This week has been a turbulent one in screenwriting. You could easily have missed it. Entwined within the drama was a historical turning point in the screenwriting world and what looks like the beginning of a new fight between creatives and capitalists. We have entered a new age of artificial intelligence that’s going to affect us in ways we never imagined how - CJ
Before we get going, it’s important you understand what the Black List is, and that’s not easy to get to grips with at first. The Black List is effectively a brand. It now encapsulates two core components;
If you’re confused then get to the back of the queue. Some argue that the conflation between the two is deliberate, especially given that the former is held in high regard and the latter less so. To me, it just seems clumsy.
Opinions on what the Black List is doing vary from acclaim to condemnation. The Annual Black List has been triumphed by many and associated with a whole host of successful projects, while some argue it is nothing more than a clique of friends mutually voting for each other’s projects, many of which are already well into production and widely known within the industry. The list fell into a little disrepute recently with the inclusion of “Untitled Lax Mandis Project”, a script written to troll Max Landis which is reportedly as poor as it is pathetic. The website is a more complex beast to judge. Services charging amateur screenwriters for exposure tend to gain support from those with survivor bias. Few are willing to accept luck has favoured them and often choose to believe the system that works for them is sound. On the flip side, it’s reasonable to assume that those critical of the system are simply butthurt they found no appreciation for their work and are looking for flaws in the thinking.
The biggest issue for me is the way one scoring system is used for the Annual List and another is used for Website. One works on positive up-voting and the other works on reductive averaging. I find that odd.
Either way, it’s always important to note that the way the Black List operates is often seen as contentious, but…
Those behind the Black List are obsessed with data, and proudly so. They search for patterns and logic within the numbers, the goal being to find a way to present them in a way that gives industry members what they need.
Don’t be led into believing that Hollywood is any different than other heavily industrialised sectors. It is corporatised. This is a world of bean counters and forecasters. There’s a lot at stake from enormous wealth to sizeable staff numbers. The marketplace is brutal and breeds the winner takes all mentality now needed to turn blockbusters into profits. These aren’t studios anymore, they are factories.
The Black List, and exposure services like it, are, in a way, a portal between the amateur screenwriting world and the machine room of the movie industry. We screenwriters get to peer through and we often don’t like what we see — we’ll come back to this.
The paid for Black List website evaluations, in a way that’s similar to how many screenwriting competitions work, use a grading system to rate scripts. Readers score a number of factors from premise to dialogue and add to this their overall likability score. As I mentioned, some competitions do the same, but are less transparent about it.
When I first learned about how these evaluations worked, I was sold on the idea. I was very new to writing and felt it was an excellent solution to meet the supply and demand problem Hollywood currently faces. With fresh Kool-Aid smeared around my lips, I exclaimed how brilliant I thought it all was to my partner, who, without even looking away from her computer, replied…
“You can’t do that with scripts.”
Okay, now a little bit of background on my partner before you jump to the conclusion she’s simply despondent within my presence. My partner (of 14 years no less) is a highly regarded library and research specialist. She would cringe at the highly regarded bit, but it’s true. With a strong academic background, she is in the business of cataloguing and researching data. When she says “you can’t do that” it means a great deal… to anyone sensible enough to stop and listen. Now, I’m just as stubborn as I am naive, so it took me something like three years to appreciate where she was coming from.
As long as subjectivity exists, we absolutely cannot quantify it. Within the sciences we can find the numbers and use them effectively. Outside of that, we may as well be investing in fortune telling. It can’t work, and any movie fan that’s ever looked up Metacritic and Rotten Tomato scores, or read through IMDB reviews will know this to be the case.
For me, I had to see the logic (or lack of) play out many times before I could really appreciate it. I’ve seen so many scripts torn apart and criticised only to go to get optioned and find an audience while competition winners get zero attention. People’s opinions on anything remotely akin to art are bullshit.
