Google’s Veo 3 is changing the game for AI-made videos, and The Sentence, a gripping short film, shows just how far this tech can go. Thrown together in just three days, it pulls you into a dark, dystopian world where a chilling new kind of punishment unfolds, all crafted with a handful of AI tools that make the impossible feel real.
The filmmaker, Hashem Al-Ghaili, behind The Sentence worked fast, pulling off writing, video creation, and final rendering in a whirlwind 72 hours. They leaned on Google’s Veo 3 for visuals and sound, Flow for editing, Suno AI for a haunting score, and ChatGPT to sharpen their prompts. “From first draft to finished film, it took just three days,” the creator said, and you can feel the urgency in every frame. One viewer called it “straight out of Black Mirror,” while another said, “It’s amazing and downright spooky that AI made this.”
No products found.
Veo 3 is like a storyteller’s dream machine. Google DeepMind calls it a “state-of-the-art generative video model” that churns out videos with built-in audio—dialogue, sound effects, even ambient noise. “It nails real-world physics, like the sound of splashing water or footsteps in snow, and syncs lips to dialogue perfectly,” they explained. Whether you’re crafting gritty sci-fi or a slick ad, Veo 3 takes text or reference images and spins them into high-quality footage. You can call out specific camera moves—pans, zooms, you name it—for a pro-level vibe.
Unlike older models, Veo 3 doesn’t just do visuals; it weaves in soundtracks, spoken lines, and effects in one seamless package. The result? Scenes like the tense exchanges in The Sentence feel alive. It also nails real-world details—think water sloshing or fabric fluttering—making everything feel like you could reach out and touch it. Plus, it keeps characters looking consistent across scenes, which older AI tools struggled with, so the story holds together.
While the creator kept their exact prompts under wraps, other users say Veo 3 shines with clear, concise descriptions—four or five sentences that paint the scene, characters, and actions. Something like “A woman with half-pink hair in a blue shirt stands on a tree-lined path, people milling in the background, delivering a tense monologue” could spark a vivid moment. The film’s polish also owes a lot to Suno AI’s moody music and ChatGPT’s prompt tweaks, which likely made the inputs crystal-clear. It’s a workflow not unlike what other AI filmmakers, like naviSTFU, have done—mixing Veo 2 and 3 with Midjourney and Premiere for a sci-fi short, burning through $125 and 10,000 tokens in just hours.
