See how companies are responsibly integrating AI into production environments. This invite-only event in SF explores the intersection of technology and business. Click here to learn how to participate.
Earlier this year, OpenAI surprised many in the technology community and the media and arts by showing off a new AI model called Sora that creates realistic, high-resolution, smooth videos of up to 60 seconds per clip. At the same time, it also raised the feathers of traditional videographers and artists. .
This technology is not currently available to the public. OpenAI said at the time in February 2024 that Sora would be “available to red teams to assess critical areas for harm and risk,” and that it would also make Sora available to a select few “visual artists, designers, and filmmakers.” ” But that didn't stop some of its first users from creating and publishing new projects with it.
Now, writer/director Paul Trillo, one of OpenAI's hand-picked Sora early access users, became one of the first in the world to demo a third-party video created using this model in March. I created something called “.The first official music video created using OpenAI's Sora.”
of video This song was created for indie chillwave musician Washed Out (Ernest Weatherly Green Jr.) and his new single “The Hardest Part.” It's basically a four-minute series of quick zoom shots of different scenes, all strung together to create the illusion of continuous zooming. Watch below:
VB event
AI Impact Tour – San Francisco
The next stop on VB's AI Impact tour in San Francisco will explain the complexities of responsibly integrating AI into your business. Don't miss the chance to gain insights from industry experts, network with like-minded innovators, explore his GenAI future with customer experience, and optimize your business processes.
request an invitation
In his account on social network X, Trillo has been posted He first came up with the idea for the video 10 years ago, but abandoned it.He also answered questions from his followers and the video was Made from 55 individual clips Generated by Sora from a total pool of 700 and stitched together in Adobe Premiere.
Separately, but relatedly, Adobe recently announced that it is considering adding Sora and other third-party AI video generator models to its subscription Premiere Pro software, but the timeline for this integration is unclear. In the meantime, those who want to emulate Trillo's workflow can generate AI video clips with other third-party software such as Runway or Pika (Sora remains private). ), you must save and import them into Premiere. It's not the end of the world, but it's not as seamless as this.
In an interview with Los Angeles Times, Washed Out/Green says: So if it's something pioneering, I'd love to be a part of it. ” The interview also describes some of the specific prompts used.
“Greene needed to create a prompt that included enough specific details not only about the image itself, but also about the angle of the shot and the movement of the characters. “We zoom into an open soccer field with bubbles popping and zooming bubblegum,” Trillo writes as part of the video's short snippet prompt. “The scene is moving rapidly, showing a frontal perspective, and you can see the students getting bigger and faster.”
I also posted one using Trillo Only text-to-video functionality in the modelRather than taking still images captured or generated elsewhere and feeding them into AI to add motion (a common tactic among artists in the rapidly evolving AI video scene).
This example shows Sola's power in creating media using AI, and one of the first demo videos, “Air Head,” created by Canadian creative studio Shy Kids, features a video of a man holding a balloon. It helps to reconfirm the recently revealed information that a man will appear. In fact, he relied heavily on other of his VFX and video editing tools, such as rotoscoping in Adobe After Effects.
And even though many other creators have criticized this technology, and OpenAI in particular, for being exploitative and infringing on the copyrights of human artists by scraping and scraping some of their music and videos. shows the continued desire of creators to use new AI tools to express themselves and tell their stories. Providing training on previous work without informed consent or compensation.
