Obviously, 2023 will be the year of artificial intelligence. In recent months, we’ve seen an almost immeasurable amount of innovation – just think of OpenAI’s ChatGPT or Bing AI, but in addition to these, other AIs are particularly popular.
One of these is Midjourney, which can generate any style, even photorealistic text-based images. While we’re at it, we shouldn’t forget Adobe, which a few weeks ago introduced an AI called Firefly, which pretty much knows the exact same things as Midjourney or DALL-E 2 – more precisely, yet. The software development company also responsible for Photoshop has turned Firefly into the world’s most popular photo-editing application, which is now capable of amazing things – previously it could complete work that took hours with just a few clicks.
He also invents what does not exist
One of Photoshop’s new features is the ability to highlight or “select” an area of the image, click the new “compose” icon, type in a few keywords describing what we want to see where our selection is, and the selection will magically appear. Shows – whether it is about anything.
The program’s new AI-based generative fill trick gives you a way to remove objects from images without creating weird errors in the result. It’s true that this was indeed possible before, but in most cases the end result was always something the user had to correct afterwards, but thanks to Firefly, the app works with almost frightening precision.
The most interesting innovation is obviously the function that allows us to change the size of the canvas and, when pressing a button, Photoshop fills the empty area with details as if it were always part of the original image.
You can see countless videos about innovation on social media – they’ve literally taken over TikTok, Instagram, and Twitter. Below, for example, you can see a compilation in which well-known memes have been supplemented with realistic detail by Adobe’s artificial intelligence that wasn’t actually there when the photo was taken. The white frame indicates the original image, and the area outside it has already been created by Firefly for the image.
It is currently in the testing phase
The previously listed novelties are only available in the latest beta version of Photoshop, i.e. intended for testers, so it still has a lot of bugs. According to reports, for example, Firefly is still struggling with human hands and facial features, just like Midjourney or DALL-E 2. Adobe has yet to announce when it will be available for high-end devices, nor if it plans deeper integration with AI into their other apps, Like Illustrator or Premier – but the answer is obvious to almost everyone.
Of course, there is another question about what kind of problems “fake” photos and videos generated by artificial intelligence will cause in the future, but it is already clear that its “power” should definitely be limited. For the sake of interest, in our previous article we also asked AI what new mental illnesses it thinks humanity should face due to the proliferation of AI systems – and OpenAI’s Chat GPT gave an honest answer. You can read our article about this by clicking here.
