Rob Bredow

Most artists, I think, our first reaction, my first reaction was like, well, that seems like a really bad idea.

I don't want a machine taking over my job.

And then when you start using it, you realize like, oh, okay, I'm very safe.

The machine can't do what I can do, but actually it does help me with this one part here.

And then you start finding little parts of the workflow that it helps you.

Then it becomes a tool like, you know, a better paintbrush.

At that point, we've achieved something that's actually can have a really nice benefit, I think.

How do you see, like, in the space that you're in, in museums, in curation, how do you see art that has AI used as one of the tools in it?

Do you see it making its way in?

Is it going to be 10 or 20 years down the line because it's going to be something that gets considered art much later in the scheme of things?

Yeah, that'll be really interesting.

And our uses of AI tend to be more about trying to get it to do what we need it to do, right?

Because our process is very artistic, but it's very commercial as well.

We're doing a lot of work.

We're doing a lot of shots in a row.

We're telling a cohesive story.

So we always need a lot of control.

And that's one of the reasons our artists, that's what our artists are excellent at.

It's getting what we're looking for, fine tuning the results to get that high quality of craft.

So it is interesting.

We're trying to force these tools that can tend to have a mind of their own, especially if you're just using the stuff that everyone has available on their social media and things.

Trying to turn that into something that is more controllable is a whole operation unto itself.

So when you're curating a new exhibit or thinking about the future of your museum in San Francisco, how do you consider all the different disciplines that can go into the craft?

Is that an important factor for you?

That's so interesting.

That's actually kind of similar in filmmaking too.

It tends to be a filmmaker or sometimes a team if it's a whole series.

They have a very particular vision and you always know where to go when you have a question about what you're making because they're the ultimate decision makers.

But all the different crafts coming together, it's a collaborative process.

On a common day on the set, we'll always have more than 100 people.

Sometimes if it's a complicated shooting day, there's 250 people working in their specialties.

And it's down to lighting and the set dressing and the rest of the rigging and paint touch-ups and all these things happening simultaneously.

It's one of the things that makes filmmaking so amazing is that all that can happen in a compressed period of time on the set just to get that moment of acting in front of the camera.

It's what makes movies kind of special, I think.

I think it's so important as we start using more and more AI tools to...

have them driven by the people who are the artistic experts in their relative crafts.

I think one of the tricky things about the existing commercially available tools right now is they tend to be designed for someone who's not going to invest very much time and wants to get a lot of output.

So a little bit of input, tons of output, and it's impressive.

It's technically impressive that it can happen, but it doesn't allow for someone who's a real expert in the craft to refine that to get them something that is, well, that is maybe...

a cut above.

I think we'll be better off when the tools are being designed for these artists that have the experience.

That'll be interesting to see because we have a bit of a hump to get over, I think, there with the AI tools.

Yeah, that's been our experience so far.

I think in concept, most people, most artists I think, our first reaction, my first reaction was like, well, that seems like a really bad idea.

I don't want a machine taking over my job.

Then when you start using it, you realize like, oh, okay, I'm very safe.

The machine can't do what I can do, but actually it does help me with this one part here.

Then you start finding little parts of the workflow that it helps you.

And then it becomes a tool like, you know, a better paintbrush.

And then I think at that point we've achieved something that's actually going to have a really nice benefit, I think.

That's such a great example to really look back.

Every time there's been a technical innovation, there's people wanting to hold it back for good reason, and then there's people who want to lean into it.

And then, you know, the example we often refer to is photography versus painting, which, of course, when photography came around, there were a lot of painters saying it took the painters' jobs.

But then the painters became more expressionistic and did a whole new thing with that craft that, of course, there's still photorealistic painters today, but many more of them.

are able to express new kinds of things because the photos are doing the job of just documenting what happened.

Well, there's so many different ways, actually.

A lot of the things we invent at Industrial Light & Magic and inside of Lucasfilm are as a result of a person or a team getting curious about whether something will work.

And our hunches in that area, I mean, of course, they're informed hunches because this is coming from experts.

