|
Your new movie
A Very Long
Carriage Ride - in a few words, what's it about?
One film, two ways. It
is the first film in the history of cinema released in two different
complete styles simultaneously—stop motion animation and 2D classic
animation. This was only possible due to the cost-saving production
efficiency of AI. Each version only cost $1,000.
In fact, this you could say is the very first pure expression of the AI
artform specific only to AI itself, as in trad film, a director simply
wouldn’t do it because it would double the budget with no guarantee that
the audience will buy two tickets.
My prior AI feature films,
Window Seat—the first fully
AI feature film in history—used live action AI video,
DreadClub:
Vampire’s Verdict was the first fully AI animated feature film. My point is these
are both existing forms—live action and animation.
Their revolution was in the cost, production methodology, and the
spiritual value in bypassing institutional gatekeepers.
A Very Long
Carriage Ride however is
something entirely new.
Releasing two complete versions forces audiences to ask: what is the
film? Each version is only half the equation; together they form a new,
fluid definition of cinema. And this in a way sort of splits the cinematic
atom. When you look at that, suddenly the entire future opens up in front
of you. Meanwhile it continues to further the same production revolution
as the others. Since our last interview I have put all these AI filmmaking
methodologies into two books on AI film theory, The New Machine Cinema and the upcoming
Post-Scarcity Cinema. I will speak more on these later.
What were your sources of inspiratiion when writing
A Very Long
Carriage Ride?
Story-wise, it’s Charles Dickens, Jane Austen, and strangely during
production The Lord of the Rings, with
Autumn’s journey seeming like a trip to Mount Doom to cast away the One
Ring. It costs her everything, but her friends accompany her along the way
and soften the burden on her shoulders.
I also became obsessed with film history. George Méliès, Alice Guy-Blaché,
Émile Cohl, D.W. Griffith, Oscar Micheaux, Lotte Reiniger; names where you
see cinema is not synonymous with Hollywood, it began with independent
cinema and independent artists; the cast of characters you see today
existed nowhere in early cinema, instead it was exactly the sort of
raconteurs and ne’er-do-wells you see in AI cinema. This is the sole thing
for me which softens the blow of continued institutional resistance.
The film is also dedicated to Walt Disney and David Lynch, who died during
production and I like to think the dreamy interludes owe him gratitude.
Walt Disney for obvious reasons, he is third or fourth in all of cinema in
terms of structural, foundational importance.
But to mark the turning point of AI we have to follow it backward to how
both the digital streamlining and the CGI revolution led to AI, and what
came before.
You first have to go back to John Lasseter and the first CGI animated
feature film Toy Story. George Lucas and the
transformation of cinema into an effects-driven industrial pipeline.
Jean-Luc Godard, who tore apart narrative conventions and rebuilt the
language of film—and John Cassavetes and the birth of independent cinema
up to the 90s indie boom.
You’d even have to go back to the 1920s, with Eisenstein defining film
grammar, and the 1890s, with the Lumière Brothers’ experiments with the
camera—to track the structural evolution of cinema, where a singular form
of story evolves into the living, breathing cinema. Whether people like it or not, AI is the most monumental and seismic shift
in cinema in one hundred years—combining practically every production
revolution in one, and now we begin to push it even further.
 |
Artistically, during the production itself, I was inspired by Terence
Davies’ A Quiet Passion, Bresson’s Joan of Arc,
Sweeney Todd,
Black Swan. The anime series
Princess
Knight. Many of these don’t reflect the film itself, but the marathon
process of making it. So there was a lot here moving through me.
With
A Very Long
Carriage Ride being a computer-generated movie,
to what extent if at all did you use artificial intelligence in the
scripting stage?
This is an important question because there has been an enormous
misconception that my films are entirely AI written. I use AI as a
creative partner, I brainstorm, discuss, research and polish with AI, but
I am not handing it authorship. In
A Very Long
Carriage Ride, I
received these comments several times, that AI wrote the film, even its
ornate dialogue.
This is my dialogue, and you would go no further than my debut 2018 novel
The Unweddable Chattaway Girls, an 18th century period piece,
preceding AI by several years, which shows I write literary dialogue in
period settings with a hint of eccentric formal artifice.
