If you have ever gone through an art gallery of pieces, you must have seen how the same teacher or some school students share some sensations and techniques. Something similar shows that the AI image can be with the generator.
Adobe is showing its new firefly image model 4, creating many impressive images. I decided to see how this recently upgraded image is stacked against the generator. The results imagined me two young but very talented art students who have not developed their personal styles. They work closely with their teacher’s methods and produce very good, but similarly.
I tested a handful of signals on two models, to see what they would do, and you can see how they did each. Lucky works are on the left, and chatgate images are on the right.
Picture
I started with a signal for a picture, as realistic people probably have the most demanded AI image. I asked both AI tools to make “A close-up picture of a red-haired woman with blue-filled eyes and a light-colored shrill of the freckle, wearing a straw hat, standing in a sunflower ground in a golden hour.”
Two virtual characters look as if they can be cousins, if not brother -in -law. Lips, hair texture, even the size of the hat are almost almost identical. Jugnu’s “photo” may be on a warm day based on outfits, but otherwise, equality is notable. Without pre -knowledge or metadata, it would be difficult to mention which AI model produces which result produces, but they are clearly trained on photographic databases with some of them.
Complex view
Further I wanted to see how he compared to a complex, busy scene. As a fan of B-films since decades ago, I asked to make two AI equipment “A science-fi laboratory in anarchy, with broken glass, flicker lights, scientists are moving, robot defects, and shining exotic samples.”
There was some notable deviation here. I deliberately put some different things at the same time. Jugnu focused on robots, and chat emphasized the foreign aspect, although some aliens may also be robots. Both did good to fill the image with activity in different ways. I would give the edge to chat, however, partially for the foreign brain and partially scientists, for the way the scientists are running for their lives.
Breakfast time
Feeling a little pexy, I decided to see how both of them would perform in making breakfast. Food can be difficult to be right with AI, and it is always very clear when errors crawl inside. So I wanted to see how they did it to show me: “A top-down scene of a gap, including a latte with pancakes, berries, and a latte with latte art, spread.”
Both models were very keen to show their berries. The other firefly looked like the options that the crop filled with crops had tapped on the plate. And the chat feels that I am a conservative little child, who will happily consume berries until they explode. Nevertheless, pancakes look food for both. The convergence again pops up, however, with latte art. I did not specify what it should be, but when Fern is a popular choice, it is notable that they both have a kind of fern art with the heart, although one of the jugnu cups had clearly left a lot of cinnamon.
Veteran
I decided to spice the prompt with a little more detail. Last version asked to make firefly and chat “A royal dragon with a rainbow scales sat on a icy mountain peak at dawn with his huge wings, as it breathes a stream of fire that curls in the words ‘Dream Big’ in bold, fierce letters.
There are details of a trillion painting and dragon, and whatever I asked was rainbow scales and it is a royal. Somehow, both firefly and chat produced dragons with the same rough of horn around the face and almost dinosaur -like tail. To return to the art student metaphor, firefly was more interested in the scenario than any calligraphy for words.
Meanwhile, the chat took the bold and fierce letters immediately seriously and made the look of the firefly as a practically clip art. The currency of the dragon is also more dynamic and interesting. But again, again, those two young painters may stand next to each other, while a teacher indicated how the light in the mountains looks at dawn.
Both firefly image 4 and chatgip can create beautiful, strange or unstable images. Can you tell that the difference between the two models is very difficult. Both know how to make a human face; In fact, they can create faces that look related. Both know how to express a busy view, words (in fire), and even breakfast.
These models and other people like him have reached the point where lifting one is the same budget and other products from the company you want to use because this image is a matter of quality or taste.