But the areas to explore are almost always as a result of someone saying, I think there's either a better way or this person needs to do this impossible thing.

And we've got to come up with some way of doing it.

So having that intense curiosity to want to chase those things down, because it's always inconvenient, right?

There's never enough time.

There's always risk that it's not going to work.

But that's one of the things that I think people in general who are attracted to this company to work,

they come with a lot of curiosity and a little bit of dissatisfaction of the status quo.

You're like, status quo is not quite good enough.

We can do this better.

Yes.

I love that.

So I'm so interested as, I mean, that's a innovative move to do at the Smithsonian.

How did you, do you run into a lot of resistance or do people who are pushing the state of the art in museums, do, is there a lot of resistance to the status quo or is there a lot of like invitation?

Yeah, absolutely.

Especially as we look at new tools, we look at AI, making sure first that we position it the right way, that it's not about replacing artists.

It's about empowering artists and making our craft as efficient as possible, giving everybody the right tools.

I think that's when if everybody gets on board with that, then we're headed in a useful direction, which I think is really interesting.

I love that question.

That's exactly what we do.

And sometimes we look back at some of those original shots in the original Star Wars and the craft that went into them is

Well, first, things like shot composition and those kind of things, those are timeless.

So there's lessons to be learned from the people who invented visual effects, some of whom we still get to work with, so we can ask them questions about how they approached it.

But also, like when we started Star Wars Episode VII, so J.J.

Abrams' film, this was about 12 years ago,

we went back to all those miniature shots, because we were going to do most of it with CG, almost all of it with computer graphics.

We went back to those original shots and said, like, what about that AT-AT walker is so charming in this original?

And what are the things that we want to preserve from that era or emulate with the new tools?

And what are the things that actually are just artifacts of the old way of doing it and actually they should just be improved with the new digital tools?

So a great example of that is the ships that when they were miniatures, the lights could only be so many feet away, like the studio we're in now.

They can only be 10 or 15 feet away because at some point you run into the wall.

You're trying to simulate a sun, but you can only do so well with that.

And it turns out like flying a ship that has mostly flat surfaces with a light that's relatively close and can only be so small.

If you're simulating a sun, it's like infinitely far away.

The computer can simulate it perfectly, but it's a little boring because it's just like an even illumination of light.

But if you have a light that's relatively close as the ship flies around, you get more interesting shading.

And we wanted the hard shadows so you believe it was the sun, but we liked the more interesting shading.

So we were able to keep the elements from the stage photography that we liked and drop the elements that made it feel a little bit artificial.

So those are the kind of details we kind of obsess over.

Well, I love that question.

I think it, especially for storytelling, the constraints make the story, I think we relate to it well because we understand the world.

We understand the constraints of the world and you set up the rules in the first five minutes of a movie and you want to adhere to those.

So that goes for, I think, the visual language of the storytelling, how the camera moves.

I feel like it's upon us as creators to actually set the rule set for our show so the audience can then relax and go, somebody's thought this through.

We know the tone, we know the feeling, and you'll hear people, I hear people a lot talk about tone of their favorite shows.

And myself, when I see something that is well-crafted, that has a consistency of style and design through it, I know I'm in good hands so I can relax and watch the movie.

I imagine that's somewhat similar when you're talking about an exhibit, you're talking about a museum, laying that out and feeling that design.

Yes.

Cool.

So through the course of my work, I've become familiar with Rafiq Anadol's work, and I'm just so excited to see the kind of attention it's getting in museums, where people sit there for half an hour and watch an exhibit because it's very mesmerizing.

It's kind of almost trance-inducing.

Is this the kind of thing that you would like to see in general in the museum community?

Is this the kind of exhibit that you would like to see continue to be explored and pushed forward, or is it just like the right dose of it is what we need right now?

Yeah.

The variety of different kinds of expressions of that might be interesting to see next.

Yeah, and this stuff is so ephemeral.

Like, it requires a certain computer with a certain GPU.

Like, these are going to be harder pieces of art to keep and preserve than a painting on a canvas.

I love that.

