I revived my dog's puppy days using Google Gemini AI – Here's how it went

Mint

2 days ago

My pet dog, Coco, is starting to get quite old now; he's almost 10. We got him in 2015, and at that time, while phones had decent cameras, I did not take many videos, primarily because they were not very good. This is why I have barely any videos of my dog when he was a tiny pug puppy, which is a big regret of mine.
Now, when Google Gemini started to roll out the feature that allows you to convert photos into videos using, of course, Veo 3 AI, I immediately jumped at the opportunity and tried it out. To my surprise, it did a fairly accurate job. Here is how it went.
Firstly, I opened the Gemini app on my iPhone (it also works on Android phones), and because I have access to the Google AI Pro plan, I have access to Google Veo 3 preview, which is the AI tool Google uses to make videos.
Then, I submitted a photo of my pet dog, Coco, when he was a puppy, being held by my brother. And next, a prompt along with it that said, 'Make a video of my pug puppy from this photo and try to keep the details intact.' The AI-generated video Google Veo 3 made. (Converted to GIF)
Gemini took around a minute or so before generating an eight-second-long video of my little Coco moving around in my brother's arms. It looks quite realistic, and I am beyond happy that I got something. While it is definitely not real, it is indeed satisfying to see a moving picture of my dog from his puppy days.
Honestly, I'm not sure how I feel at large because this video isn't real. It's something that never actually happened. But maybe that's the point. It offers a glimpse of what could have been and perhaps, that in itself is comforting.
This is possible with the Veo 3 preview only. It gives you 8-second videos. Veo 3 only launched in May, and Google is now shipping the preview for users. This capability is only for Gemini AI Pro and Gemini Ultra subscribers, which means you will have to pay roughly ₹ 2,000 to access Google Gemini (there's a free trial available).
Google says people have been creating a wide range of content using Veo 3, including everything from ASMR videos to cutting through lava.
Having said that, I also tried it with photos of myself when I was a kid, but it did not allow me. Google is also bundling in safety measures, including adding a visible watermark to show they are AI-generated and an invisible SynthID digital watermark.
And, you can definitely use the Google Veo 3 preview to generate a wide range of content from making talking monkey vlogs set on the streets of Delhi to flying, talking giraffes climbing Mount Everest. The world is your oyster, though, of course, within the limits of Google's guardrails.