VEO3 AI Filmmaking Is Here...Buckle Up.

THIS POST WAS WRITTEN ENTIRELY WITHOUT AI.

Google’s new AI video generator, VEO3 was released May 2025, and just weeks later has the already fractured film & television industry gasping for air.

AI image and video generators like Midjourney and Sora have existed for a “while” in the exponentially accelerated technology space, which to the rest of us seems like the blink of an eye. But VEO3’s debut has done more than caused a stir, it has left veterans from the film, fashion, and entertainment industries completely shaken.  

VEO3 IS UNLIKE ANYTHING WE'VE SEEN BEFORE

Both programs have their strengths and weaknesses like video length, realistic motion, and the ability to interpret and generate cohesive narratives - so why is VEO3 catching so much attention? 

One aspect is its ease of usability, as the user interface of VEO3 makes this AI video generator more appealing to non technical users. More access to the technology means more users, which rapidly accelerates AI video’s effect on entire industries as well as society.

There is an increasingly disturbing difficulty in identifying the difference between an AI generated videos and reality. Messages on social media reveal these growing concerns about the potentially disastrous uses of the technology. Without a doubt, the best way to see what VEO3 can do is to see it yourself.

AI FILMMAKING AND YOU, DEAR CREATIVE.

Hollywood is already facing an existential crisis as masses of long-timers in film & television have pivoted or left the industry after Covid and the strikes weakened an already broken business model, disrupted by streaming television. Creatives used to commercial budgets and continuous work will need to adjust their expectations even further with pay rates cut by more than half, work outsourced to international locations, equating to fewer roles available for less money.

Network television and now even streaming networks are seeing their models crumble and are adopting the AI technology as quickly as possible to try and salvage what they can of their empires. Most major studios have stopped green-lighting projects unless they are AI, so entrepreneurial-minded individuals and creators, especially those with industry experience are seizing the moment, seeing a unique opportunity to produce and develop their own projects like films, series, and even sitcoms on streaming video platforms like YouTube as well as taking steps to understand and adopt AI technology into their production pipelines.

UNDERSTANDING THE TECH, EVEN IF IT HURTS.

While many readers may not have an understanding nor interest in the technical details, Producers, Editors, Directors, Videomakers, and other ATL (above the line) and post-production people will benefit from having some knowledge of the basic framework for AI video generators like Sora and VEO3, and having some terms to use in the very near future. 

So take a breath, go slow and buckle in for the tech of it all. 

The first thing to know is that the main difference between Sora and VEO3 is the model on which they were built, which thereby affects how the AI will interpret the written prompts.

AI video generators are built with different models. 

Each model is a different approach consisting of a cocktail of AI technology that tells the AI how to read and then interpret the data you feed it. All of this falls under the umbrella of “Machine Learning”, a branch of technology that studies quite literally, how machines learn. 

COMPARING SORA WITH VEO3.

SORA

Sora is built with a hybrid of transformer models that includes a Generative Adversarial Network (GAN) model that gives Sora an ability to create 1080p video. Sora has a reputation for realism, able to replicate realistic textures and objects, natural movement, able to generate videos up to 60 seconds in length.

VEO3

By contrast, VEO3 is built with (operates on) a diffusion-based architecture model that includes a process of cleaning up noise to transform it into images and video in 4k results in striking, cinematic video that is optimized for YouTube Shorts and the Google ecosystem. The video length is limited to 20 seconds.

Graphic from iamdave.ia 

As always, art and storytelling come in many forms and can be expressed across countless mediums. It remains to be seen to what extent Google’s VEO3 AI video generator release will affect or even interest creatives in the beginning but there is no doubt that the business of entertainment will never be the same. 

Save this article to check back in on your AI video knowledge as it drips its way into the creative atmosphere. Comment to let us know if you’re using any of these programs!