Following Apple’s much more subdued Image Playground, which will launch this fall, and Samsung’s somewhat unsettling but generally entertaining sketch-to-image function, Google is the fourth phone vendor to unveil AI photo editing features this year. The Pixel 9’s response is a new tool dubbed “Reimagine,” and after a week of use, it’s safe to say that none of us are prepared for what’s to come.
Reimagine is a natural progression of Magic Editor from the previous year, which allowed you to modify the sky to resemble a sunset or pick and remove certain portions of a landscape. Nothing shocking happened.
Reimagine, however, knocks down the entire door rather than just going one step forward. To create something in that space, you can pick any nonhuman object or area of a scene and provide a text prompt. The outcomes are frequently startlingly real. Generally, the perspective, lighting, and shadows are accurate to the original image. Yes, you can embellish with delightful things like rainbows, wildflowers, or anything else. However, that is not the issue.
However, when users tested the limits of Reimagine using their Pixel 9 and 9 Pro, some quite unsettling results were produced. Some of these needed imaginative prodding to get past the obvious boundaries; with careful word choice, you can get it to produce a quite believable figure beneath a sheet covered with blood.
Adding automobile crashes, smoking explosives in public spaces, sheets that seem to cover bleeding bodies, and drug paraphernalia to photographs are a few examples that were shared throughout the trial week. That appears to be problematic. Remind yourself that all of this is integrated into a phone that was only released last week, not some specialist software that users went on to use.
Also Read: Google Imagen 3 AI: Now Available in the U.S. with Enhanced Image Generation
Google’s terms and conditions don’t worry about someone with the worst of intentions, either. The most concerning aspect of all of this is the dearth of reliable resources for locating this type of online content. Our capacity to produce troublesome pictures is far greater than our capacity to recognize them.
There is only a tag in the metadata when you alter an image with Reimagine; there is no watermark or other visible indication that the image was created using artificial intelligence. That’s all well and good, but all it takes is a quick screenshot to remove normal metadata from an image.
Since Pixel Studio photos are entirely synthetic, Google tags them using a more advanced technique called SynthID. However, Magic Editor-edited photos do not receive those tags.
Undoubtedly, manipulating images is not a novel concept. Since the invention of photography, people have been incorporating strange and misleading elements into pictures. The ease with which you can now realistically include these elements in your images is what makes the difference.
Also Read: Google Gemini Live vs GPT-4o AI Assistant: Which is better?
A year or two ago, it would have required a lot of effort, knowledge of Photoshop layers, expensive tools, and time to add a realistic automobile crash to a photograph. These obstacles are no longer there; all that is needed is a new Pixel phone, some time, and a little text.
Furthermore, it’s never been simpler to spread false images. The same device you use to take and share your images with the world also houses the tools you need to edit them convincingly.
Who knows, maybe everyone will use Reimagine to add rainbows and wildflowers to their images and understand and follow Google’s AI standards. That would be very nice! However, it can be wise to approach images you encounter online with a bit more caution, just in case they don’t.
Also Read: How to use and access Gemini Live? Check the Features of Google’s New Voice AI Assistant