What the Black List tries to do is crazy, and the industry demand does not validate it. It only serves to show this madness creeps into some powerful corners of the movie making world. Yes, it’s possible to find patterns that look like cause and effect, but that pattern finding can also lead to people believing an octopus can predict the results of association football matches.
Three days ago, the Black List announced the release of a new service called ScriptBook. Priced at $100 a pop, this service was promoted as a way for screenwriters to extrapolate information from their scripts to help both them and industry members.
First off, ScriptBook from ScriptBook NV, is a pre-existing piece of artificial intelligence software which;-
“…indicates the commercial and critical success of a project, along with insights on the storyline, character analysis, target demographics, market positioning, distribution parameters [and more] prior to any made costs.”
The partnership being offered for that $100 would create reports like this one for Fences and, I believe, share some of that data with the Black List website listing so the script in question is better indexed.
Now, it’s important that you know that ScriptBook NV claims ScriptBook is already in use, but doesn’t disclose by whom exactly. The cost industry side is also kept private. The technology itself is mysterious, as AI often is, but I believe it can reference in the region of 5,000 scripts written in the past 40 years to help make comparisons. That’s often how these systems work.
That report for Fences will probably shock you. Software telling you how likable your characters are and what the budget should be. We live in times that are both as exciting as they are daunting. The machine age is truly upon us and I’ll discuss that right after we get into the drama.
If only there was AI that could predict how a service will be received by the marketplace, right? Irony alert.
Upon launch, the screenwriting world exploded negatively with the kind of force and spectacle that Michael Bay has wet dreams about. Which is unusual for a group of people usually more than happy to bend over and forgive the absence of any lube.
Reddit reacted negatively. Done Deal Pro reacted negatively. Twitter reacted negatively. Craig Mazin was temporarily unfolded from the foetal position and called upon to give his negative view. The resistance was so venomous that the Black List pulled the service within 48hrs and issued refunds to anyone who bought a report. Not too surprising when so many said it reinforced their belief the brand seems to have cynical motives behind its actions.
Some have applauded the response but it’s difficult to put the horse back in the stable now. The fact is, those behind the Black List felt it was reasonable to ask amateur screenwriters, those with the least amount to spend, the most to lose, and already facing a whole host of other costs to find exposure, a further $100 a time to use this service. Yet at the same time, they didn’t believe in it enough to give it a fair run, honour an offer of ten free trial reports, or let it prove its value. That’s really telling, because this is one of the more transparent offerings from this brand.
I for one welcome the robotic revolution. I think machine learning and artificial intelligence has a place where being human has its limitations and emotions can get in the way — such as medicine and law. But I’m not ignorant to the negative affect it’s going to have either. Millions are going to lose their jobs because they are being used as human robots.
The image we have is that it’s only blue collar workers who will suffer. That’s not going to be the case. We have made the roles of many white collar workers and creatives robotic too. Sadly, this has extended to the jobs of studio readers and low level executives who are tasked with rushing through an endless pile of screenplays and decide which will and won’t make money.
The Black List as inadvertently revealed that the future is potentially even worse than what we have now, the fleeting moments of a rushed reader’s emotional intellect replaced with light speed algorithms.
We should not be headed in this direction, but it may be too late to change course.
Figuratively dehumanising the reading process is one thing, literally doing it is another. The insult of having to pay the premium for it is really just the icing on the cake. We cannot allow this to become normalised.
Arguments will be made that the machines will free up staff to better spend their time, but industrialisation does not typically work like this. Computers were supposed to give us the four-day working week and here we are at the mercy of our office email 24/7. It is going to be too tempting for studios to sack readers and buy software subscriptions.
Corporate Hollywood wants to remove the humanity from the movie business because it’s a frustrating inconvenience. It will throw the baby out with the bathwater if it gets a chance, while completely forgetting that good stories are what makes us human in the first place. If we bring this down to ones and zeros then what do we really have left? Perhaps the ultimate dystopian future where life literally imitates art.
