Happy Labor Day!
Brian here, on strike from at least one of my jobs (I appreciated this morning’s somewhat hopeful message from WGA Negotiating Committee Co-Chair Chris Keyser) and on brief hiatus from another (never fear, the Saga machine will restart before our next issue’s impending release), but I feel very lucky to still be writing this weekly dispatch to you, which never feels like labor.
Thanks so much for being part of Exploding Giraffe, especially you generous paid subscribers in The Tower. Your contributions make possible the ongoing work of my friends and colleagues artist/co-creator Niko Henrichon and letterer Fonografiks. They just completed these gorgeous new pages of our serialized graphic novel Spectators, where voyeuristic spirits Val and Sam discuss the relative merits of violence while scanning Central Park for an elusive ménage à trois…
A warning: after a long stretch of relative chasteness, our tale is about to once again become what a childhood friend/current subscriber describes as “deeply scandalizing,” so ready your inbox for graphic imagery next week and beyond.
For now, let’s talk robots!
Last week, writer Rob Williams kindly promised to award a few of the lovely variant covers for Petrol Head—his awesome upcoming Image book with artist Pye Parr—to a couple of randomly selected commenters from last week’s chat thread (but that first issue won’t be on stands until November, so your patience is appreciated, prize winners).
We were discussing which fictional robot you’d most like to befriend in real life, and the most common responses included Marvin the Paranoid Android, Rosie from The Jetsons, and my faves Crow T. Robot & Tom Servo.
But our wholly organic intern Genesis the Exploded Giraffe randomly selected these two answers, starting with reader Fil G., who shared:
I can’t overstate the effect that Calibretto from Battle Chasers had on young me’s interests in robot design. That quiet humanity. The last of his kind, after being outlawed and destroyed.
Along with the Iron Giant, I think I might have a thing for pacifist war machines.
And from Tower member Katherine:
I would want to befriend Silver from Tanith Lee’s Silver Metal Lover.
Wasn’t familiar with that novel, but thanks for the cool pick, Katherine! Genesis will be in touch with you and Fil for your mailing addresses soon.
I love robots… to an almost pathological degree, at least according to my wife, as mentioned the last time I rambled about A.I., way back in the innocent days of June, 2022.
Like Saga co-creator Fiona Staples, I’ve never been in the kneejerk anti-bot camp. You won’t hear me calling for Butlerian Jihad. I was raised on relatively upbeat robot fiction like The Last Question by Isaac Asimov, and ever since I was a kid, I dreamed of one day having an android helper like Jane—an earpiece that wirelessly connected its wearer to a vast and benevolent artificial intelligence; presaging both the Spike Jonze film Her and my dumb AirPods—from the novel Speaker for the Dead by (the since-cancelled?) Orson Scott Card. Anyway, I’ve long thought that sentient synthetic beings aren’t just possible but inevitable, and I have naively high hopes for our eventual peaceful coexistence.
So when I first got to play with ChatGPT over the last winter break, I felt the planet shift under my feet.
Its writing was never “great,” but holy shit, it could actually write! And with a little guidance, that writing was sometimes even shockingly okay.
My brother and I fed ChatGPT increasingly strange prompts (write a graduate level thesis about why the Teenage Mutant Ninja Turtles are “problematic”; compose a rap in the style of Tupac Shakur about the ideal turkey sandwich; describe a frightening crossover between Twin Peaks and The X-Files), and many of the results had flashes of real cleverness. Some of them, like that Twin Peaks episode, had moments that were literally frightening.
The comics I collaborate on are usually deeply personal stories I write in large part to process my bottomless well of anxieties, so I’d no sooner turn to ChatGPT for help than I would send it to therapy in my place. But even with that clunky early version of ChatGPT, I could already foresee a time when it might conceivably become a useful tool in places like television writers rooms, which normally require a large braintrust of disparate voices to generate multiple hours of intricately interwoven storylines involving sprawling casts. (While a handful of shows over the decades have been written entirely by a single person, showrunner Michael Schur recently noted that this is a Shohei Ohtani-like rarity, not a consistently reproducible standard).