To counter this idea that I don’t write my films, as with
DreadClub:
Vampire’s Verdict (2024), I am publishing the entire
production diary which includes my complete collaboration with LLMs in the
writing of
A Very Long
Carriage Ride. Meaning
people will be able to see exactly how I work with AI in the writing of my
films.
I am also planning to publish over 40 hours of direct screen recordings
making
A Very Long
Carriage Ride in real
time. I feel a responsibility to document, not only as a way of protecting
the record, but in preserving it. I am not against fully machine written
films, it’s not a priority for me personally, but I reserve the right to
experiment with that later.
You've made two versions of
A Very Long
Carriage Ride - so do talk about the two distinct stylistic
choices for the two versions, and why these choices, and why go with both?
Because this is my second 2D animated feature film after
DreadClub:
Vampire’s Verdict,
I had the process down pat and was able to make another one with more
evolved technology, and I was able to take bigger risks.
A Very Long
Carriage Ride really does capture the aesthetic of
classic golden age animation. Your review had remarked the story isn’t
always appropriate for a children’s film, but I also think there were
Disney films like The Hunchback of Notre Dame and Pocahontas and
The
Lion King that did have dark, emotional gravitas.
One film I have written called Strings is
going to share this animated aesthetic, but it is going to be the darkest
animated film ever made, as a commentary on how classic animated films in
the 1930-1950s were ignoring the social themes of their time. Strings will
be my most provocative vision ever, because it takes place right in the
South, post-slavery. However, I am also considering doing it live action
technicolor, inspired by Jean Renoir’s The River.
Ideally I could do it both ways again, but I do not believe I have another
dual aesthetic release in me at the current level of the technology.
The stop motion animated version of
A Very Long
Carriage Ride I
am extremely proud of, because this version is the most technologically
advanced AI feature film ever made up until this point. It has
authentically captured actual budgeted stop motion productions.
The exact film would cost some $15-$25 million in stop motion, likely
more. This goes even further to what you said, there is no doubt it is
going to attract kids who are curious about the film and end up seeing a
serious adult drama full of debates about God. Maybe they would get a kick
out of the characters and world, and the general energy of it, while
without understanding a single thing that’s happening. Regardless, it’s a
fun film anyone can enjoy.
Between the 2D- and the 3D-versiion of
A Very Long
Carriage Ride, what were the differences achieving them, and
which was easier to do, and why?
I made both the stop motion and 2D animated version simultaneously, often
in 10-20 minute increments. It was like climbing up a mountain, then
descending only to climb it again a second time at every post. Because 2D
works on a flat plane, it was easier to refine, and I genuinely had fun
making it.
The stop motion, because it was both my first stop motion animated film,
and because it had the responsibility of being the first AI stop motion
animated feature film in history, it was a nightmare. The detail and
nuance is so much harder to get right, but the benefit is that once it’s
right, it is so right. This is the
version I tell people they just have to watch, even if they end up
preferring the timeless feel of the 2D classic animation; the stop motion
is a must-view.
You used AI for
pretty much every aspect of
A Very Long
Carriage Ride, also for music, sounds and voicework, right? So
what were the challenges of getting the right score and dialogue delivery
to the screen?
AI music has advanced since
DreadClub:
Vampire’s Verdict, where
my job was to be a curator—to generate until I found the piece I liked. On
A Very Long
Carriage Ride you now had an ability to go in
and regenerate segments of a song, so you can assemble your music like a
Rubik’s Cube. The musical centerpiece of the film is the title song which
plays in the closing credits, and it is the first song in the OST.
No such job exists yet, but a fully AI composer would be helpful, because
a process that once took a few days can now take several months, opening
up brand new possibilities. But at the same time it also violates the
central tenet from the New Machine Cinema: One Person, One
Film.
All that said, please take us
through the whole step-by-step process of making
A Very Long
Carriage Ride?
I can take the opportunity to speak about The New Machine
Cinema and its foundational ideas.
One Person, One Film - the cinema of the future should
be made only by one single director with no other credits in the film.
Speed of the Mind - this is a maximalist credo: with
more production available to us for less, a film should have more content
crammed into less time.
