Runaway Gen 2 AI is ChatGPT for videos & you can try it now
Runaway Runaway Gen 2 is now publicly available. The AI tool allows you to create a four-second clip based on a text prompt like ChatGPT and is available to try for free.
The onslaught of artificial intelligence tools has no signs of stopping, with tools like DALL-E, ChatGPT and more becoming increasingly popular. Now, Runaway’s latest tool, simply named “Gen 2” is available for free, and allows you to generate up to 105 seconds of video using its text-to-video interface. This equates to around 26 videos in total, given the tool’s limitation of four-second videos using a free account.
To use the tool, you first need to head over to Runaway’s website and create an account. Once completed, head to this link to access the tool. Runaway boasts that the tool can help users storyboard, conceptualize, or just generate a short video out of ideas in their heads.
The tool outputs a four-second MP4 clip. Though, in our testing, we found that the tool has some issues with animating motion. Additionally, the output is relatively low-resolution at just 768×448. The resolution of the output can also be enhanced with Runaway AI’s paid-for options.
How to use Runaway Gen 2 AI
Runaway Runaway Gen 2 AI retains almost everything that made ChatGPT a success, including a simple UI and text box where you enter a prompt in order to generate an output from the AI itself. Additionally, if you wanted a certain look, you can also upload an image for the AI to take notes from. Once complete, the tool will have generated a video that you are able to save to your account’s profile, and locally.
We’ve detailed the full process below:
- Sign up for a Runaway account
- Head to the Gen 2 tool
- Input your prompt
- Optional: Add an image to use as “inspiration”
- Wait for the AI to generate the output
It’s a relatively simple process, but our results were wholly mixed. Prompts without an additional dose of inspiration can occasionally look simple, and others have very little movement or animation in them, too. Though, supporting this with “inspiration” for the AI to look at seems to have improved the results.
We found particular success when attempting to generate landscapes, while it struggles to animate single subjects. Prompts involving more surreal subjects like “A frog man” also did not work out too well. When attempting to draw a prompt of “an angry man” with my face attached, it also happened to change the race of the subject in the generated video, calling into question whether or not racial bias in AI is indeed an issue for the fledgling industry.
AI is being implemented everywhere from Windows to Photoshop and more, and it seems that Runaway’s Gen 2 text-to-video is just another step in the direction of forthcoming AI dominance, where experts have urged for more regulation in the industry as a whole.