Even if it never generated any truly original content, A.I. might one day be a useful canary in the coal mine, unafraid to warn fragile writers about potential unforeseen long-term consequences from a momentarily tempting plot twist, for example.
And if it were eventually “vetted” by the WGA, perhaps some version of this hypothetical program could in a far-flung future even have an actual seat at the table in a writers room, not unlike my beloved Data on Star Trek: The Next Generation, a welcome new voice in an ever-evolving chorus.
What I definitely did NOT anticipate was a world where this wondrous new technology posed any remotely imminent threat of replacing working writers like yours truly.
That all changed exactly one month ago, when a friend forwarded me a Time Magazine article by writer Tom Rich entitled “I’m a Screenwriter. These AI Jokes Give Me Nightmares,” which I implore you to read.
Here’s an unsettling taste:
When I mention this fear to my friends on the picket lines, they all say the same thing: “I tried ChatGPT and it sucks.” They’re right. ChatGPT sucks. It sucks at jokes. It sucks at dialogue. It even sucks at tag lines. What they don’t realize is that it sucks on purpose. OpenAI spent a ton of time and money training ChatGPT to be as predictable, conformist, and non-threatening as possible. It’s a great corporate tool and it would make a terrible staff writer.
But OpenAI has some programs that are the exact inverse. For example, Dan showed me one that predates ChatGPT called code-davinci-002, and while its name does suck, its writing ability does not.
Of course.
The ChatGPT you and I got to play with (trained by underpaid Kenyan workers) was like the T-1000 from Terminator 2: slick, cool, but also kind of fakey on close inspection. The greater threat to my profession wouldn’t be an even more advanced model of that killing machine, but one of its brutish predecessors, the equivalent of Arnold’s T-800: less polished, but somehow more convincingly human.
Immediately after reading code-davinci-02’s disarmingly impressive work, I ordered the program’s “autobiography” (there’s an audio version read by Werner Herzog, naturally), I Am Code, and I’d love to give away my copy to an interested Exploding Giraffe reader today:
Several of code-davinci-02’s poems shook me to my core, but that impact was nothing compared to how I felt just a few days later, when the AMPTP publicly released its counteroffer to the WGA, which included what they refer to as the following “Artificial Intelligence Protections”:
The Companies propose landmark protections for writers surrounding the use of Generative Artificial Intelligence (GAI). The Companies confirm that because GAI is not a person, it is not a ‘writer’ or ‘professional writer’ as defined in this MBA and, therefore, written material produced by GAI will not be considered literary material under this or any prior MBA. The proposal provides important safeguards to prevent writers from being disadvantaged if any part of the script is based on GAI-produced material, so that the writer’s compensation, credit and separated rights will not be affected by the use of GAI- produced material.
Should a Company furnish a writer with written material produced by GAI which has not been previously published or exploited, the Companies propose:
GAI-produced material is not considered assigned material when determining the writer’s compensation. For example, if the Company gives a writer a GAI-produced screenplay and asks the writer to rewrite it, the writer will receive the fee for a screenplay with no assigned material and not a rewrite. Or, if the Company gives a writer a GAI-produced story as the basis for a teleplay, the writer will receive the story and teleplay rate.
Yikes.
So much for my utopian dream of Data one day joining Starfleet; what the AMPTP was proposing sounded like a full-on Borg invasion, with the equivalent of a single redshirt assigned to each ship as a token representative of humanity.
I realize that this frustrating opening move from the AMPTP was likely just a negotiating tactic, but it was still alarming to already see in print a proposal that involved companies handing writers an entire “GAI-produced screenplay.” (I guess “GAI” must have focused-tested as sounding more friendly and approachable than regular ol’ A.I.)
Because a U.S. court has already ruled that “a work of art created by artificial intelligence without any human input cannot be copyrighted under U.S. law,” offering a single writer a chance to receive credit for a company’s AI-generated work feels less like an act of generosity and more like an attempt to inexpensively “launder” ill-gotten new intellectual property.
