A secret about screenwriting that only the tippy-top screenwriters know
One of the weirder things I’ve learned about screenwriting over the years is that there are things that work in a script that don’t work in a movie and there are things that work in a movie that don’t work in a script. Understanding why this bizarro process takes place can help you elevate your scenes – especially your dialogue scenes – to another level.
For example, montages work great in movies. Montages don’t work very well in scripts. That’s because movies are great for images and music. But you can’t see images and can’t hear music when you’re reading a script. In a screenplay, a montage is just a bunch of factual description a reader has to read through. Where’s the fun in that?
On the flip side, limited characters and limited locations work great in scripts because everybody who reads a screenplay wants it to read fast. So they love it when there’s any setup with 2-3 characters, minimal description, and a lot of dialogue. So if you have two characters trapped on a tower talking most of the time (“Fall”), a reader can get through that script in 30 minutes. It’s a dream read.
However, limited location movies, especially if they’re set somewhere stagnant, like a basement (the Duffer Brothers’ “Hidden”) become very boring on screen. That screenplay that read faster than lightning is boring to watch because it’s aesthetically boring to look at. Same background and little-to-no movement aren’t exactly cinematic.
This opens up a “4-D Chess” strategy to screenwriting because you’re now making choices about whether to write something that’s going to work in the script even if you know it’s not going to work in the movie (good idea for spec writers), or something that’s going to work in the movie even though it doesn’t work in the script (good idea for a writer who’s already hired). In other words, do you write for the ‘great read’ or do you write for the ‘great movie?’
It’s an interesting topic that I was thinking about the other day because I watched this show where, in a scene, two girl friends (I forgot their names but we’ll call them Sara and Nancy) needed to have a discussion about whether Nancy was going to marry her fiancé.
Whenever you have two people trying to work out a problem via dialogue, there’s a lot of variety in where you can set the scene. Theoretically, you can place a conversation anywhere. So, in this case, the writer placed it at a gym. Sara and Nancy were doing a workout routine with one of those giant tires that you flip over again and again. And, while doing so, they talked about whether the fiancé was marriage material.
The scene was actually okay, mostly because you were invested in the characters. But also because the movement and pauses and catching of their breath gave the conversation a little extra oomph.
However, I noted that, if you had read this scene in a script, nothing about this workout would affect how you read the scene at all. You could’ve set the very same scene at a cafe and it would’ve read the exact same. Because dialogue reads the same on the page whether your characters are working out or hanging out.
I’ll see this a lot, also, with characters doing the dishes together. Instead of placing your characters on a couch to have a conversation, you give them something to do, like dishes, while they have their conversation. Again, in a movie, that’s a good idea. But in a screenplay, it doesn’t matter where that conversation is had because we’re just reading the dialogue. We don’t care if Husband Joe properly places the egg-beater in the right dishwashing compartment.
The more I thought about this, the more I realized that the real problem here was that you wrote a scene that was so stale, you needed all this artificial movement — whether it be a giant tire workout or loading the dishwasher – to make up for the weak scenario. A scenario should never be so boring that it needs an extra on-screen boost.
Instead, you should be looking to write scenes that work on the page AND on the screen. To achieve this, you need a two-pronged approach.
First, you want the CONTENT of the scene to be good enough on its own that it wouldn’t matter where you placed it. You could have something as static as two characters sitting in bed together before they go to sleep. If the content of the conversation is compelling, we don’t care where it takes place, either on the page or on screen.
There’s no better example of this than Marlon Brando’s famous “I coulda been a contender” scene in On The Waterfront. It was the culmination of his entire life, this “contender” admission. So it was a very important moment. And, originally, the director, Elia Kazan, had this elaborate background projection sequence planned that would play behind the two characters as they had this conversation in the car.
But the projection broke, forcing them to film the scene with the back window closed, making it the most boring background ever. But it didn’t matter because the content of the scene was so strong. So make sure the content of the conversation is awesome first and foremost. That will take care of 75% of bad dialogue scenes.
Second, if possible, you want to use the physical scenario the characters are in TO ENHANCE the conversation, not just be incidental movement. Let’s say our married dishes couple is just getting things back on track after some infidelity by the husband. It still stings for the wife. So there’s pain there.
