Luma Labs Dream Machine: Bringing AI Characters to Life (and Possibly Hollywood)

Luma Labs' new text-to-video model is here, and it's bringing AI characters to life! See how our virtual friend Kateřina went from static image to animated camper.

Luma Labs Dream Machine: Bringing AI Characters to Life (and Possibly Hollywood)

Is your AI-generated character feeling a bit stiff? Luma Labs' new text-to-video model might just be the caffeine shot they need!

By now, if you're into AI Art (and let's face it, who isn't these days?), you've probably heard whispers about text-to-video generators. Like me, you're probably refreshing your browser every five minutes, waiting for the next big breakthrough. Well, grab your popcorn, folks, because Luma Labs just dropped their new model in beta, and it's more impressive than my attempts at recreating movie scenes in my living room.

Will Denis Villeneuve be using it for Dune 3? Probably not. But hey, that doesn't mean we can't take a peek at where the future is headed. After all, today's "neat party trick" is tomorrow's "how did we ever live without this?"

For those of you who've been following this blog (hi, Mom!), you might remember Kateřina, my AI-generated muse who's been lurking around in my images like a friendly ghost. If you're new here, no worries! Check out the links below to catch up on my "Crafting Cohesive Characters" series. It's like a soap opera, but with more pixels and less evil twins... for now.

Kateřina Goes Hollywood

I decided to take one of the images from Kateřina's campsite photo shoot (because nothing says "AI model" like a fake camping trip) and animate it. And let me tell you, it turned out surprisingly well! It's like Kateřina suddenly came to life, probably wondering why she's stuck in a perpetual camping trip.

Here's how the magic happened:

  1. I took this image generated with Stable Diffusion 1.5 (imagine a charming campsite scene here, minus the bugs and with perfect lighting).
  2. I uploaded it to Luma Labs, crossing all my fingers and toes.
  3. I used the following prompt, channeling my inner film director:"Happy woman at a campground, lantern by her side, warmly laughing and smiling gently"

And voilà! Kateřina was suddenly more 'real' than my houseplants (which, admittedly, isn't saying much).

0:00
/0:05

The Future is Moving... Literally

I'm consistently shocked at how these AI tools keep evolving and changing the way we look at the world. It's like waking up every day in a sci-fi movie, but with better special effects and fewer alien invasions (so far).

I can imagine that within a year or so, Denis Villeneuve might be using these tools for "Rendezvous with Rama" or whatever else is on the docket in Hollywood. Who knows, maybe the next big blockbuster will be "Attack of the 50-Foot AI Model" starring a gigantic, animated Kateřina. Hey, Hollywood, if you're reading this, I want royalties!


If you've created something cool in Luma Labs recently, I'd love to see it. Maybe you've brought your own AI character to life, or perhaps you've animated your pet goldfish reciting Shakespeare. Whatever it is, hit me up on Warpcast or BlueSky. Let's create a support group for people addicted to animating AI images!

Have fun.


  1. Crafting Cohesive Characters: Consistency with Stable Diffusion Embeddings
    Warning: Reading this article may result in an irresistible urge to create your own AI character and plan their entire life story.
Crafting Cohesive Characters: Consistency with Stable Diffusion Embeddings
This is the first article of our series: “Consistent Characters” Understanding Embeddings in the Context of AI Models Embedding in the context of Stable Diffusion refers to a technique used in machine learning and deep learning models. It involves the transformation of data, such as text or images, in a
  1. ControlNet batch fun, a camp photoshoot with Kateřina
    Discover how to stage a faux photoshoot with your AI character. No actual camping skills required!
ControlNet batch fun, a camp photoshoot with Kateřina
In this article we will be using the txt2img tab to batch up a ControlNet job on multiple depth maps to transfer a character setting to a character position Here’s our situation, we have the lovely Kateřina as our model today doing a camping photoshoot, let’s generate some images with