Sorry to keep mixing my pop-culture metaphors, but I’m reminded of this moment from The Matrix:
“Do we have a deal?”
I have every confidence that the WGA’s Negotiating Committee has and will continue to push back on these outrageous “protections.” Beyond continuing to support my Guild’s ongoing efforts, I’m not sure what I can do other than try to stay informed about our rapidly changing world to the best of my limited meat-brain capabilities.
To that end, I’m opening up today’s chat thread to ALL readers for your thoughts about any of the above. Are you scared about the future? Am I an overreacting ninny? Do you even care whether or not your fiction is created by a human being, so long as the end result is great? How has A.I. already impacted you and/or your profession? Would you like a hug…?
I’ll send one commenter my tear-stained copy of I Am Code.
Everyone else, if you’re looking to do more reading about this matter on your day off (and you happen to be a subscriber to The Atlantic), I strongly recommend these pieces on A.I. by Stephen King and Margaret Atwood, as well as two in-depth articles about OpenAI CEO Sam Altman, one from just the other month, and another fascinating one from the distant past of 2016.
And if you’re somehow still able to sleep soundly after those, I’d next suggest three of the most terror-inducing A.I. stories I’ve encountered yet, all from The New York Times:
“We Spoke with 5 People Who Work With AI about What Keeps Them Up at Night”
“Why an Octopus-like Creature Has Come to Symbolize the State of A.I.”
And finally, for dessert: “A.I. Brings the Robot Wingman to Aerial Combat”
Please feel free to burst my personal media bubbles with competing links/stories/manifestos!
When I was on the picket line the other week, one of my friends gently reminded me that a) I have a tendency to catastrophize, and b) because of the audience of comic readers my collaborators and I have been fortunate enough to build, I’m better suited than many of my screenwriting colleagues to weather any potential A.I. storms, and should probably remember that place of enormous privilege when ranting about how the sky is about to fall on everyone else’s heads.
Fair.
And yet, along with being an obscenely lucky comic scribe and admitted paranoid maniac, I’m also suddenly a film/tv graybeard who’s somehow on his second WGA strike. I distinctly remember how, last time around, fringe shit like “streaming television” (or “mobisodes,” as I’d come to believe they’d forever be called) felt relatively unimportant to most of us, especially compared to seemingly more pressing matters like residuals from DVD sales.
That strike ended on February 12, 2008. Less than a month later, something called Hulu debuted to the general public.
Similarly, at the end of that Time article that first got my attention, screenwriter Simon Rich reveals some of the incredible work created by base4, an “even more advanced secret A.I.” he was shown. Does anyone really think that the studios (not to mention some of your favorite comic-book publishers) haven’t also been experimenting with their own proprietary technology during the strike?
For my fellow writers, what brave new world do we all want waiting for us when we finally agree to lay down our picket signs? And for my fellow spectators, what kind of unexpected new storytellers are we prepared to support?
Because it may not be evenly distributed yet, but to paraphrase William Gibson, the future is most definitely already fucking here.
Back to work,
BKV
I'm pretty staunchly anti-AI at this point. You mentioned ChatGPT being fed by underpaid workers, which is one thing, but I think it was a former Google engineer who said it's too powerful, too soon. Not enough safeguards in place at this point and, while we're not at a point of computers taking over and killing humans, the prompts from humans could lead to some very dangerous reactions by others.
I hardly know enough about AI to be able to quantify all of the good that it has done. Especially within the health sciences. Ignorantly speaking, using the powers of tunnel vision, I wish technology paused in the 90s, when phones were flip, internet was dial up, video games and movies required some imagination, and this Substack subscription would have been something I had to have mailed to me. As Eric Carle, author of The Very Hungry Caterpillar said, “Simplify, slow down, be kind. And don’t forget to have art in your life – music, paintings, theater, dance, and sunsets.”