AI fashions that may generate movies from textual content or photos have woefully lagged behind conversational and picture generator AI instruments for much too lengthy.
But the text-to-video house is lastly catching up.
Just final week, two enormous startups within the AI sphere, Stability AI and Meta AI, unveiled their very own text-to-video instruments—Stable Video Diffusion and Emu Video. Both appear promising, however in my opinion, they nonetheless have progress to make.
Today, one other startup, Pika Labs, is about to affix the get together with their very own AI video generator software known as Pika 1.0.
What is Pika?
Pika is a platform that may generate movies from textual content or photos utilizing synthetic intelligence. The goal of the software is to revolutionize how movies are made and edited by means of a state-of-the-art video foundational mannequin.
According to its creators, anybody can channel their inside Scorsese or Spielberg.
Our imaginative and prescient for Pika is to allow everybody to be the director of their very own tales and to convey out the creator in every of us.
I couldn’t agree extra. Their aim resonates deeply with my perception that these AI applied sciences may empower peculiar folks such as you and me.
Here’s the way it works
I attempted Pika on their Discord server. Under one of many “generate” channels, hit the “/” key in your keyboard to convey up the menu.
From right here, you’ll be able to both generate a video from a picture or a textual content description. Let’s attempt an instance with this immediate:
Prompt: An astronaut driving a horse on the moon
It took lower than a minute for the AI to generate a 3-second video in the identical channel on Discord.
