How to Generate Videos With AI.

Learn how to generate videos with AI using RunwayML Gen-2. Just input a description, an example picture or a video and the artificial intelligence will generate a video for you.

Generate AI Videos With Gen-2.

Gen-2

Gen-2 can be found as one of the tools within Runway, an AI visual effects suite.

You can generate videos by offering text, photo or another video, and the tool will generate a 4 second video with movement.

Tips and Tricks for Generating Videos With Gen-2.

While there are no strict rules for prompting in Gen 2, this formula tends to yield better results: style, shot, subject, action, setting, and lighting.

Begin your prompt with a style keyword, like cinematic, action, animation, black and white, etc.
The shot refers to the camera angle, such as wide, medium, close-up, or extreme close-up.
Then specify the subject, which can be anything from characters to objects or locations you want to include. Keep character descriptions simple, like “man with grey hair wearing sunglasses”
Action should describe what the characters are doing, but remember that Gen-2 is better at simple actions like walking, talking, or running. 

The setting describes the environment, such as a city, beach, or spaceship. 
Lastly, use keywords like sunset, sunrise, day, night, horror film lighting, or sci-fi lighting for the desired lighting effect.

Using a reference image can help guide Gen 2. Try to create characters and settings within mid-journey and use them as storyboards for Gen 2. Keep descriptions simple for better consistency.

Gen-2 by Runway Review.

While Gen-2 may not always produce exactly what you expect, it can lead to interesting and artistic results.

Remember that upscaling improves the resolution significantly, and Gen 2’s Discord version has some additional features that might be added to the web-based version in the future.
The IOS app is quite reduced in it’s capabilities, the video generation tool works quite well.