If these two had a dinner party with another couple and the wife thought the husband was too flirtatious with the guest wife, then that dishes scene is going to be a little more interesting. Just the way that the wife is handling the dishes. You could have her GRAB them and SHOVE them into the dishwasher to add an exclamation point to her frustration.
But guess what? You could make this scene EVEN BETTER if you used the scenario to enhance the conversation, not just use it as a minor visual distraction.
One thing you could do, for example, is to set the scene at a restaurant. The husband is taking the wife out to patch things up. But then, they get a really attractive waitress. And the waitress keeps coming in. She seems to be giving the husband more attention than the wife. This starts upsetting the wife. And now we’ve got a scene where, if you’ve taken care of the content part, we’re using the scenario they’re in to poke the bear – to fire up the coals beneath the dialogue. It becomes a much more interesting scene cause there’s an extra element going on.
I just watched this Portuguese show called “Gloria,” about the Cold War. Our hero, a spy who was transporting highly classified audio tapes across countries, regularly has to drive through checkpoints. But he knows all the checkpoint people so he’s able to shoot the sh— with them and move on.
In one of the scenes in the pilot, though, he’s chatting with a checkpoint friend but then, behind them, a new checkpoint guy strolls up. And this new guy decides to inspect the trunk, where our hero has hidden one of his audio tapes. So you’ve got your dialogue between the hero and his checkpoint friend. And then you’ve got the other thing going on behind them that, if it goes badly, our hero could be thrown in prison.
That scene’s going to work on the page and on the screen. Which is what I’m asking for here. First and foremost, make sure the scene works on the page. Cause you have no control over whether your script is going to get produced. All you have control of, at the moment, is making the read as good as possible.
Once you’ve got that down, see if you can improve the scene so that it also works on screen. Maybe it already does cause you got lucky. But if it doesn’t, see if you can find a way to create a scenario that enhances the good dialogue you already have. If you do that, you’re going to turn that good dialogue into great dialogue.
Genre: Thriller
Premise: On a private island off San Francisco, a nanny goes to work for a mother who is one of America’s most powerful tech entrepreneurs. Things slowly begin to devolve as the mother’s hyper-monitoring and surveillance become suffocating.
About: Another script that made last year’s Black List. Laura Kosann has won the Nicholl Screenwriting Contest. She wrote an interesting script I reviewed last year called, “From Little Acorns Grow.” Definitely a writer to keep your eyes on.
Writer: Laura Kosann
Details: 113 pages
Anna Kendrick for…. ANNA!
Give me an island and I’m a happy script reader.
I don’t know what it is about islands but the second I see one in a logline, my eyes light up. It’s probably because of the isolation – the fact that you can’t run away. Especially these days, with everything being so interconnected – with help being a pushed button away – islands are the last places in the world where you’re truly screwed if things go wrong.
I also love islands near giant cities. I’ve told myself for years I’m going to visit Catalina Island (a ferry ride away from Los Angeles – the island doesn’t allow cars, how cool is that!) and I still can’t get over the fact that there’s an island (called “Treasure Island”!!!) in the San Francisco Bay, just sitting there, that you or anyone can live on. Which segues perfectly into today’s script, because that’s where our island is located – in San Francisco. Except it’s not a public island. It’s private.
Ever hear of Elizabeth Holmes? The creepiest tech CEO in history? Well, wait until you meet Nicole. We’ll get to her in a moment. First we must meet Anna, a 30-something former tech engineer who left her job abruptly under mysterious circumstances. Now, she just focuses on nannying.
Anna gets a call for a prestigious nannying job that will require her to live in the mother’s house. When she shows up to meet the mommy she’ll be working for, she instead is greeted by a boat! A drone boat which tells her to hop on. Anna reluctantly does and gets shuttled across the bay to a glitzy home on a private island.
Here she meets Nicole, a care-free tech CEO with more money than South America. Nicole is in her 40s and has set her sights on using technology to eliminate miscarriages. She had a bunch of miscarriages of her own due to the stress of her job and wants to make it so that women can work in stressful environments and still have babies.
Aiden is her miracle baby, her only child, now just months old. And she needs Anna to take care of her while she works from home.
Aiden’s got some problems. His skin blisters in the sun, so he always has to be kept under a special netting when he’s outside. He spits up and chokes wildly, so you have to be very careful about what you feed him and how you feed it to him. Basically, he’s the most delicate baby ever. Which means Anna rarely gets an opportunity to nanny him. Nicole is always over her shoulder telling her what she can and cannot do.
