Samsung’s new AI symbol-making software is a little too cool

Dispatching pirates in Elliott Bay was something, but it was a minor misty bee that sent me over the brink.

Samsung wants us (and its shareholders) to understand that its new phones are the AI-est phones ever to be AI-ed, and the Crease 6 I’ve been trying out comes with a new software called “Sketch to is called “image”. , Draw a rough caricature of a photo or a blank note page, and generic AI will be required to transform that thing into a photo. When Samsung introduced it on stage at Unpacked I ignored it as another AI factor – but y’all, it really is excellent. So excellent that it worries me a little.

Using the caricature to symbolize software in a note is very intuitive: you draw one thing, highlight it, and use a handful of tools like “3D cartoon” and “Illustration” to transform your doodle into something else. Let’s select one of the styles. Colourful. Your symbol is delivered to the cloud, and then within a few moments, you’ll see some options to choose from. In most cases the results are adorable and hilarious; I took the request from my two year old and we pulled out weird looking unloaded vehicles and college buses. Sometimes you get a teddy bear from a lot of hands, but nothing serious.

Using a caricature to depict a photo is where things get weird. I’m the worst artist in this area, and this software turned my extremely simple sketches into photorealistic pictures. The AI-generated parts are believably incorporated into the images – scaled and combined with the environment in some way that allows them to be sterilized enough to be identified as fakes.

In this way I reached a state like a bee. I took a photo from a dock just south of downtown Seattle with some plants in the foreground. Because they are so close to the camera and my focus was on the distance, they are quite blurry. I created the world’s worst bee cartoon on a sow, thinking the AI ​​would insert an in-focus symbol of a bee – giving it away as just a gimmick. out of place!

The AI ​​bee is blurry, similar to the flower it is landing on. If I didn’t know the story of AI B’s foundation, I might refuse to think carefully about it when scrolling past that symbol on Instagram. I think the photographer took the picture at just the right moment or was waiting for a bee to fly into the frame – things that influence ability and patience. Now it is not so. In fact, now I’m not even sure I’ll be able to see the “AI-generated content” watermark in the corner of the picture.

I’ve toyed with caricatures to draw a dozen in the last year, and the effects aren’t always “blurry bee” excellent. Often, they will have obvious indicators of generic AI artwork – words written in foreign-looking language or unusual textures that don’t look quite right. Looks convincing, but if you look beyond the 2D, you’ll realize something is wrong.

Sometimes the material presents itself – I don’t think anyone believed I saw a giant pirate ship anchored in Elliott Bay or a giant orange cat at an intersection in West Seattle. However, even if the photos are so strange that no one would mistake them for authenticity, they still give the impression of looking real.

Generally, important issues will seem obviously fake. But it’s really easy to add another motor vehicle to a photo of a busy road or a sailboat in the distance, and most people may be none the wiser. Apart from that AI watermark – which has been cropped out nicely – there’s really no way to discern that there’s anything else strange regarding the symbol. This is strange!

I don’t want to waste it. The use of caricatures for symbolism is not entirely mandatory, and many people will not even be able to find it in the Gallery app. Bee out of focus not able to undo our family stuff. Although I think we’re in a more bizarre park with AI. Surely, you have long been ready to add an out-of-focus bee to a photo in Photoshop. However, by keeping this ability within The same software you use to display and distribute the image Are there any other factors? The functionality and reach of generative AI tools goes beyond our shared understanding of what might be real and what might be fake when you scroll through Instagram.

I think I think about this trait most strangely when I’m showing it to my baby. He’ll grow up understanding that, with the click of a button, you can turn a rough caricature into something more charming. Or with a little effort, you can enhance the photo of some train tracks by adding a train. Is this a good thing or a malicious thing? I don’t know, but I certainly sense a dissonance between how I saw the creative opening as a child and how he would see it.

None of this has stopped me from having dozens of laughs with everything from caricatures to logos. There’s a curiosity about the output of generative AI that is more or less hilarious – like when I tried to add a green monster with its head sticking out from Puget Pitch and it turned my drawing into a giant green polar bear , in which the position of the muscles was undulating. Edge. Or when it became a caricature of a stick figure, which turned into a life-size stick figure, with a silhouette in the garden beneath it.

Is the definition of images changing before our eyes? Is our speculation beyond the reality in the pictures working out an incredibly uncertain future for our self-governance? Yes, but additionally, I took a picture of a rabbit and the AI ​​let me put a simple management hat on its tiny head. What a future for survival.

Comic Strip for Image is available on Galaxy Z Fold 6 and Z Fold 6. Samsung hasn’t said publicly whether it will make the feature available to alternative Galaxy phones, but judging by Monitor reports of the company aggressively expanding Galaxy AI into previous-generation models, I think. That this is highly likely to happen. Samsung has also committed to bringing AI features to 200 million phones this year alone. If the slight blurry bee is any more or less indicator, I’d say things get a little weird when that happens.