The Living, Breathing Cinema - I sought to prove this
concept in building
A Very Long
Carriage Ride, it
states ‘a film is no longer a film, a film is a series of possibilities’
Machine Pleasures - in the production of
A Very Long
Carriage Ride, there were over 60 speaking AI machine actors. What
I found is that sometimes the bizarre machine readings came off wrong, but
were somehow preferable to regular readings. I call this Machine
Pleasures. It is when wrong is right. The reading sounds
odd, but hits your emotions deeper, and you wouldn’t want it any other
way.
The Automated Film - most of the anti-AI contingent
today act as if we are already at the point of automated films, it is
ultimately their biggest logical fallacy. I figured we must define the
furthest end point first to validate that we what we are doing is in no
sense automated.
The ideas in the book shaped
A Very Long
Carriage Ride’s production
to the point I was considering adding the full list to the beginning of
the film, currently they are in the front of the trailers.
The $64 question of course,
where can A Very
Long Carriage Ride be seen? My vision for
A Very Long Carriage Ride is that a
prompt begins with the film: Classic Animation or Stop
Motion Animation. They click. That’s the official way to
watch the movie. This has been fulfilled on the Blu-ray release.
Traditional streaming limitation means I can only submit one version, so I
am deciding whether to submit the stop motion version—the more advanced
version—or the 2D animated feature—the crowdpleaser. I decided to roll the
dice and submit both and hope they’re both picked up. Anything
you can tell us about audience and critical reception of
A Very Long
Carriage Ride?
Critics respect the movie and its dual-nature accomplishment, but they
don’t quite realize what is in front of them. Part of the divide is, what
happens when you have a literary, Jane Austen-esque film with a largely
female ensemble, aimed toward female audiences, but only men review it?
The ‘one person, one film’ tenet in AI is
not only philosophical but a budgetary necessity, requiring me to act as
PR. There, I have found that specialized audiences are strictly gatekept,
meaning you have to go broad, and then all your outreach lands on the
wrong viewership.
Film festivals, forget about it. Most of those are curated from direct
phone calls and handshakes with organizer presidents. Films submitted
blind are looked at under a microscope, while films curated from
relationships are given a pass sight unseen from name alone.
All this creates a paradoxical situation for A
Very Long Carriage Ride. What do you do when you are facing a wall of bad
faith resistance? With my previous AI features, I could bear that I was
tracking the limits of technology in the moment, but now, bad faith
opinions have turned into a genuine injustice; the limits are no longer
with the film, but with society around it.
A film like this becomes a stress test for the system itself. If
A Very Long
Carriage Ride isn’t given a platform, the entire thing is broken
beyond repair—therein it sort of exists as both the problem and the
solution at once.
Another way to say it, AI cinema currently exists as a kind of symbolic
protest vote. My job is to bring out the question, what if
it’s actually better?
For now, the historic markers are my sole driving force because it is the
only thing available to me. So I do it for the young me, who did not know
what would happen but knew something would happen, creating a whole
continuum of fate that wouldn’t exist without these works. Without them,
my entire past would be lost.
I do not create in a bubble. Every criticism, I go out my way to address
loudly on screen. Viewers will now see one minute long takes with complex
staging, full character consistency, a richer narrative with humor, drama,
pathos, it will now be in full color. They will see animation a leap
ahead, stop motion at a professional level, bigger and more refined music,
dozens of two-shots, four-shots, even a ten-shot.
These are all improvements in methodology, technology, and my growth as a
filmmaker. So to still be painted in the broad brush that because AI is
not perfect, nothing on screen has any merit—the very same response to
my early, primitive AI work in 2023—it’s disappointing. It becomes
apparent that I am growing while opponents have remained stagnant.
One critic said the film has too many characters. Yes, it has 60 speaking
parts! This is a feature, not a bug. In tradfilm, you would have a
nightmare of paperwork, unions, and payroll. Even at the professional
Hollywood level, we’re talking $100 million productions for a cast so
large.
The production cost and efficiency in AI now allows us to have more, and
with more for less, we must aim toward more. I use slow cinema as an
example of the reverse. Slow cinema is the natural haven for the resource
scarcity of budgeted independent film—the arthouse is full of slow cinema
because working in a form cast in the mold of low budgets is a necessity
(slow films are these long films with 5-10 minute takes, for instance,
unbroken shots of people cooking in their kitchen). I love many of those
films, but with AI, a film should be grand, majestic, and fast, otherwise
it is not taking advantage of the benefits of AI.