It isn’t long before Anna senses something is off with Aiden. And it takes even less time for her to deduce that Nicole is bats—t crazy. There’s a great scene early in the script where, out of nowhere, Nicole goes into this seizure-like crazed fit that’s utterly unsettling. Just when it seems like she’s dead, she smiles and is fine again. She explains that she just wanted Anna to know how Aiden’s choking episodes looked like so she could recognize them. But that isn’t even her biggest “crazy” tell. She swims in her private pool with Beyonce’s “Single Ladies” blasting on max volume!
Anna senses that if she doesn’t get Aiden away from this witch, he will most likely die. But, you see, Anna doesn’t have the best judgement skills either. She’s got a couple of wackadoodle episodes in her own past linked back to a traumatic event in her life. So she wonders: Is this woman really endangering her baby or I am just going nuts?
I have to say, I LOOVVVEED the first half of this script.
It was everything I wanted my contained thriller on an island screenplay to be. It was very much a female version of Ex Machina. I’m sure that was a big inspiration for Kossan.
Nicole is a particularly fun character. She takes the “Oscar Isaac” role and ups it several notches. Her helicopter child-parenting and helicopter nanny-parenting is so fun to watch. You’re wondering what weirdo thing is she going to do next? And watching Anna slowly realize that she’s surrounded by cameras no matter what she does really makes the suffocation feel unbearable.
I would give the first half of this screenplay an A.
Here’s the problem though (and we’re going to get into semi-spoilers so you’ve been warned). There’s clearly something sketchy going on with Aiden, the baby, to the point where we know the writer is hiding something.
What do I mean by that?
If you are going out of your way to keep details away from the reader, we’ll eventually figure out that you’re cheating. We never see Aiden for more than a brief few seconds. He’s always under netting. Or when Anna takes care of him, we never really get a good look at Aiden. He always seems to be conveniently distorted. And because we don’t get any extended looks or interactions with him, we know something is up. So when the big twist comes, we already kinda figured something like that was the case.
Big twists are tough. I’m not going to pretend they aren’t. If you give too much information, we figure it out. If you give too little information, the twist feels random. Honestly, the only way to get twists right is from repeated feedback. You have to test the twist. Get feedback. Rewrite it with adjustments. Then test it on ANOTHER PERSON. Cause the last reader already knows the twist is coming so their feedback is worthless. Then you just keep making adjustments until it’s perfect.
This probably needs 4 or 5 more drafts to make this work. You need to give us more time with Aiden where Aiden is acting normal so we don’t suspect anything. And then you can’t highlight so many times where something is weird about Aiden. Because we can put two and two together. We know this is a tech CEO. She’s talking about making artificial wombs a lot. Artificial sperm. We suss out what’s going on here long before the twist arrives.
Which is too bad for this draft because the script started off sooooo good. It was easily in “impressive” territory. I just think there were too many questionable decisions in the second half. I didn’t think it was worth it to have Anna questioning her sanity. Every writer does this. And it’s never done well. Someone questioning their sanity always feels sloppy.
You had enough going for you with your original concept: A tech CEO invites a nanny onto her private island to take care of her baby then starts acting crazy enough that the nanny wants to rescue the baby. That’s a movie there. The “Am I also insane” question overcomplicates it.
[ ] What the hell did I just read?
[x] wasn’t for me
[ ] worth the read
[ ] impressive
[ ] genius
What I learned: When you choose to reveal information is so important to every story you write. It can completely change how the story reads. For example, Anna goes to meet Nicole on this beach where she assumes her house is. But instead, there’s a boat and it’s talking to her, telling her to get on. Kossan could’ve easily told Anna and us beforehand that Anna was going to an island. But then you don’t get the fun reveal of both her and us being surprised by this boat. Ironically, I think that Kossan makes some questionable decisions about when to reveal information later on in the story. But in this boat reveal, she hit the bullseye. That’s information that leads to a way better story moment if you wait to reveal it. It’s a wasted opportunity if you tell us about the island beforehand.
Today’s script is something Spielberg would’ve directed in 1988. And that’s a good thing!
Genre: Action/Adventure
Premise: When aspiring magician, Harry Houdini, discovers a mysterious puzzle-box, he must use his talent for illusion and escape to unlock the box’s powerful secrets and keep it out of the hands of a vengeful sorcerer.