This is the pitch I heard when George Lucas talked about opening his museum in L.A., which is still under development, and I don't have any inside information about it.

But what I heard early on was that it was going to be classic painting next to a movie poster that was also a fantastic painting.

And I don't know if that's how they're going to end up curating it, but if they do, that sounds delightful to me.

Yes.

Yeah, they really do complement each other.

When we have a shot in The Mandalorian that I talked about in my TED Talk, where we have stop-motion animation combined with real-time computer graphics combined with a real set, and it's seamlessly blended between all three of those live on set where 100 people can compose their shots and make a show inside this both high-tech and low-tech system, which I think that kind of combination is really nice.

That is 100% the right word for it.

It's the magic trick when you could never guess how it would be done.

Even people who've seen the behind the scenes are like, okay, I know they used LED walls.

I know they did digital extensions.

They would have never guessed there was also some stop motion in there.

And as soon as you start mixing the media, it informs what it looks like.

Oh, I mean, the story is king, 100%.

The story has to be the driver.

And the characters, you know, when I first got into visual effects, especially when I first got to the point where I could help design sequences, because we do a lot of, let's say there's something like the Kessel Run or something like that in Solo.

The visual effects supervisor has a lot to do with the design of that sequence, trying to make sure we understand how we're going to approach all those shots.

It would be too easy to lose track of the people when you're just designing spectacular ship shots.

And even knowing that, it's still easy to design in too many ship shots and not enough shots where the action is going to take place on the actors' faces.

As soon as you shoot the whole thing, you realize, oh,

Of course, every time I always get reminded the story is on the actors' faces.

It's not even just their lines.

It's their lines, but it's more their reactions to other people's lines, to what's happening in the world.

And you do need a few shots to keep yourself organized in the geography and to understand the story.

But most of it's on actors' faces, which is lovely because that's what we love to look at.

Oh, I think the opportunity to continue to give that moment on set the most information as possible.

So some of the reasons the LED walls are fun is because the actors get so much more information than acting in front of a blue screen.

There's a temptation to build less and less and to lean into the digital techniques more for cost reasons, for efficiency, to move fast, all sorts of good reasons.

But being able to make sure that set stays a magical space.

When we loaded the cast of Solo into the Millennium Falcon and there was a wraparound projection screen and we didn't tell them in advance that,

It could do everything.

It could do the whole Kessel Run.

We just had stars up there.

So as far as they could tell, it was a painting.

So they got in there.

And this was Ron Howard's idea, not to tell them, just let them get up there and experience.

He's like, well, why don't we do a dry run?

They're just still holding their script.

They're half ready for the shoot.

So they sit down in their spots.

And this is Donald Glover, Phoebe Waller-Bridge, Alden.

And I can't remember who else was in that particular shot.

But their first time in the cockpit, Alden was playing solo.

So Ron does the countdown, and he goes, OK, ready, Rob, cue hyperspace.

And so we cue hyperspace, and the special effects guys shake the Millennium Falcon.

And they're just all freaking out in there, because they pulled the lever, and they went into hyperspace.

They didn't expect this to happen.

And everybody was yelling and having fun.

And then I heard Donald Glover say, after the take was all over, he leaned back, because I was listening on the headphones.

And he goes.

this is the coolest thing I have ever done, which was just, I mean, to do that and have that be their first rehearsal, it changes what they're going to do on screen.

Absolutely.

Giant space monster appears on the screen.

Everybody knows right where to look, exactly when to react.

You get a different performance than if you're just saying, look at that tennis ball and ready, one, two, three.

And actors are great.

They can do it.

But how much better is it when you give them the real thing?

When you're thinking about how a piece is going to resonate emotionally with people walking through the museum or what story you're telling and what you want people to experience emotionally as they've gone through a whole part of an exhibit, what are the ways you try to use the art to kind of create certain emotional feelings?

Yeah, that is very similar in filmmaking where we're

Our best heads of departments, the teams that run all the different departments that work on the show, they spend so much time with the script, understanding the story, understanding the theme, and then using as many tools as possible to reinforce those themes, those emotions.

