hoangbre@gmail.com
Smartphones in 2025 are competing on more than just hardware, and how each phone maker approaches AI is key to the perception of its new phones. Central to many companies’ efforts is Google, and both Gemini and Google Cloud are helping usher in a new era of AI on smartphones.
Several Android phone makers are using Google Cloud features — the same ones that power Gemini features — to build incredible new features powered by AI. Motorola offers a suite of Moto AI features, with some utilizing Google Cloud, while the OnePlus AI Mind and Realme AI Planner features also leverage Google Cloud models.
One of the most interesting uses of Veo on a smartphone is the Honor 400, and its new Image to Video feature. Powered by Google Cloud, it uses the Veo 2 models to generate a 5-second video from a single image. Here’s how it works and why it’s both creepy and incredible.
How Honor AI Image to Video works
A picture is worth a thousand words, but a video is worth a million, especially in its ability to convey emotion and incite a specific feeling. Honor’s new AI Image to Video allows you to turn any static image into a short video, and the results speak for themselves.
The interface and process leave room for improvement, but the results make it worthwhile. To generate a video, it’s as simple as navigating to the Create tab at the bottom of the Photos app and then clicking on generate. It’ll then ask you to select the photo, and it’ll begin generating.
The Snapdragon 8 Gen 3 powering the Honor 400 is a flagship processor, but it’s 18 months old and lacks the powerful NPU of more recent processors. The result is that this feature requires an internet connection, takes around 1-2 minutes to generate, and will occasionally result in an error. Once you look past this and the 10 images per day limit, the feature is both outstanding and creepy.
It’s much better than I thought it would be
I’ve generated ten different videos from images of a variety of scenes, and the results are phenomenal. From artwork in museums to people indoors and outdoors, or even physical objects, the Image to Video feature is somewhat akin to magic. Here are some examples; in each of the videos below, the thumbnail is the original image used to create the video. Interestingly, you can create a horizontal video from a vertical image and vice versa, which is a particularly handy tool to have.
This is an interesting test, as it captures a static image of a car parked amidst traffic in the background, making you feel as though you’re there, especially with the movement of vehicles in the background. Those cars aren’t fully formed, and there are glitches in the generated product, but it’s incredibly lifelike and could easily pass for an actual scene.
This video is particularly interesting as you can see the top part of a white van through the windows of the bus. Honor’s AI feature recognized that and ensured it was in the final video. As a result, the video is very similar to the actual scene that occurred a few moments after this photo was taken.
As you can see, this feature can be used to animate a wide range of subjects. Whether it’s vehicles, street photography, people, or flowers, the Image to Video feature can be fun to play with, especially as it’s a new image or concept each time you generate it.
Generative storytelling with a difference
For every capable creator, hundreds of people have ideas but lack the time, knowledge, or equipment to bring them to life. If this sounds like you, features like Honor’s AI Image to Video are a key step towards achieving what I consider true generative storytelling.
The Image to Video feature could be very creepy, especially when used for nefarious reasons, but I can also see uses for it that truly delight users. Imagine creating a video for a wedding and wanting to include a short clip of someone who’s passed away. Or imagine museums posting short 5-second videos of key exhibits as a way to attract more customers.
Lastly, I can imagine this feature being particularly handy for filmmakers on a budget who have an image and need something more creative or artistic to fill a scene in a video or movie, especially if it’s about a historical subject, where video footage is grainy or unusable. Yes, it’s only 720p so it’s low quality, but it’s usable enough for some scenarios.
Most AI prompts require text input and a vague idea of what you want. The Image to Video feature does the opposite: share a picture and let Honor wow you. It’s a different take on generative AI, but it’s great, and I want this feature on all the best phones.
