Ever feel like your AI videos are stuck in a rut? They either look cool but are silent, or they’re just a talking head that can’t do anything.
It’s a classic ‘show’ vs. ‘tell’ problem.
Well, I just stumbled upon an incredible post from Tianyu Xu that completely reframes this. He’s figured out that Google’s new model, Veo 3, can do BOTH in the same clip! He calls it the ideal model for “show and tell,” and after seeing his examples, I think he’s spot on.
To get these amazing results, Tianyu Xu shared the 5 rules he follows to nail his prompts and increase his success rate. This is pure gold!
⚙️ Tianyu Xu’s 5 Rules for Veo 3 Success:
- 📌 One Job at a Time: Don’t try to make the AI multitask in a single prompt. Keep your request focused for better results.
- ✍️ Plain English is Key: He says to describe the visual effects or actions in simple, clear language. No need for overly technical jargon.
- 🖼️ Focus on the Image: Put double the effort into describing the static image you want to see. A strong visual foundation makes for a better video.
- ✅ Embrace Imperfection: Don’t aim for a perfect 10/10 clip every time. He points out that you can use the best parts of a generation to get what you need.
- 🚀 Use the Right Tool for the Job: He wisely reminds us that Veo 3 is fantastic, but it isn’t the only model. For “overly wild” effects, he notes that other models might be a better choice.
Bottom line: Tianyu’s tests show that Veo 3 is a game-changer for anyone in education or storytelling. It’s not about crazy effects, but about mastering that perfect blend of showing AND telling.
This is just my takeaway, you have to see his original post for the full details and examples. He even links to his course on it!