And sometimes that can be technique, you know, the choices that are made in the craft.

Sometimes that's just a feeling that everybody's going for.

Like some of my favorite movies have these wonderful color scripts that even before we know what all the designs are, we know what the mood of the various scenes are going to be because of the colors that we're going to establish way early.

So then you get back into that and you're like, okay, how are we going to make that magenta feel in here as you're adding in each of the designs?

So it gives the cohesive flow over the course of a movie.

Completely.

Dennis Mirren, who is a visual effects supervisor who worked on the original Star Wars, he said, on Star Wars Episode IV, that was our first big movie, right?

They'd done all the shots.

They could only see the mistakes that were left in the shots, all the things they wanted to make better.

And then he went to the premiere down in Los Angeles in Hollywood.

And he watched the movie with the audience and all he was seeing was those mistakes, but he could see everybody else was seeing the storytelling.

And that's what he said, that's when he became a filmmaker instead of a shot maker, right?

Someone who could see the whole story instead of just all the different details, which I think is kind of beautiful.

Absolutely.

Perfection isn't necessary.

And Dennis would always be teaching us this.

Dennis Murin and Ken Ralston and the teams that worked on those original films that inspired so many of us in our craft, they're really good at saying, oh, that detail over there, it's imperfect, but it doesn't matter.

People aren't looking there.

I'd rather spend the time working on the thing that people are looking at.

So there's lots and lots of great examples of films.

Yeah, exactly.

Perfection is not always better.

It has really been fun.

I love learning a little bit about your world.

So thanks for taking the time.

The opening shot of Star Wars, episode four, the original one, it broke all new grounds and really redefined that era of filmmaking.

Plus, it's just like perfectly designed.

Now, behind the scenes, you may or may not know the camera rig.

This is the Dijkstra Flex.

It's a computerized motion-control camera rig.

It was innovative.

It could be moved from stage to stage.

It enabled these new dynamic camera moves on models, on miniatures, for the very first time.

Again, redefining that era of filmmaking.

Today, we're entering a new era of technology.

But instead of being introduced to it with a spectacular opening shot of a movie, we're reading about it in the press.

And as someone who's been a visual effects artist for over 35 years, some of these headlines are pretty disappointing.

They seem to say, AI is coming, and it's coming for our jobs.

And the tools, they're getting pretty good, too.

You know, I work on movies, I should be safe for a little while, right?

It's going to take them, well, you know where this is going, Google, Meta, OpenAI, all showing sophisticated video generation systems.

These appear to be doing some of the same things we are doing right now, frame by frame, with multiple artists often working weeks at a time to make a single three- to five-second shot in movies you love.

It's going to be pretty intimidating.

And we're not the only ones who are a little concerned about the future of AI.

OK, we've been here before.

And I believe this is an interesting collision, right, of tech and art.

And I believe what George Lucas created in Industrial Light & Magic, his visual effects company, ILM, he founded this 50 years ago to solve the visual storytelling challenges in his films.

And what he did there is he put artists side by side with engineers.

They were working together to be able to innovate and create the future.

This is artist-driven innovation.

I'm passionate about this.

So what we're going to do today is we're going to look at 50 years of history at ILM.

We're going to pull some of those cinematic game changers that have happened over this period of time, and then we're going to try to draw some universal themes out of that.

And in the end, I'm going to premiere for you, for the very first time, a short created by an artist using the latest generative AI tools, so we can see the trajectory we are on and the trajectory we want to be on next.

Let's turn the clock back to the Jurassic era.

This is the Jurassic Park era, my Jurassic era, 1993.

On this film, they were going to use as many full-scale dinos on set as possible.

Just an amazing, amazing movie.

I mean, this film really ended up redefining the way visual effects were done in movies.

But it wasn't planned that way from the beginning.

They were going to do these huge, full-scale dinosaurs, which they did, and they were amazing.

But for the shots where they couldn't have a full-scale dinosaur run like a T-Rex, they were going to do stop-motion animation.

And all that changed one day when Kathy Kennedy, the show's producer, she was walking the halls of ILM, and she noticed this test looping on a monitor.

