Despite divisive opposition to AI from many in the film, video, and other creative industries, AI seems here to stay. And, as we’ll likely see in various comment sections when this article is posted, many might not care which AI is better for video.
Yet, with personal anxieties and opinions aside, there are us filmmakers and video cadets who are interested in exploring the opportunities AI presents. The arms race to develop the best image-to-video editor is raging in full force and it’s important to keep up with the latest technologies.
Two apps in particular leading the artifical arms race are Runway (which we’ve written about before) and Pika (which we’ve also covered).
While the technologies are both still evolving it’s certainly worth it to take a look at how the two stack up against each other—at least for those of us who simply want to learn this new technology and explore how it might be helpful for our own AI-powered video endeavors. (We’re no scabs!)
Runway ML vs Pika Labs
Alright, before we dive into the side-by-side comparisons between the two, we need to do a bit of history on each company for those who might not be too familiar with each.
Runway (Runway ML) was launched in 2018 and has made inroads in the world of both amateur and professional video editing for several years. Most famously, Runway tools were used on Oscar-winning films like Everything Everywhere All at Once by the film’s visual effects team.
Over the past year Runway rolled out several new features powered by its Gen-2 update, notably its Director Mode and custom camera controls.
On the other hand, Pika Labs developed on a path more akin to Midjourney as it has mostly operated on its own Discord server. Despite that it is still one of the most promising AI apps for generative video.
Most notably, with the Pika 1.0 update launch, it’s become one of the best true idea-to-video AI tools in the space with its Camera Control features.
Comparing Runway and Pika
Now, for the test we’ve mentioned, we’ll be looking at a comparison done by the YouTube channel AI Video School, which—as you might imagine—is a channel focused on exploring the vast and growing world of AI video.
In the video, our host goes into the background of each app as well as explores how developed each platform is. (Note: there have been a few updates since this video published in September 2023, but for the most part these tests are about as up-to-date as anything in this AI space can be).
The test centers on a cinematic view outside of the Louvre museum in Paris, France, exploring how the different AI algorithms can handle adding motion and camera moves to a stationary 3D object. As you can see in the test, the results are similar but noticeably different in several ways.
Reviewing the Comparison Results
Ultimately, these results focus on just one prompt and sample from each app. However, we do get to see how these two AI algorithms stack up against each other. In the video, we hear what our host favors between the two, with Pika seeming to be taking the lead with its smoother and more accurate camera controls.
You can judge yourself by watching the video as well as trying out both, as each app is available to use with minimal investment (some plans even starting free for a couple minutes). It’s hard to say at this point if this sample is super indicative of which AI will be better in the future though, as there are so many other aspects of video that we want to eventually test.
With new AI technologies and innovations developing every day, and industry leaders like Runway and Pika both hard at work refining their algorithms and adding new features, we’re in for a wild ride as these technologies continue to advance at breakneck speeds.
Author: Jourdan Aldredge
This article comes from No Film School and can be read on the original site.