There is a place for these numbers people and it’s not within the arts. However, those who are good with numbers excel (pun intended) at making themselves and what they do look incredibly valuable. Plus they believe in what they are doing as passionately as we believe in what we’re doing.
The BBC tried this recently with The Secret Science of Pop which showed that being able to analyse is one thing, but being able to predict is entirely another. They couldn’t do it, they couldn’t reverse the process, and these were professors working with producers.
We got a glimpse through the portal here because a brand that’s managed to envelope the submission system for profit got a little tone deaf with its latest advancement. What we saw isn’t just ugly, it’s terrifying, and we can’t put our guard down because the monster’s hidden under the bed,
When it comes to the subjective, formula is for the foolish and statistics are for the stupid. It cannot work and we have every reason to stop paying for it and every reason to start fighting it now, before it’s too late, and a single line or single word we have written is the difference between the computer saying “pass” or “recommend”.
Comments
Fiona Faith Ross
Sat, 2017-04-22 19:02
Permalink
I read about this new script
I read about this new script-analysing software on a screenwriters' group on Facebook. As you say, the initial kneejerk reaction is one of distaste. This software could be useful in a first pass through a script, to check all technical elements are in place, e.g. Inciting Incident and other plot points (forgive me implying a write-by-numbers approach here. I don't like it either, but the hard truth is that plot points are important in script evaluation). However, I would raise certain pitfalls/issues arising from using software like this.
One. A film is a visual emotion-stirrer. How many times do you see the word "emotion" used in the filmmaking industry, from company names to reviews, to feedback, to audience reaction? This software is Artificial Intelligence. It can tick boxes but it cannot make value judgements. It is not Artificial Consciousness, which may be able to do such a thing in the future, if and when it comes into existence (see the work of Futurologists like Michio Kaku and Ray Kurzweil). In one sentence, how can an algorithmic software like this one judge the likely success of a film with a human audience? I would argue that it can't.
Two. Let's bring on the debate about "likeable" characters. (Jack Nicholson in As Good As It Gets. Would the software pan that script because Nicholson's character is a pain-in-the-a for the first part of the movie?). Who says they have to be "likeable"? Do you want to see a story unfold where the character learns nothing from their (whatever it is) life experience in the film, a story in which the "likeable" character stays in exactly the same place from start to finish? Okay, maybe your primary character does this, but at least have one secondary character who shifts their viewpoint somewhat (The Hangover). Not everybody buys a cinema ticket to experience Mary Poppins. Afterthought: I suspect they mean "relatable", implying a wider scope of character definition, which is important for screenwriters, and I also suspect, fairly crucial in getting the "likeable" box ticked on the score sheet in question.
Three. The "judgement" of this software, and the parameters set that underpin the number-crunching that pops out the "scores" for each element, are themselves, based on value judgements, that's right, real human emotions, so there will be bias in the scoring system, intentional or not. Imagine you have two software engineers (although usually it's a small team that would do this through an agile or SCRUM planning lifecycle), anyway, let's say these two have been tasked with setting the algorithm for "Likeable". As an example, I'll quote from Shrek since it's one of my favorite scripts. In Shrek 2, (spoiler alert), the ogre drinks a magic potion and he wakes up HOT, not as in, running a temperature, but hot hot (you know what I mean), so he acquires a "cute button nose", "tight round buttocks" etc. Imagine Engineer 1 is emotionally biased towards this template, but Engineer 2 is biased towards more of the "Captain Hook" kind of looks. How big a range of "Likeable" are we going to get? Either broad or narrow. On the narrow end of the spectrum, "button nose" scores 100, "hooked nose" scores 0, "white skin" scores 100, a "save-the-cat moment" 100, "blonde hair" 100, "brown hair" 75, "red hair" 25 - Anne of Green Gables, anyone? "Youth", say under-25, 100, "elderly" 0. Where would that leave Maggie Smith's brilliant character in Alan Bennett's acclaimed, The Lady In The Van? Or Mulan's Granny, who is adorable? I could go on.
