SlatorPod #250 HeyGen CTO Rong Yan on AI Video Generation and the Language Challenge
May 16, 2025
Rong Yan, the CTO of HeyGen, shares insights on the company's shift from a Metaverse startup to pioneers in AI video generation. He highlights the rapid growth achieved, attributing it to a focus on quality, consistency, and controllability. The latest Avatar IV model brings full-body animations and emotion to life, setting a new standard in video realism. Looking forward, Rong envisions tools that let anyone create videos from simple prompts, revolutionizing storytelling and making high-quality video production accessible to professionals everywhere.
AI Snips
Chapters
Transcript
Episode notes
HeyGen's Pivot to AI Avatars
- HeyGen started as a metaverse-focused startup exploring 3D CG avatars, which didn't work out initially.
- The pivot to AI-generated virtual spokespeople led to rapid growth from zero to $1 million ARR in six months.
Quality fuels sustainable growth
- Viral growth is influenced by luck but must be backed by high product quality to sustain users.
- HeyGen focuses on upgrading avatar quality from 70% to 95% for real user value, not just viral spikes.
Avatar IV transforms realism
- Avatar IV generates full-body avatars with synchronized gestures, breathing, and emotion, a game changer in realism.
- It supports profile views, animals, cartoons, and even sketch drawings that can be animated to speak.