I term this simply as possible: post-scarcity cinema. Without
budgetary restraints, we are free to put down anything we can imagine, so
we must dream big.
AI cinema should be big—because it can. We don’t have
executives anymore telling us it has to play to the back of the audience.
Further, the modern era is what created the need for AI cinema. If the
prior model were preferable, AI cinema would not exist. So perhaps the
zeitgeist should be puzzled, confused,
angry, and at a loss for words, perhaps it means the medicine is working.
You have made AI-animated
movies for quite some years now, pretty much since the technology was new
- so how have things evolved over the years on a technical level? And
where do you see the technology go from here?
It has naturally gotten a lot better since
DreadClub:
Vampire’s Verdict. One thing I have always joked about
from the very start is that opponents mistake me for the CEO of AI. I tend
to be casted all the blame of the technological limitations as a flaw in
my ability—but then they turn around and give all the credit to the AI
when I really nail it. There’s no winning.
As for what’s next for me, it is currently animation. My job is just to
give every snapshot of the technology the very best possible feature film.
This led me toward animation out of necessity. I consider myself a live
action filmmaker, it’s just there was no means to make a live action AI
film since Window Seat
to that standard
in fluid, realistic performances.
But in the process of working in animation, I have become a huge fan of
the artform and its history in a way that I wasn’t before. I watch
cartoons and anime around the clock.
Currently, I am working on a CG animated movie called My
Boyfriend is a Superhero!?, which is a superhero rom-com, and I want to
get to Strings as well. So I will be
running tests soon to see if I can capture nuanced performances for a live
action Strings, or whether it will be an
animated film.
This is another thing about working in AI. Every single movie has a
workflow totally unique to itself. Every one is a record in time.
Your/your movie's website, social media, whatever else?
hooroojackson.com
instagram.com/hooroojackson
Anything else you're dying to mention and I have merely forgotten
to ask?
AI movie stars. Rogers, the machine actor who played both Cousin Martini
in A Very
Long Carriage Ride and Triswald in
DreadClub:
Vampire’s Verdict and
now Reejus in My Boyfriend is a Superhero!? I
consider him the first AI movie star. Every word out his mouth is gold,
and his line readings most encapsulate machine pleasures.
Joan, the machine actress who played both Betty Gray in
DreadClub:
Vampire’s Verdict, Katy
Bloom in A
Very Long Carriage Ride and now TV-Sami in My Boyfriend is a Superhero!? I
also consider her a machine movie star, with the sort of charisma you only
see on stage.
The last thing I would mention is that
A Very Long Carriage Ride was the most difficult production in my
career—two full features built side-by-side, an impossible timeline, and
even a brief stint editing the film from a hospital bed. But the film
proved that a single filmmaker can deliver stop-motion and hand-drawn
features on thousand dollar budgets, and the workflow breakthroughs are
already feeding my next two films.
 |
Feeling lucky?
Want to
search
any of my partnershops yourself
for more, better results?
(commissions earned) |
The links below
will take you
just there!!!
|
|
 |
It will be a lot easier to make films in multiple ways in the future, but
not like this. I wanted to offer the foundation by doing it from scratch,
that will one day be seen the way we see solving cryptography with pencil
and paper—by then, you will be able to change the entire aesthetic of a
film trivially, with one click. So I offer this film as proof, and the
biggest gift I could offer is that it’s not a tech demo but a fully
fledged feature film, to which I gave my all.
But art has a cost. I used to romanticize that my films were everything to
me and worth whatever cost. So if any young filmmaker looks at this,
because it is a notion of the young to romanticize sacrifice at the
expensive of everything, I will urgently warn them. Art comes at a cost we
may not realize. You don’t want to know your limits once they’re already
broken, and there are limits you don’t realize you have.
Despite this I am optimistic and I am still working. Stoicism is more
important than ever. I haven’t lost a step, but I am certainly
shell-shocked.
Thanks for the interview!
|
amazon.com
amazon.co.uk
amazon.de
eThaiCD.com