So, there's a kind of paradox in this software, that leads me instinctively, CJ, to say, your partner is right. "You can't do that with scripts". You can do it for some elements of screenplays, but not for all of them. So I would say to you, forget the software and just do your thing anyway. If that closes off some avenues of sale for the bigger visionaries and creatives who come up with the disruptive stuff, who find magical and new ways to deliver "give us the same but different", then so be it.
Keep writing and don't get despondent if a robot doesn't have the heart to evaluate your work of art.
CJ Walley
Mon, 2017-04-24 09:50
Permalink
Yeah, I think, after a few
Yeah, I think, after a few years, we tend to grow a little tiered of the subjectivity argument, but it's always worth reminding ourselves that everything comes back to that. So many of us try to shoehorn logic into this art and, while a lot of it has academic merit, it's only ever a naive attempt to bring order to the chaos.
You can objectively categorise scripts. You can break them down to components. You can extrapolate data. But, like any form of art, you can only use that data to build libraries and not forecasts.
Gabe Spangler
Sun, 2017-04-23 05:27
Permalink
Great posting, CJ. I think
Great posting, CJ. I think society has jumped the gun on instituting "A.I." solutions. Why? Because A.I. doesn't exist yet. It has become a catch-all term for "smart" hardware and analytical software. Computers are not yet capable of true human reasoning and logic. They may be in the future, but they are certainly not there yet. But "A.I." has become a marketing term now, and people literally think it's a reality in our present-day lives. That is the alarming fact to me.
CJ Walley
Mon, 2017-04-24 09:57
Permalink
That's a good point, Gabe. I
That's a good point, Gabe. I worked closely to a robotics company through 2004-2008 and watched how those clumsy machines started recognising objects and manipulating them with pre-fixed co-ordinates. On the surface it looked like intelligence, but really it was a few algorithms benefiting from the latest sensors and motors.
It's very easy to push something to the market and claim its intelligent. It's also hard for us to determine if a plane's autopilot is boasts more intelligence than our phones face detection.
The ScriptBook report feels very similar to logging into Google Analytics to me. It's dubious data arranged into ambiguous graphs. Sure, you can find some trends and comparative data, but you can't live or die based on those numbers.
CJ Walley
Mon, 2017-04-24 14:42
Permalink
There was a quip on Reddit a
There was a quip on Reddit a month or two ago asking if we've reached peak Black List. Franklin laughed it off, but I think that Redditor may have had a fair point. Many seem to feel the veil has been lifted now and ScriptBook has confirmed their worries about the Black List brand having a very cynical method of operation. This looks to have been the straw that broke the camel's back.
Part of me feels bad for Franklin. I am ready to believe he's just an absolute data junky who's trying to do the wrong things in an industry that rewards him handsomely. On the other hand, I feel he's a businessman in a creative's world who knows how to turn desperation into huge profits. Either way, what he's bringing to screenwriting needs to be stopped before it becomes normalised.
Shawn Davis
Fri, 2017-04-28 14:39
Permalink
Fantastic blog, CJ!!!!
Fantastic blog, CJ!!!!
Well spoken.
Shawn.....><
CJ Walley
Fri, 2017-04-28 17:28
Permalink
Thank you very much for that,
Thank you very much for that, Shawn.
