Video generation has matured: you can now direct models instead of prompting and hoping

MilitantAI · x · 2026-09-03

MilitantAI reflects on how video generation leapt from the rough early days of Stable Video to today's high-quality output in just a few years. Kinks remain, but users can now 'direct' models rather than prompt and hope for the best—which puts consumer AI video products at an interesting spot: as capability improves, interface and interaction design become the differentiators.

Related event: Video generation evolves from prompt luck to directorial control(2 posts)→

Original post →

More from Multimodal

Multimodal channel →