About: This script finished with 12 votes on last year’s Black List. The writer is brand spanking new on the scene.
Writer: Matthew Tennant
Details: 116 pages
The Black Phone’s Mason Thames for Houdini?
This is a clever idea.
Studios have been desperate to find the next Harry Potter. It still baffles my brain that they’re going to turn that into a TV show with a new cast. Justice for Daniel Radcliffe. If they tried to remake Star Wars with different actors, I’d have a conniption fit.
So it’s a good idea to come up with another magic movie that focuses on a young character. You know the market is eager for it.
Also, Harry Houdini specs always get love. Even the bad ones. There’s something about that name that gets people excited to read a script.
Now, all we need, is for the execution to be great. Easy as a warm cup of cheesy tea-sy.
New York. 1889. 16 year old Erik “Harry” Weisz is a magician! Or, at least, that’s what he’d have the many small crowds he performs for around New York City think. In reality, Harry is an apprentice for a locksmith named Hugo Crane. Hugo has been on the wrong side of the law for a good portion of his life and, therefore, has taught Harry some trickery in relation to how to pickpocket.
Our opening scene has Harry performing a “magic trick,” which is really just a distraction so he can steal a rich guy’s pocket watch. Harry makes a run for it as the cops chase him which doubles as an establishing shot of 1880’s New York. We hit just about every fire escape and building top in the lower east side. I don’t know if this was actually happening on the lower east side. I just like saying “lower east side.” Makes me sound like a real New Yorker!
The cops end up catching Harry but Hugo, posing as a dastardly orphanage headmaster, convinces the cops to leave Harry with him, where he’ll be mercilessly worked to death over the next two years. “Can I have summor playse?”
Back at their locksmith shop, an old rich friend of Hugo’s stops by, Sir Neville Ballantine. Ballantine gives Hugo a puzzle box. Says he needs it opened and Hugo’s the only guy who can do it. The rumor is that there’s REAL MAGIC inside. No sooner has Ballantine left than Aleister Crowley, a 20 year old professional douchebag wearing a cape, shows up. Crowley wants that puzzle box!
When Hugo refuses to reveal where he’s placed the box, Aleister stabs him and leaves. Harry tries to save his dying friend, but Hugo says it’s too late. He tells Harry to get that box back to Ballantine. He’ll know what to do next. And with that, he breathes his dying breath. So off Harry goes!
Harry heads to Ballantine’s mansion where they’re having a ball for his 17 year old niece, Sophia. Harry is able to sequester Sophia away and tell her why he’s here. He needs her help to get this box to her uncle! Except no sooner does he tell Sophia this than Aleister appears again! What a douchebag.
Sophia, who doesn’t like Harry, is forced to team up with him and run off. They hop into one of the first ever “electric carriages” (a car) and it’s a car chase through the city. Beep beep! But after the chase ends with them in the Hudson River, the two will have to come up with a new plan to both open this box and defeat Aleister Crowley!
This was a solid script!
When I read these big adventure stories, I concede that we’re going to be following the Hero’s Journey. And even though that’s a familiar template, it’s a template that works. So I’m perfectly okay if your story feels familiar in that respect.
Where I push back is if your world and plot feel too familiar. Have I seen these set pieces before? Have I seen these situations before? Does every scene feel like a movie I’ve already watched?
I’m going to be honest here – I was feeling that a lot during this script. The opening set-piece where Harry steals the watch and runs through New York felt almost identical to a dozen scenes Spielberg has filmed.
In fact, I couldn’t stop thinking about the opening scene in Indiana Jones and the Last Crusade. It almost felt like a beat-for-beat remake of that scene, to the point where I could hear John Williams fun little “bompf” musical cues when Young Indiana would fall down.
But hold on, Carson. You just said you liked this. When does the liking start?
It’s the exact same lesson as the one I talked about yesterday with “Silo.” I loved the characters. I liked Harry immediately. He’s fun and funny and clever. But I obviously liked him even more when his mentor died. It’s a cheap move but boy does it work. We feel sympathy for anybody who’s lost someone close to them, especially if we’ve met that person and liked them ourselves, which was the case with Hugo.
What’s strange is, I usually don’t like the love interest characters in these movies, especially these days, because the industry pressures writers to make female characters perfect little Mary Sue badasses. And Sophia was a *little* annoying as she had some of those qualities.