Steve Garry
Sun, 2017-04-30 22:08
Permalink
I'm a confirmed "computer guy
I'm a confirmed "computer guy", and if I were an executive I might run a script through the Scriptbook process just to see if it caught something I'd not considered, or to see how much it disagrees with what I concluded. But $100 for a single analysis is too much. It harkens back to the early days in computing, when software licenses were outrageously expensive. I recall the first word processing software, that had $1000 site licenses. That's "per computer"! Then the first WP dictionary software add-on: Wow! Cool! And another $1000. But, we still had to proofread our typing, because then as now you can't depend on the dictionary to catch everything. Thankfully, you know that many executives in the business don't place 100 percent trust in these algorithmic solutions. I'm a nobody of nobodies, and I still crack big companies all the time to obtain responses through my cold pitching. That would be an absolute zero level possibility if they were totally and completely committed to this high-tech malarkey. Look, even Microsoft has competition, so I don't think an Algorithmic Armageddon is coming any time soon. Being from the computer biz, I gotta admire my fellow computer geeks who continue to try to find some angle to infuse their enthusiasm into a new field of endeavor. But just when you think the feared Dystopia has arrived, or is inevitable, some script comes out of nowhere, to be discovered and exploited and... wait, it wasn't on anybody's list anywhere! That drives writers crazy with hope, but it also keeps executives up at night, as they worry about the competition finding it, something, whatever it was, before they do.
CJ Walley
Mon, 2017-05-01 10:30
Permalink
It's interesting you make the
It's interesting you make the software cost comparison, Steve, as I've been doing the same in my head. I've been doing it with service costs. For ~$50 a month, I can license the entire Adobe Creative Cloud Suite. Everything from Lightroom, to Photoshop, to Premier Pro. It allows me to do all my freelance work, be it laying out brochures, designing logos, or even putting together videos. The Black List want double that just to run a report through an algorithm. For $10 a month, I can license all of Microsoft Office, and get 1TB of cloud storage thrown in. The Black List want 150% more just to upload a script to their website.
I honestly cannot believe how expensive screenwriting can be compared to other professions. Sure, there is virtually zero barrier to entry, but everyone is pushed toward spending $50-$150 a month on something that is little more than a lottery ticket or spin of the roulette wheel.
There's no way any industry members will be paying $100 for a ScriptBook report in my opinion. Nobody on that side of the fence is that foolish. It turns out, to the surprise of the Black List, we're not so stupid on this side either.
Steve Garry
Mon, 2017-05-01 21:49
Permalink
I gathered that the
I gathered that the Scriptbook "offer" was for the screenwriters to ante up, but that the BL member agent/managers/producers would also get access to it. Nice. Well, I see where it's good for the writer to consider the business side of the business, and that such a report will help us see the buyer's point of view, but as we are all agreeing - it's just too expensive.
CJ Walley
Wed, 2017-05-03 14:39
Permalink
I don't think it's something
I don't think it's something they've marketed industry side. ScriptBook NV do that themselves I believe by offering prodcos and industry members their own subscriptions that would most likely be a monthly/annual cost than one off transactions since they'll be running a lot of submissions through the software.
Xavier Zinn
Thu, 2021-02-18 20:28
Permalink
As I read this, I’m thinking
As I read this, I’m thinking of the book “Appetite for Self-Destruction: The Spectacular Crash of the Record Industry in the Digital Age” - A great read for anyone who’s interested. Basically, the industry players aren’t looking at the industry and saying how can we work together to make it better - Instead, how can I win to the determent of everyone else.
Apple comes in pitches iTunes that each record company had independently created but discarded because nobody wanted to play nice in the sandbox and next thing you know songs are 99cents and they get something like 10% or less all because of ego, lack of vision, and stupidity.
Anyone remember BetaMax? it’s an interesting story that is also in the book, as are a bunch of others like the power of the record store in some decisions they made.
The point is, at least as far as I can see it, If players in an industry work in a vacuum, or think they are an island unto themselves, someone with a vision and the ability to execute will destroy them simply because nobody is looking at the bigger picture and making the industry a place where everyone can benefit.
As for AI specifically, it all comes down to money. I work in the mortgage industry and there is only 2 ways to make more money (excluding lying, cheating and stealing), fund more, ie lower your standards, or have more business come through the door. Back in the day we needed 40 million in the door in order to fund 15 million per day.
