Tomorrow Tuesday — July 2026 (Issue #2)
Welcome to another issue of Future Tuesday Tomorrow Tuesday, where we cover the revolution running through film, video games, music and the creative industries.
A newsletter by SOUTS, where we explore what happens when artificial intelligence stops being just a tool and starts changing the way we create and consume stories.
SIGGRAPH 2026: Let’s Nerd Out a Little
One of the most anticipated moments of the year for nerds — and completely invisible to people with a social life — has arrived: COMIC-CON SIGGRAPH.
SIGGRAPHis the big annual conference dedicated to computer graphics, animation, visual effects and research applied to digital imagery. It’s one of those places where technologies get shown that today look like weird experiments and a few years later end up inside the tools we use every day.
We didn’t make it this year, but we’ve been going through announcements, papers and demos. Here’s the coolest stuff:
🤖 AI agents, AI agents, AI agents
Everybody has their agent and everybody wants to plug it into creative software.
There were demos of integrations with tools like Adobe, Blender and Unreal Engine. Adobe is also expanding its assistants so they can run multi-step processes inside Creative Cloud.
Until now, you could ask a chatbot how to do something in Blender. Now the agent can start operating the software for you.
We like this because the technical handling of the software tends to be the first barrier to entry when learning a new discipline, especially with things like 3D programs.
📦 Generative 3D is almost, almost there
There was a lot going on in image-to-3D, mesh generation, textures and rigging.
Over the past few years we’ve seen spectacular image-to-3D demos that weren’t actually usable in real life.
Now it looks like we’re basically there.
The goal is no longer just to generate something that looks 3D from one specific angle, but something that can be edited, animated and dropped into Unreal, Blender or a video game.
Generative 3D still has a way to go before it’s fully reliable, but it’s coming.
🌎 World models and robotics
This is probably the buzzword of the year.
A world model tries to learn how an environment works: how objects move, what consequences an action produces and how a situation can evolve over time.
This can be applied to video generation, animation and video games. But the real money is in robotics.
Training robots in the real world is expensive, slow and dangerous. That’s why there’s so much money going into training them in virtual worlds before sending them out into the physical one.
🩰 Better movement, more control
We’re already bored of generating images. The new obsession is getting control, especially over movement.
There were quite a few papers and demos about tools that let animation studios train movement on their own animations, as a way to preserve the style of that movement.
Same thing with actors and performers. Not just reproducing a person’s face or voice, but capturing the specific way they walk, react, occupy space and build a performance. Because Leo DiCaprio’s performance isn’t just Leo’s face, it’s how Leo moves.
AI in Entertainment
🌳 Black Forest Labs introduced FLUX.3, their new multimodal model. It’s not an open release yet and, for now, we’re seeing experiments and limited access. But it makes you really want to try it. It generates image, audio, video and action prediction. It’s being talked about as competition for Seedance 2.0.
FLUX.1 won our hearts, partly for having been one of the best open source models. FLUX.2 lost a bit of that enthusiasm, mostly by not being open source.
Let’s see what happens with FLUX.3.
🏃♂️ Runway looks like it’s in the middle of a pivot. They haven’t put out a new model in a while, and instead have focused on other things: MCP, Agents and now they announced their Media Router. You no longer worry about picking a model — depending on what you need, Runway picks the image, video and audio model for you. You just define how much you care about price, quality and speed and the choice gets made for you. Doesn’t quite add up for me, but okay…
📽️ A big chunk of the industry is buying into the convergence between different media: streaming, video games, music, podcasts and books are all converging toward a world where stories can be liquid across formats. That opens the door to a Universal Entertainment App. The idea is a sort of Netflix, but one that doesn’t just have movies — instead it has every possible kind of media in one place. Every content platform is going to be fighting to be that Universal Entertainment App. Will this actually go somewhere, or are we going to forget about the concept in a couple of years?
💰 We finally have a number: Netflix revealed they paid $587 million to buy Ben Affleck’s company, InterPositive. It’s focused on post-production and special effects, though we still have no idea what it actually does.
👿 Higgsfield keeps positioning itself as the Nestlé of generative AI: the less you know about the company, the nicer it seems. Higgsfield just announced its new terms and conditions for the platform. Now, everything you create on Higgsfield is going to be yours (good), but they get perpetual, irrevocable, worldwide and royalty-free rights to use your inputs and your outputs for model training, product development, marketing and other uses (bad).
New Media
⛵️ The Odysseyis heading into its second weekend. It’s been the spectacle of this second half of the year so far, which doesn’t say much since we’re only one month in. It has already made more than $260 million in the United States, which shows that people do want to go to the movies — but they want to go see original stories, and they want it to be an event.
But beyond film, the media world has been moving a lot.
🧒 Microdramas keep growing and their audience keeps getting younger. A couple of years ago, men made up only 1.1% of the audience for this kind of content. Today they already make up more than 30% of the audience. It’s also skewing younger and younger, which suggests microdramas are going to keep carving out space within mobile-first entertainment. It’s not a medium for older ladies anymore.
So much so that TikTok is testing its own microdrama platform outside the main app — they’re calling it LimeShorts. We’ll have to see how it does, since the result is probably going to be the most decisive impact on this new category.
👾 Universal just signed a deal with Atari to make potential movies based on 10 Atari games. The most striking part is the titles involved, which include Pong, Asteroids, Centipede and Breakout. It almost feels like a desperate grab: any IP people recognize seems to have gained a whole lot more value, to the point where none of these franchises has a narrative, a story or even characters behind it. All they have is a name.
And LATAM?
While there was some Latin American presence at SIGGRAPH, it was pretty minimal. Some basic research using ChatGPT tells me less than 1% of the technical papers presented had Latin American participation.
I don’t know if that’s bad in itself — it’s a conference in the United States. It’s expensive to go, and it’s not like universities around here have that much infrastructure.
Our opportunity, beyond the academic side, is in applying these technologies. Every paper presented is available for anyone to take advantage of.
At SOUTSwe try to always stay on top of what’s happening, and every time there’s something relevant we’ll bring it to you in this newsletter.
See you next Tomorrow Tuesday.
Yves