Now, the animators had actually positioned this monitor, hoping she would see it.

This is the early days of CG, computer graphics, and this looks promising, enough to commission a test with skin and lighting.

So they did that test just real quick, and Dennis Mirren, the film's visual effects supervisor, and famous stop-motion animator Phil Tippett, they took this down to Spielberg and to Kathy Kennedy, and they looped it a few times in their screening room.

And as the lights came up, they knew cinema history had changed.

It wasn't going to be the same.

Spielberg caught Tippett's eye, and he could see his expression on his face.

He said, how are you feeling, Phil?

Phil said, I feel like I'm going extinct.

But I think artists right now, many of us are feeling that same threat of extinction.

But there's good news here.

This is not what happened on Jurassic Park.

Innovation thrives when the old and new technologies are blended together.

Jurassic broke all new ground in computer graphics.

All new techniques for the first time had been used at this scale.

But many of the techniques were exactly the same as before.

The animators taped themselves on video for reference.

They drew these amazing storyboards of these shots that are now indelibly imprinted in all of our minds.

Engineers at ILM even went so far as to make the dinosaur input device.

It's a stop-motion armature.

Each of the joints, it has a little encoder on it so it can send the data back to the computer.

So the stop-motion animator can do frame-by-frame animation like they normally would.

The computer would record it.

That animation could be sent across the bay from Phil Tippett's shop to the artist at ILM, who could put the skin on and light it and render it, even touch up the animation digitally to create the final shots in the film.

literally a combination of the oldest animation technique we know, stop motion, and the brand-new computer graphics techniques back in 1993.

This has been true recently with AI technology as well.

Harrison Ford from Indiana Jones and the Dial of Destiny, performing on set in that iconic role he created, but he's looking a little younger for the opening of this film.

Our artists hand-tuned generative AI models on past performances of Indiana Jones, of course with Harrison Ford's permission.

In addition to that, our artists also created a full CG computer graphics version of his performance.

So why did we do it two different ways?

Well, it turns out the AI system is really, really good at getting that likeness, pulling all the details out of Harrison's performance that he could do unencumbered on set, get that into a likeness and the correct age that we were looking for for this character.

It wasn't as good at the details and the controls.

That's when our CG model came to the rescue, and our artists created a blend of the two to create the opening 20 minutes of this movie, and I'd say some of the best work of this kind ever done.

And it's not bad if Harrison Ford says, it's pretty good, it feels like me.

So that's blending the old and new, right?

Leaning on the experienced artist to drive even the brand-new techniques.

That takes us to how technology can actually work hand in hand with creative.

I believe that we're designed to be creative beings.

It's one of the most important things about us.

That's one of the reasons we appreciate and we just love it when we see technology and creativity working together.

We see this on the motion control on the original Star Wars or on Jurassic Park with the CG dinosaurs for the first time.

I think we just love it when we see creativity in action like this, tech and creative working together.

If we fast-forward to 2020, we can see the latest real-time virtual production techniques.

This was another creative innovation driven by a filmmaker.

In this case, it's Jon Favreau, and he had a vision for a giant Disney Plus Star Wars series, first time we were ever going to do a series of Star Wars on TV.

This is for The Mandalorian.

And it was a big show, but we didn't have a normal film schedule.

We had to shoot it all very, very fast.

So rather than flying the cast around the world to try to find these exotic locations, we brought locations to them on a single soundstage in Manhattan Beach, where we had configured this LED wall to act as a virtual production background, so we could immerse our actors, our entire crew, all of the departments in these virtual worlds, and even some locations that weren't from our world, they're from outside of our world.

Interestingly, Kathy Kennedy, the same producer who produced Jurassic Park, she now runs Lucasfilm, she was one of the people who got behind this idea early, before any of us knew that it was going to work.

The technology leading up to this had been created hand-in-hand with filmmakers over the years.

So I was supervising Solo, a Star Wars story, and we thought, could we take what they'd done before and could we make the content high enough quality so you could point the camera directly at it?

And we started making shots like this for the very first time.