Unfortunately, underwriters can only go so fast, so if you don’t want to lower your standards, ie increase your risk, you need more business in the door, but then you need to hire more underwriters if you want to maintain SLAs. You want 100 million coming in the door; you need to double your staff. Because it’s all about EBIDA (Earnings Before Interest Depreciation and Amortization) nobody wants to spend the money if they don’t have to - enter fancy computer program designed to replace staff (probably over budget and late). Here, it’s auto-adjudication. Since everything in getting a mortgage can be reduced to a number, creating an algorithm is “easy” it spits out a number and based on your criteria it is either declined, funded or sent to a human for investigation, instead of 40 underwriters doing 100 million, you’ve got 5 on staff to look at “anomalies”.
AI is attempting to do the same thing, reduce it to a number and say this screenplay is good, this one is bad and then have your attention focused on the one that will make money - this way you don’t have to think too hard just look at the number and send it to its appropriate destination, or if you’re really savvy, automate it entirely.
If we assume for the moment this software did what it said and worked flawlessly, The Black List and its ilk would cease to exist because the company that created this awesome software wouldn’t license it to The Black List, they would license it to the studios and tell every screenwriter and wannabe screenwriter send us your screenplay and if it’s worth it, it gets sent to the studio powers directly. They’ve now created a situation where the AI company is now the gatekeeper of all things movie and they will gouge the studios, or take a percentage of revenues/profits and the studios will pay because it’s a sure win, and if they choose charge the screenwriter a fee as well.
They’ve just eliminated how many middle-men and ancillary markets?
And if we take it further, what purpose would an agent or manager serve in that specific scenario? Budgeting and such wouldn’t be needed either since this fancy software would probably have a buy-in module that analysed a million things and spat out a budget, not to mention what actors would be best suited to this role, and what specific profit could be expected based on X,Y,Z.
So, in thinking they’re smart, and this software is differentiator in the market, they’ve just signed their death warrant and created the possible the destruction of the movie industry as we know it.
A little far-fetched today, I admit, but the law of unintended consequences is real and always lurks in the background waiting to rear its ugly head just when you think you’ve done something great.
Balbeer Bahi
Tue, 2023-07-18 12:46
Permalink
Hi CJ, great article thanks.
Hi CJ, great article thanks. Just put two scripts through an AI analysis for the Black List. It was free based on a submissions deal on Film Freeway where one of my scripts had placed as a Finalist.
The script that had placed as a Finalist got a score of 35% possibility of being placed on the Black List, and the 2nd script, which has had three options and been read by over a hundred professionals and always well received, got a 5% possibility of being accepted on the Black List! To be fair, there are plenty of disclaimers stating that the analysis is based on simple word recognition, not on its ability to understand those words in context. (It counts nouns, adverbs etc)
So saying, it made me want to research it further which is how I ended up here, reading your article.
IMO it would be dreadful if producers start depending on this kind of nonsense to assess scripts.
So saying! I did once pay for an assessment on a reputable screenwriter's website where the analysis stated my script was so awful, I almost stopped writing! However, the same script, unchanged, also won awards and positioned as a Finalist in several competitions. Which thankfully renewed my faith in the script and my writing skills. So subjective human opinions can also vary and contradict one another. I'm not sure one AI program would say 'It's terrible', while another would say, 'I love it'! Cheers again.
It's all too weird for me!
CJ Walley
Tue, 2023-07-18 16:39
Permalink
Balbeer, given the nature of
Balbeer, given the nature of how scripts get onto the official Black List (votes from industry members), it seems insane to me that anyone could even think of trying to predict what's most likely to make it, via AI or otherwise. The list infamously featured a parody one year, written to mock Max Landis by a producer, and had no place being in any professional list at all.
But, of course, the tech entrepreneurs love anything AI right now, as it means being able to potentially charge for a product that requires next to no man-power, but feels potentially more sophisticated than any human can deliver.