But I can’t lie. I really wanted to see the two end up together. There was something intense about her disdain for this street urchin and I couldn’t help but wonder how he was going to puncture that wall and make her like him. If you liked the Peter Parker MJ dynamic in the Spider Man movies, I would put this above that. Tennant got a little more out of this relationship than the Spider Man people did theirs.
I only have one big complaint and that’s this new trend of side quests.
Boooo! Boooo! Hissss!
Side quests work like this. Let’s say your hero is pursuing something. The Lost City of Atlantis. And they’re running around the whole movie. They’re trying to find it. And they finally find someone who can take them down to the city. But the person says, “First, I need you to steal something for me. Then I’ll help you get to the city.” THAT’S a side quest.
You know how flashbacks stop the forward momentum of your story cold which is why they are evil and must all be placed at the bottom of a nuclear waste dump? Side quests aren’t that much different. They’re better than going backwards. But they’re still stopping the forward momentum of your story so you can go off on this little side thing that you never technically needed. Which is the real problem. A side quest is almost universally unneeded. It’s only there because the writer created it.
They just did this in John Wick 4 and it was so unnecessary. Everyone complains that that movie was 30 minutes too long. Well, you could’ve gotten rid of ten of those minutes in a heartbeat if you didn’t do that stupid side quest.
Same thing happens here. Harry and Sophia get the puzzle box to a woman who can open the box. But she says, first, you gotta go steal back this trinket that some bad guys took from me.
NOOOOOOOO. No side quests. Side quest bad. Side quest very bad.
I know why this is happening more and more these days. It’s because the first person video game generation is becoming adults. And they’re incorporating what they know from video games into screenplays. Which is stupid video game side quests.
But you have to remember that you can’t spell “movie” without “move.” A movie’s gotta move. A side quest THINKS it’s moving. It creates the ILLUSION of moving since your characters are going after something. But the main plot is stalled and therefore we feel stalled.
Outside of that one issue, I thought this was really good. I could see Spielberg circa 1988 attaching himself to this in a heartbeat. Unfortunately, in order to do that today, you’d need to set the movie in a newspaper room that’s covering the Cuban Missile Crisis. I kid, I kid! But actually I’m dead serious.
The end.
[ ] What the hell did I just read?
[ ] wasn’t for me
[x] worth the read
[ ] impressive
[ ] genius
What I learned: Houdini really is a spec-hack. If you can come up with even a semi-cool idea that revolves around Houdini, you’re going to get reads. Matthew Tennant came up with two key angles that made this feel fun. We meet Houdini as a teenager. That’s one. And we bring in the possibility of REAL MAGIC, which is two. Those two things made this concept pop.
Guardians of the Galaxy 3 is, probably, the best movie Marvel could’ve released right now, the reason being that IT’S ITS OWN THING. The problem Marvel’s been facing lately is that all of its movies have been intertwined with one another, and while that was great when the MCU was cooking, it’s become the world’s draggiest anchor ever since it entered its Marvels/She-Hulk/Dr.Strange era.
Still, I couldn’t muster the enthusiasm to get myself to go watch the end of Starlord’s trilogy. There were two main reasons why. One, the film looks sad! It looks like it’s going to lean heavily into its feels and that’s not why I go to see big Hollywood films. It’s why I watch smaller films and some TV shows. But when I go to see a big movie, I want to have fun. I want my spirits to be lifted, not dragged down to Sadsville.
And to be honest, I don’t think Guardians has earned the right to have a big emotional ending. This isn’t Iron Man after 20 films. This isn’t Luke Skywalker at the end of the greatest trilogy ever. It’s Chris Pratt, people in weird makeup, and CGI creatures acting goofy in space. Let’s be real here. People aren’t asking for Manchester by the Sea when they’re watching gun-wielding raccoons.
Then, of course, there was that second movie. That second movie was awwwwwful. It was weird. It was bumpy. It ditched its main character in favor of focusing on its two villains. Had that movie been good, I probably would’ve seen Volume 3 regardless. But the stink from that misfire still lingers in the back of my nostrils.
I’m so hot and cold when it comes to James Gunn. Never connected with his pre-Guardians content. Love Guardians Volume 1. Hated Suicide Squad. Loved the Peacemaker show. Hated Guardians Volume 2. I’m actually excited to see what he does with DC because he has an opportunity to dethrone the flailing Marvel. But this one? This Guardians 3 movie? I’m sitting this one out.