This is hyperspace in the cockpit of the Millennium Falcon, and you can see my friend, the DP, Bradford Young, and he invented this shot.

We'd shot hyperspace tons of times before.

But in the moment, he saw the light coming in and he realized this was a moment of hope in the film.

So as he pans over to Han Solo, you can see the reflection of hyperspace in Han's eyes.

Just a beautiful collaboration between tech and art that elevates the storytelling and gives our artists new tools.

That's what we love.

So that catches us up to The Mandalorian, where, as always, it starts with words on a page, the script.

And then our artists break it down.

And I'll break it down for you, how they add all the different elements.

Computer graphics shift, add it in post-production.

But the rest of the background, that was live on the day, on that LED wall, for all the departments to work together to create life in that scene and block these shots out.

The walkers in the background, you'll never guess, they were stop-motion animated over at Tippett's shop.

Yes, the same Phil Tippett, far from extinct, all these years.

So it's a combination of the oldest technique in the book with the latest real-time computer graphics to bring an illusion together on the screen in front of the whole cast and crew.

Our artists worked really hard to make it a seamless blend between the two.

And now we can pop these LED stages up anywhere in the world when it's the right creative fit for a filmmaker.

OK, so that's blending the old and new, how tech and creativity working hand in hand create things we just love.

And last, maybe most importantly, our theme about artist-driven innovation.

So what happens when you put the latest AI tools in the hands of talented artists, both to see how good these tools are these days and what does it do to our artists' imagination?

Where does it take their imagination?

So I'm going to premiere for you a brand-new short film.

This is just a test.

This is not a final product.

This was created by one artist.

His name is Landis Fields.

He created this in two weeks.

to explore what it would feel like if you sent a probe droid out to a brand-new Star Wars planet.

What would it see?

And more importantly, would this kind of show, if we made this, and I'm not announcing a new show, this is just a test,

If we made this, what would it feel like?

Pretty fun to see artists, expression, leveraging the latest new tools.

Now, this, as I mentioned, is just an early work in progress.

For a show, for a movie, we need a team.

For so many reasons, we at Lucasfilm on Island, we don't think we're just around the corner from one or two people making a film in a darkroom by themselves.

We benefit so much from the diverse skill set of the artists and the technicians that work on our films.

They bring so much creativity to our projects.

That's what we want to capitalize on.

What you see here, you can kind of think of it as like a mood board, but a moving one.

And how much more powerful is it to have a tool like that to get an early impression of a show as you're deciding what to do next?

It is natural for us to think about jobs as these tools continue to evolve and get more powerful.

If you go back to the days of models and miniatures and motion control, there was only a couple dozen people in the whole world who know how to build and operate these complicated machines.

Today, there's literally hundreds of thousands of artists around the world who know the latest CG techniques, the latest computer graphics, visual effects techniques.

There's a huge opportunity for creative, talented people with these skills.

And as our tools continue to evolve, and we take our responsibility seriously to grow and develop the talent that's working on our projects, this is how we've been doing it for the last 50 years.

In fact, some of the artists who worked on the original Star Wars series are still working with us at ILM today.

So that's artist-driven innovation.

Like you've seen on our productions over the decades,

Now, the tech companies on their own, they don't have the whole picture, right?

They're looking at a lot of different opportunities.

We're thinking about it from a filmmaking perspective.

And storytellers, we need better artist-focused tools.

Text prompts alone, they're not great ways to make a movie.

And it gets us excited to think about that future where we are going to be able to give artists these kinds of tools.

And as these tools do get so powerful, we do need to use them thoughtfully, with the full permission of the talent, like you saw in the example of Indiana Jones.

where Harrison Ford gave his permission for his likeness to be used, and we had full access and the rights to the training data.

So that's what we're focused on at Lucasfilm and ILM.

the next generation of creative tools leaning on all these amazing innovations.

And then we can just imagine that next game changer, that next Star Destroyer moment that's going to light up screens around the world.

It's going to be even more imaginative, can be even more creative than that original that inspired us all in 1977.

Thank you very much.