Ever get frustrated that your AI videos can either look cool OR have a character talk, but rarely both at the same time? It’s a huge pain for storytelling.
Well, I just stumbled upon an incredible post from Tianyu Xu that completely blew my mind. He’s been digging into Google’s new Veo 3 model, and it’s a total game-changer for exactly this problem.
According to Tianyu, Veo 3 is special because it can both ‘show’ (create the scene) and ‘tell’ (have a character act and seemingly speak) in the same clip. I think this is massive for creators and educators!
His 5 Golden Rules for Great Results:
- 📌 One Job at a Time: Don’t ask the AI to multitask in a single prompt. Keep it focused.
- ✍️ Keep it Simple: Describe actions and effects in plain English. No need for fancy jargon.
- 🖼️ Focus on the Image: This was a huge insight for me! Tianyu says to double down on describing the still image you want. The video will follow from there.
- ✅ Good Enough is Great: Don’t chase perfection. He advises using the good parts of a clip and building from there.
- 🚀 Use the Right Tool: Tianyu notes that while Veo 3 is amazing, it’s not the best for overly wild effects. Other models are better suited for that.
This is just my quick summary of his awesome work. For all the details and to see his Veo 3 avatar in action, you HAVE to check out the full post from Tianyu Xu himself!