Another Unpacked is nearly upon us. On February 25th, Samsung is expected to announce the Galaxy S26 series of flagship phones — potentially including the S26, S26 Plus, and S26 Ultra — with another round of AI-based features. That will come as a surprise to nobody. But as more AI creeps in on Samsung’s AI phones, there’s one important consideration: the slop factor.
I’m talking, of course, about AI slop. Last week, Samsung put out a blog post promoting a new “seamless Galaxy camera experience” on its newest phones. Based on the examples in the article, that experience includes turning a picture of a puppy into cute little stickers, filling in the part of a cupcake you took a bite out of, brightening video in low light to make it more lively, and editing a picture of a cow in a field to make it look like it’s being abducted by aliens.
Harmless, right? You can do most of that with Samsung’s AI editing tools already, and the blog post implies that you’ll be able to make these kinds of edits with natural language “simply by asking in your own words.” But here’s the line that has me worried we’re not just in for an easier path to cute puppy stickers: “Mobile cameras are moving beyond capture,” it says. Let’s — I’m so sorry to say this — unpack that just a bit.
Smartphone cameras have long employed algorithms to squeeze every bit of image quality possible from their tiny sensors. They’ve been doing that for the better part of a decade. Most recently, the AI features have been creeping into the camera app itself — particularly in Google’s latest phones. The Pixel phone’s Add Me feature will help you fake an image in which everyone is posing together, even though it’s really two photos merged into one. And on the Pixel 10 Pro and Pro XL, you can use gen AI to fill in missing details at high digital zoom magnifications. These things are problematic as hell, but they’re ostensibly in service of creating an image that resembles one you saw, or one that could have happened if you had a tripod on hand. But if the cameras themselves are “moving beyond capture,” then that raises the “what is a photo-pocalypse” threat level to code red.
I don’t want to get too caught up by one line in one blog post, but then there are the (partially? totally?) AI-generated ads Samsung also recently posted on its social channels. Two of them seem to show low-light video being brightened, presumably with an AI tool, and one simulates a phone camera zooming in on a distant car to reveal a dog wearing sunglasses inside it. But at least one of those low-light videos seems to have much more AI going on than just some lighting tweaks.
Take that skateboarding video: The onscreen transition between lighter and darker treatments suggests this is a real clip that’s being brightened with AI. So why do parts of it look completely AI-generated? Was the real video merely a starting point for AI to extend the clip? That seems weird!
And what the heck is going on with that glasses-wearing dog? According to the fine print, that image contains “an AI-generated background image with edits.” What edits? Which part is AI? Which part is the camera responsible for? Is there even a dog there? I guess it doesn’t really matter if we’re moving “beyond capture.”
“Beyond capture” could be a very weird destination: one where AI is no longer a tool in service of recording reality or even “memories.” It could very well be Slopsville USA, where you just point your camera at a sunset, press the shutter, and tell it how you’d like to embellish the scene. Or maybe just describe the video you want it to create of your friend on his skateboard and let the camera app do the rest. After all, people like Sam Altman seem to think we’re headed for a future where the line between real and AI content is so blurry that we no longer care about the distinction. So what’s the harm if the slop happens to come from your phone camera and not Sora? And is a camera that has moved “beyond capture” even a camera? I’m not so sure.
It’s possible that Samsung is just adding natural-language photo and video edits to its gallery app and this isn’t an extinction-level “what is a photo” event. We might even see some genuinely cool non-camera uses of AI on the S26, like a privacy screen that can intelligently obscure certain kinds of notifications from onlookers based on viewing angle. I’m all for that! But if not, then I don’t think I’m going to like whatever lies “beyond capture.”
