Image credit: Google
Google just dropped its latest AI video generator, Veo 3, and the internet is buzzing! These clips look so real that most people online can’t tell them apart from videos made by human filmmakers and actors. It’s pretty amazing—and a little unsettling—how the line between real and fake is starting to blur.
Why This Matters
The realism of Veo 3 videos is blowing people away, but it’s also got some folks worried about how hard it is to spot what’s authentic anymore. Shared online, these clips are leaving viewers both impressed and a bit spooked.
Unlike OpenAI’s Sora, which rolled out more widely last December, Veo 3 stands out by adding dialogue, soundtracks, and sound effects to its videos. This AI model is great at handling detailed prompts, turning your ideas into lifelike scenes. It follows real-world physics, nails lip-syncing, keeps continuity smooth, and even creates people with natural features—like five fingers per hand. Examples from Google and users online show almost no signs of that typical “synthetic” look. Take a viral X post by filmmaker and molecular biologist Hashem Al-Ghaili, who shared short films of AI actors rebelling against their creators—it’s wild!
The Big Picture
A New Era for Filmmaking
Special effects, video apps, and camera tech have been shaping Hollywood for decades, but AI-generated films bring a fresh challenge for human creators. In a promo for Google’s new Flow tool, which includes Veo 3, filmmakers like Dave Clark shared how it gives them new freedom with a touch of spooky independence. “It feels like it’s almost building on itself,” Clark said, hinting at the tool’s unique vibe.
How It Works
Veo 3 was unveiled at Google I/O on Tuesday and is now available to Google AI Ultra subscribers in the U.S. for $249 a month. Google says it drew inspiration from working with creators and filmmakers, and some are excited about these tools. But the flood of videos online is also upsetting many pros and art lovers who feel differently.
The Debate Around AI Videos
Some call any AI-generated video “slop,” no matter how good it looks—though, as Axios’ Ina Fried notes, that’s a matter of opinion. On the flip side, AI analyst Ethan Mollick sees potential for commercial marketing and media projects. We don’t know exactly how Google trained Veo 3 or how that might shape its creativity. For instance, 404 Media found it spat out the same cheesy dad joke for users asking for a stand-up comedy video. Last year, YouTuber Marques Brownlee tested Sora with a “tech reviewer at a desk” prompt and got a video with a fake plant nearly identical to his own desk shrub—hinting the tool might’ve learned from his content.
What’s Next?
As hyperrealistic AI videos get easier to make, we’re just starting to figure out big questions like authorship, consent, rights, and what this means for the film industry. It’s an exciting time to watch!
What do you think about Veo 3 and its impact? I’d love to hear your thoughts!
