In an age where new AI startups seem to be everywhere, Guo is making his company Pika stand out with software that allows anyone to type in a few words and get crisp video clips in response. It is working.
Mr. Guo started Pika in April 2023. This was born out of her work as a graduate student at Stanford University, which focused on creating content using AI, and a setback when she was trying to create a short film. (She designed an algorithm that took live-action footage and made it look like animation.) “I realized that the process of creating video was still very difficult,” she says, recalling her thoughts at the time. Masu. “She couldn't have imagined that five years from now, AI would have the same process. It's so complex and inefficient.”
Guo left Stanford to focus on Pika and, along with co-founder Chenlin Meng, released the first AI video creation software late last year. This program allows users to generate short videos from written prompts, still images, or other videos. Sound effects capabilities make clips more realistic, such as an AI-generated clip of an impossible pink waterfall combined with the sound of flowing water.
For now, Pika is one of the few companies that offers text-to-video software. Perhaps the most high-profile competitor, OpenAI Inc.'s Sora software, does not yet have a release date. Pika has raised $55 million in funding.
Guo expects things to continue to change rapidly. “Every month, every three months, we see huge advances in the industry,” she says. “And I think we are one step closer to achieving our dream.” —Rachel Metz