So, after my No Guardians For Me temper tantrum I just made you endure, did I give up on the weekend? Absolutely not. I did watch something. And that something ended up totally surprising me. It surprised me so much that I did research on the creator, Graham Yost, only to find out he was the writer of SPEED, one of the best action movies ever!
The show I watched was called Silo, which is a post-apocalyptic story about people who live in a giant underground silo city. Every once in a while, someone demands to go outside, convinced that the air isn’t poison and that humans can, once again, live on the surface. But all of these people make it a total of about 20 steps before they fall to the ground and die.
The pilot episode is about the silo Sheriff’s wife, who suspects that the governing body of the silo is lying to them, and that the outside is, indeed, livable. (**spoiler**) So she goes outside. And dies just like everyone else. The sheriff is left heartbroken. But as time goes on, he considers giving the outside a shot as well.
I actually read the book the show was based on (called “Wool,” which refers to the wool everyone has over their eyes) which started as a short story that the author, Hugh Howey, shared with an online group. The enthusiasm inspired Howey to turn the story into a novel (the first chapter in Wool, which was the original short story, is utterly riveting so I’m not surprised people fell hard for it). It’s very much like the “Lost” narrative where there are a lot of secrets and reveals, which makes it ideal for a TV show.
So then why isn’t anyone talking about it? They probably will as word gets out. But I think it has more to do with the fact that Apple has zero concept for how to promote a TV show. None of these streamers do, really, beginning with Netflix. But the thing about Netflix was it was a destination site. You went there looking to watch something then you saw the latest greatest Netflix show being promoted so you checked it out.
Apple TV not only doesn’t have enough material to be a destination site, it’s buried under too many layers of menus and buttons. Every time I fire it up, I feel like I’m turning on a nuclear reactor. Is this where the original shows page is? Or is it over here? I suppose Apple gets some credit for making me feel like a genius every time I find the show I came for. But the poor ease-of-use severely hurts its chances of anyone watching a show on its service.
Then, of course, it doesn’t spend on advertising. Which is bizarre for a company with a 1 trillion dollar market cap. Great shows are going to find an audience no matter whether they advertise or not. But everything else needs to build awareness. Apple literally has a media strategy whereby they don’t tell you about a new show and make it hard to find any show you do hear about. With that strategy, the ONLY way anybody’s going to be able to find your show is if it’s Game of Thrones level awesome. It baffles me that there are smart executives getting paid millions who don’t do anything about this.
Is Silo a great show? I don’t know yet. But the pilot is darn good.
I’m always surprised when something pulls me in. Since I know all the buttons and levers writers are pushing to get me to buy into their story, I’m hyper-aware that whatever I’m watching is being written. For that reason, it’s hard for me to get lost in a show/movie. But I got lost in Silo’s pilot.
What wizardry did the writer use to achieve this?
Well, they start by giving us an intense opening scene then flashing back. Yes, this is a cliche. But they do it well. They start us in the present, briefly setting up the world of the silo before the Sheriff tells the government he wants to go outside. The story makes it clear that this is a death wish, which is a nice way to intrigue the viewer.
That wasn’t what hooked me, though.
What hooked me was the story of the Sheriff and his wife once we flashed back. The entire episode takes place in the past and shows the Sheriff and his wife trying to conceive a baby.
I can’t emphasize this enough – when someone becomes hooked on your story, when they buy in – it’s almost always because of the characters. And it’s almost always because you’re being truthful with those characters. You’re showing us characters who are going through the same trials and tribulations that real people go through. It’s that authenticity that makes us care about them.
All the Sheriff and his wife care about is having a baby. That’s it. That’s all that matters to them. And each day that goes by, each day that they become a little less hopeful, pulls us closer to them, makes us feel more sympathy for them, makes us root a little harder that tomorrow will be the day they get pregnant. By the end of this pursuit, I was in love with these two. I was ready to run through a wall for them.
So when the wife says she wants to go outside, the equivalent of committing suicide, I was heartbroken. Cause I knew what that meant. I knew she wasn’t coming back and I knew that he would be wrecked.
