - as the writer of Shepard’s Tone, I think I have a responsibility to be an actual creator, and not just a curator. On this note, please know that I’ve been working on Shepard’s Tone for years, and it’s only with AI and modern day CGI that I have decided to take on creating the thing. The bones of everything you are reading about Shepard’s Tone was created by me and when I did use ChatGPT to help me, there was never a moment where I didn’t modify what ChatGPT suggested, in order to make it my own.
Using AI to mimic the specific style of known authors or creators. This one is a bit more difficult for me. When two humans are working together to creatively write a screenplay, I can imagine that almost everyone has “style” conversations. So what’s different about having style conversations wit AI? For me, I think there is a line between using ChatGPT to discuss different styles and ordering ChatGPT to amalgamate something from different styles.
Generating character backstories that may inadvertently perpetuate stereotypes. Yikes. This one was unexpected for me, but I really felt like ChatGPT was filling Phi’s character to way too many cultural tropes. I’m now going back and removing those tropes and creating unique features that feel more true to me.
Leaning on AI to resolve complex character arcs without human intuition. This one is more complex for me. I’m sooooo addicted to having ChatGPT help me with this one - just the pure utility of it is a massive opportunity. However, I can feel it in my bones that I’m going to have to watch out for this one - I’ll need to be the artist at the end. I’ll need to limit ChatGPT to utility, and not hand over the artists brush to it.
Unacceptable Uses of AI for Character Bibles
Directly lifting or plagiarizing characters from known works. You can’t rip the Character Bible of a character you find interesting in an existing movie, and simply change their name, or sex, and plop them in your movie.
Creating character babies from one or more characters in existing creative works. While amalgamating historical figures is ok in my book, amalgamating characters created, and owned by the way, by someone else - that’s not ok. Every character should be a unique work of art. Reminder - I’ve gone back and removed this type of amalgamation in Phi’s Character Bible.
Creating the physical appearance of a character by blending two or more existing actor’s appearances. Actors should own their images, whether real or CGI - period.
Ignoring AI-generated biases without human oversight or correction. Each LLM that we use inherently has biases - and that’s scary. While I’m a technologist and deeply understand this one - I don’t think the average writer has a clue as to what, and perhaps more importantly, how this happens - and further how to avoid it. This one, while seemly not that big of a deal right now, will be one of the biggest influences on society in the future. I urge each of you to deeply think about this one, and deeply think about taking a position on it.
Claiming AI-generated content as wholly original work without acknowledging AI's role. Transparency is everything when utilizing AI when creating a film. I firmly believe that viewers have a right to know. Many feel that any use of AI is unacceptable and that no viewer should want to watch a movie tainted by AI. My dial is slight more towards AI (but only slightly) whereas I think humans should embrace technological progression, but should do so carefully, and with forethought. I hope that everyone who is reading my posts appreciates the transparency that I’m embracing in creating Shepard’s Tone.
In our exploration today, we've delved deeply into the nuances of using AI for crafting Character Bibles, touching on the empowerment it offers and the ethical dilemmas it surfaces. This journey has been enlightening, highlighting both the promise and pitfalls of intertwining technology with artistry. As we move forward, I'm excited to embark on our next venture: utilizing ChatGPT to breathe life into the world of Phi. We will venture into the intricate realms of Aegis city, the desolate wastelands, and the mysterious underneath. This upcoming sprint promises to be as intense and revealing as our journey through Character Bibles. Stay tuned as we further demystify AI's capabilities and confront its ethical intricacies in the realm of world-building.
Thanks, and I look forward to sharing next week’s sprint with you.
What is The Brief and Who should read it?
I release a weekly digest every Friday, tailored for professionals ranging from executives to writers, directors, cinematographers, editors, and anyone actively involved in the film and television domain. This briefing offers a comprehensive yet accessible perspective on the convergence of technology and its implications for the movie and TV industry. It serves as an efficient gateway to understanding the nexus between Hollywood and Silicon Valley.
Who am I?
I'm Steve Newcomb. Functionally, I’m a recovering Silicon Valley founder that is finally old enough to have a bit of care. I’m perhaps most recognized for founding Powerset— it was the largest AI and machine learning project in the world when I founded it. It was later acquired by Microsoft and transformed into something you might recognize today - Microsoft Bing. Beyond Bing, I had the privilege of being on the pioneering team that witnessed the inaugural email sent via a mobile device. My journey also led me to SRI (Stanford Research Institute), where we laid the groundwork for contemporary speech recognition technology. Additionally, I was a co-founder of the debut company to introduce a 3D physics engine in Javascript. I've held positions on the board of directors and contributed funding to massive open source initiatives like NodeJS and even the largest such project, jQuery. My experience extends to academia, having been a senior fellow at the University of California, Berkeley's engineering and business faculties. Recently, I ventured into Layer 2 internet protocols and assisted a company named Matter Labs in securing $440 million in funding to bolster their endeavors.
What am I doing besides writing these posts?
Typically, I allocate a year between groundbreaking ventures. My exploration for the upcoming project commenced in May 2023, and the sole certainty is its nexus with the film, television, SMURF, and AI domains. Sharing insights on my research endeavors helps me discern between feasible prospects and mere illusions. My hope is that for this venture, I appropriately consider the ethical and sociological repercussions.
If you are interested in contacting me, being interviewed, being helped, or yelling at me, my email is steve.e.newcomb@gmail.com.