You may have heard of the amazing results achieved by text-to-image AI such as Dall-E, Stable Diffusion, and Midjourney. As you might expect, this revolution is moving forward, with text-to-video AI tools next in line.
Google and Meta teased their text-to-video capabilities in research papers from their AI labs, but this advanced technology is not yet available to the public. If you’ve been eagerly waiting for the chance to try creating an entire video using simple AI cues, now’s your chance, thanks to QuickVid.
Before your expectations climb too high, it’s important to realize that this is not the same as generating thousands of Stable Diffusion still photos and combining them to make a video or accessing the world’s most advanced AI system to generate a real video . This is an early entry in the text-to-video solution race.
The first step in the AI process is to generate a script based on your prompts. I created a short YouTube video to test the system with the following text: “A video of a DJI drone flying over astronauts on Mars, ending with a shot of the surprised astronaut reacting.”
The AI wrote a full 79-word narrative based on my prompts, then synthesized the speech with a choice of male or female voices. TechCrunch notes that the background video chosen for the resulting video was pulled from an inventory library, apparently with lots of “astronaut on Mars” footage.
As a dubious finishing touch, QuickVid overlays the script as the title and adds a thumbnail generated by the Dall-E API. The resulting YouTube clip seen above is… interesting. Perhaps, it will handle more Earth videos better.
In an interview with TechCrunch, QuickVid’s developers said improvements are coming, with more personalization options arriving in January. Eventually, QuickVid will also include subtitles and supporting avatars.
More text-to-video solutions are likely to emerge next year, along with other visual marvels such as AR glasses and more advanced VR headsets. This should be exciting.
Editor’s Choice