It’s something I try to remind myself of all the time, whether I’m helping another writer or working on something myself. Don’t get distracted by the bells and whistles – the twists, the mysteries, the mythology, the double-crosses. If you make us care about your characters, we’ll follow you through any door. Hell, we’ll follow you right out the silo! Into that poisoned air.
Assuming we can find your show, of course.
If AI was a whisper two weeks ago, it’s become a Wilhelm scream over the last two days. That’s because with screenwriters striking, everyone wants to know if the studios are going to turn to AI to write their scripts. Let’s be honest. This is the dream of every studio in town. To be able to generate scripts without dealing with pesky neurotic inconsistent writers who never entirely deliver on what the studios wanted in the first place.
Can you imagine if you didn’t need to hire and fire Damon Lindelof for the next Star Wars movie? You could merely go to an AI chat bot, feed it an idea, and have the script written in 98 seconds? And pay nothing in the process. That sounds like a sweet deal to me. In fact, we can get a preview of what that script looks like right here. I’m going to prompt Chat GPT to write the opening scene of the recently announced Rey movie.
Riveting stuff there, Chat GPT.
I’m here to tell you that AI has no chance of replacing screenwriters. And I’ve got five ironclad reasons why.
It Doesn’t Know How To Be Funny
AI has no concept of humor. Which makes sense. AI is a logic-driven process. It crunches data and approximates how to use said data to give the user his desired result. It’s very nuts & bolts. It doesn’t do funny. And if you can’t do funny, you eliminate a gigantic component of screenwriting.
Even if you’re not writing a comedy, you need to inject humor in places. You’ll always have at least one character who’s comedic. The screenwriters who get the most attention are the ones who have strong voices, and “voice” is synonymous with how a writer uses humor.
I prompted Chat GPT to write a new Seinfeld scene. Jerry and the gang have found his apartment transported to the future, in 2023, and they’re trying to get back to 1992. For clarity’s sake, I prompted GPT to have them already in the future. Yet is still started like so…
And this is a sitcom that Chat GPT already has tons of data for. So it knows the characters and the kinds of episodes that are written and where the laughs usually come from. Even with all that data, that’s the best it could do. Now let’s see what happens when I tell it to write the same scene with a totally original sitcom, where it has to make up everything. Here’s what it gave me.
I don’t think there’s even an attempt at a punchline in this scene. It’s safe to say that comedy is the first area where screenwriters are safe.
How Can Something That Isn’t Human Document The Human Experience?
Every story that has ever been written is a reflection of the universal human experience. When someone writes about love, it reflects their own experiences with love. When someone writes about the devastation of losing a loved one, it’s because, either directly or indirectly, they have experienced that feeling of loss. Heck, when someone writes about losing their virginity, they are drawing upon their own experience of losing their virginity.
How can a computer ever cover these things in an authentic way? This is the secret sauce in screenwriting that you don’t really think about. You need to bare your soul in your writing if you want your stories to resonate with others. Neil Gaiman talks about this. That nobody paid his writing any attention until he started writing “the truth.” The truth being our true life experiences, the ugly stuff we’re ashamed about, the things we’re embarrassed of, the messy stuff that occupies the shadows and crevices of our existence. That’s what makes a story connect with other people.
I’ve been watching the Hulu show, Dave, lately. It follows a rapper trying to make it big. Dave spends a lot of time in his show focusing on his insecurities about his body, particularly in relation to sex. There’s a lot of fear and shame and confusion that occurs whenever he has sex. That’s an exclusively human thing that a computer could ever replicate. Why? Because a computer has never had to go through that before.
It’s a Litigation Nightmare
The dirty little secret about AI’s journey into art is that it steals. It says it doesn’t. It says everything it generates – pictures, paintings – all comes from its “imagination.” But a funny thing happened not long ago. Someone prompted the AI to generate a photo of a slide tackle in soccer. It came up with this.
Look a little closer. You see something there in the middle right side? It almost looks like a watermark. Here’s another REAL photo that happened to already be on the internet.
Looks similar huh? And what’s that there? Is that a Getty watermark? Ahhhh-haaaaa. Looks like we just caught AI in a big fat case of plagiarism. Obviously, AI is taking real photos and real paintings to create art. It’s stealing. And you can bet your bottom dollar that the second AI “writes” some scene into a movie that’s eerily similar to a scene from an unsold screenplay on the internet, or even a scene from another produced movie, that that studio is going to get sued up the wazoo.
