Generated Weekly: What is good AI art?
Plus slopoganda and a new model by Alibaba.
There’s a battle raging online between digital creators. Some creators, usually those with a track record of shooting traditional video, see the work of AI video artists who share something visually stunning but without soul.
For example, take this by “AI filmmaker” Dinda Prasetyo:
Is this the future of action cinema?
I challenged the AI to blend high-stakes urban parkour with a heavy-duty mecha transformation, and the result is more intense than I ever imagined. The prompt was a chaotic mix of "exploding bridge" and "biometric armor deployment," and seeing the suit snap together mid-run honestly gave me chills.
The video follows a futuristic heroine sprinting through a collapsing cityscape, narrowly dodging debris and sliding under moving trucks before suiting up in winged armor to take on a massive, terrifying monster. It captures that perfect "main character" energy with every frame.
What do you think?
Made with Dreamina AI Seedance 2.0
Edit Adobe Video Premiere 2026
Upscale: Topaz Labs
Are you building an AI product? Work for a brand looking for AI video experts? We’re holding the first AI marketing video awards in New York this May! We’ll be posting more information this month but until then visit the website and sign up if you’d like to join as a brand sponsor.
Now we can assume that Prasetyo is producing content in earnest and hoping to express themselves as a creator. They’re not paying lip service to the idea of “filmmaker” even though their product is essentially a clip that could have come from a video game.
Then comes the criticism. The responses range from “You’re out of a job, Spielberg” to “This is junk.” Some choice comments:
You gotta have an actual story and characters people care about. Bored girls in mech suits and video game Kung fu fights aren’t it.
You can telll how many people in this post are scared of losing their jobs to a Robot. It’s going to be the future, you have to adapt just like cars or any future production. If you don’t you will be left behind. I know it’s not what many of them want to hear they rather call Ai slop but Ai is the future.
AI films will always appear mindless without an idea. And I don’t really sense any idea by way of origination and the first step here. Indeed AI was allowed to run the show, which is why we now have a very executional film that feels random, meaningless and a story only retrospectively being added (so quite contrived with it).
I hope this isn't the future of action movies. This looks way out of touch with reality and doesn't feel relatable at all. Films like Mission Impossible, The Raid and Christopher Nolan's films all feel exciting because we know humans are pushing themselves to the limit to get the great shot or do the impossible. AI video just cannot capture that and looks like Illogical slop.
Neither viewpoint is wrong. The classicists are looking for what Robert Pirsig would have called “quality,” whereas the technologists see the tools and their evolution. Both viewpoints are vital to the growth of this industry. To an artist, a paintbrush is made to capture reality on canvas. To a toolmaker, the paintbrush is great at slathering matte latex onto a wall. But when they both stand in front of a Monet or Pollock, they’re both in awe.
The critics are not just angry. They are pointing at something real. When everything is possible, nothing matters unless you choose to make it matter. A collapsing bridge means less when no one had to build it, light it, or stand on it. A stunt means less when no one risked a fall. There is a human memory tied to effort.
But the defenders are not wrong either. Every tool has looked crude at the start. Early digital video looked flat. Early CGI looked fake. Then it got better. Then it became normal.
What is missing right now is not skill. The systems can already generate spectacle. What is missing is restraint. The sense that a creator chose this moment and not the thousand others the machine could produce. The sense that behind the image there is a mind that cares about what comes before and after.
Right now, most AI video feels like a highlight reel with no game attached. A series of peaks with no climb. You see everything at once, and so you feel very little.
The older creators are asking for story, for character, for stakes. The newer creators are asking for freedom, for speed, for reach. Both are right, and both are incomplete.
Because the truth is simpler and harder.
Tools do not create meaning. People do. A camera never made a film. A prompt will not either.
The next phase will not be about better explosions or smoother motion. It will be about someone using these tools and deciding to hold back. To build a scene that takes time. To make a viewer wait. To risk being boring for a moment so that something else can land later.
In other words, the next phase will be the art, not the science, of generated video.
Alibaba just revealed it’s behind a viral AI video model dominating leaderboards
Alibaba has revealed that it is behind “HappyHorse,” a previously mysterious AI video generation model that has been quietly topping industry benchmarks, according to a CNBC report.
The model gained attention after appearing on global leaderboards without a clear origin, outperforming or matching leading systems in areas like motion quality and consistency. Its strong results put it in direct competition with tools from OpenAI and Google, including their video models Sora and Veo.
Alibaba’s confirmation ends speculation about who built the system and signals a more aggressive push into generative AI, especially video, which is emerging as the next major battleground after text and images.
AI-generated Lego videos and Trump’s poo-bombing: welcome to the Iran-US slopaganda wars
AI-generated videos are starting to reshape how political messages spread online, blending humor, animation, and current events into fast-moving, highly shareable clips.
Iran, for example, has been dumping out LEGO animations featuring pointing critiques of the Trump administration, a clever and fascinating way to spread propaganda through vitality.
Seedance 2.0: The Most Controversial AI Video Model in the World Just Landed in the US
CapCut has rolled out a new “Video Studio” that integrates Dreamina Seedance 2.0, a fast-rising AI video model, directly into a canvas-based editing environment.
Instead of a traditional timeline, the new tool lets users create videos from start to finish in one space, handling everything from idea generation and storyboarding to scene creation and final edits. The goal is to simplify the process so creators can turn prompts, images, or rough ideas into finished videos with minimal manual work.
Seedance 2.0 itself is a text-to-video model that has gained attention for producing more realistic motion, stronger visual consistency, and better control over scenes compared to earlier tools.