This is going to happen with characters, scenes, concepts, dialogue. They could potentially get sued from two-dozen different directions if you believe that AI is crowd-sourcing data from every corner of the web to come up with their screenplays.
If you’re a studio, why put yourself in that position? Just buy a single writer’s script and you don’t have to worry about any of that.
Its Costs Are Going to Skyrocket
Recently, it was discovered that AI used Reddit as a huge language source for the way it thinks and acts. So Reddit said, we’re not just going to give you that data. We’re going to charge you for it. This is going to be a business that explodes over the next few years.
Every single company, Reddit, CNN, the New York Times, Twitter, Instagram, blogs — anything that has lots of data – is going to start charging AI to use it. Which means AI is going to have to pass those costs on to others. YOU. You’re going to have to start paying for AI. And for every corner of the internet that AI will have to pay money to use, those costs to the consumers are only going to rise.
What that means is that a producer is not just going to be able to get a Chat GPT screenplay for free. They’re going to have pay for it. And the costs are going to keep getting higher as people figure out how AI is getting its information and making sure that they have to pay to get that information.
It Cannot Offend
This is the whole ball of wax, as far as I’m concerned. It’s the smoking gun as far as why screenwriters will never have to worry about AI replacing them.
AI cannot offend. It’s not allowed to. It’s being reprogrammed EVERY SINGLE DAY to be less offensive than the day before. This is happening because the AI companies are terrified of AI saying something offensive. Because that means THEY’RE offensive. And they can’t let that happen.
The thing about writing is, if you want to write anything that has any sort of power to it, it’s hard to do so without offending someone.
AI could not have written Inglorious Basterds. German people could say it’s offensive and stereotypes them. AI could not write The Whale. It’s offensive towards overweight people. It would refuse to write Veep, as there’s way too much swearing and mean-spiritedness. It wouldn’t write Academy Award winning Promising Young Woman, as the subject matter of rape and revenge are tricky to navigate and may trigger someone. It couldn’t write 13 Reasons Why. The series might cause someone to commit suicide. It would refuse to adapt Anna Karenina, as there are no people of color in that story. It couldn’t write Beef because the show routinely makes Asian-Americans look dangerous and incompetent. It couldn’t write Super Mario because Italians don’t want to be stereotyped as cartoon plumbers.
American Beauty, Doubt, The Hunger Games, Jojo Rabbit, Get Out, Mad Men, Joker, Three Billboards, Spotlight, The Passion of the Christ, Room, Silver Linings Playbook, American Sniper, South Park, Deadpool, The Sopranos, Vicky Cristina Barcelona, Juno, Crash, White Lotus, Suicide Squad, both the British and American Office.
AI would not have touched a single one of those stories as they each have something in them that could potentially be perceived as offensive.
I’m not even convinced it could write Top Gun Maverick, as the bad guys are vaguely implied to be Russian, and the AI would refuse to stereotype an entire nation.
What this means is that AI would only be able to write the most vanilla of vanilla screenplays. The safest of the safe. You wouldn’t feel a single emotion because AI would go out of its way to avoid anything that might cause emotion!
People have said, “Yeah, but AI is only going to get better.” That’s true. But we’re also only going to get more sensitive. So it doesn’t matter how good it gets if it still has to avoid ten million offensive pitfalls every time it writes a screenplay.
The only entities who can write stories are people because people are still allowed to shine a light on the dark crevices of life. People are allowed to shine a light on the corruption inside the Catholic Church. People are allowed to write a pro-life movie like Juno and tackle the contentious underbelly of that subject matter. People are allowed to highlight the kinds of inadvertently offensive people who we find ourselves around every day, like Michael in The Office. AI is not allowed to do that. Which is why writers will always be safe.
Wrapping this up, do I think that AI could come up with a really clever plot like the movie, “Missing,” which I just watched? Do I think it could come up with a whopper of a twist, like “The Sixth Sense?” I honestly don’t know. I think it has a better chance of figuring out how to plot a movie since plotting is technical.
But plot is so closely interwoven with character. And “character” is the thing that I don’t think AI will ever figure out. Because it’s elusive. You can’t come up with a mathematical equation to write a great character. And, like I said earlier, AI doesn’t understand the human condition. It can’t. It’s never lived a human life. For that reason, every screenwriter in Hollywood can sleep soundly. You’re not going to be replaced.