Meteoroscape photography expressed the “Diffusion of AI in society”.
“The Horse in Motion” is a series of photographs of a galloping horse in motion, shot from the side.
The photographer was Englishman Eadweard Muybridge, who was visiting the US in June 1878 to further his career as a photographer. There, Leland Stanford, an industrialist and former governor of California, asked him to photograph his mare, Sallie Gardner, with multiple cameras.
Somewhere I heard a legend that prior to that experiment, a bet was made as to whether all four of the horse’s legs would be in the air when running at full speed, or whether one of them would be left on the ground.
This “work,” which was thought to be a scientific experiment, quickly became a hot topic of conversation at the time and, interestingly, influenced that Thomas Edison to create the kinetoscope, which is said to have been one of the origins of the current film industry.
Later that, Hollywood was formed, people came to love movies, Marilyn Monroe became an idol, and animated films were created. Isn’t this exactly what “Photography as a Contemporary Art” is all about?
It is no exaggeration to say that “愁篇ノスタルジア Exhibition” was given birth from it, too.
So, this time, based on “The Horse in Motion”, we redesign the work that is employed “defusion model algorithm” of drawing AI but is made up by human power. The plan is to create a mysterious work that is a reproduction of the original, but which is also the original. Luckily, “The Horse in Motion” consists of 12 cards, so we’re surely to see a horse in the end.
I say “surely” because, as of January 1, it hasn’t been finished to make at all. The exhibition will be open from March 9.
Our battle is about to begin! Please look forward to Itsuki Kujo’s next work!!!
So, after spending well over 200 hours, I finally generated one series of works. Basically, these pieces of photos are just manually arranged, so it cannot be done in a few dozen seconds with a single button press.
Now, this work may be another typical example of ‘Art’ that is difficult to understand at first glance.
To begin with the overall composition, it is a series of 16 cards, consisting of 4 ‘noise’ cards followed by 12 cards based on “The Horse in Motion” mentioned above. Only the last card is a ‘drawing’, which is made with Stable Diffusion, based on the 12th card of “The Horse…”. The other cards are ‘photographs’ assembled by Kujo himself over a long period of time.
The method is the same as the prototype “人為的アルゴリズムで…”. the same format is used to prepare a set of image data sets, and then gradually generate new images by layering them on top of each other. The difference is that this time, the number of images in the set is 256, and the size of each frame is not reduced for the sake of printing in real life.
Next, focusing on one of the layer cards, it is clear that this technique treats a single photograph as a dot and creates a larger image. But, the technology itself is already well established, not to mention the highest-paid work at NFT, and there is even software that can create the image automatically, thanks to its frequent use in advertisements. Therefore, if you pick up a sheet of the series of “The Algorithm in Motion”, it is almost the same as the one in the above.
Therefore, what makes this work different is that it adds the idea of creating an image from the data set with different brightness and transparency (specifically, this time, a total of 16 sets with 4 types of brightness and 4 types of transparency were prepared) and layering the new layers in turn, influenced by the old layers. It could be said that this is similar to oil painting or watercolor painting.
Thanks to this method, the algorithm, we are able to generate a new ‘original’ that is not bound by the ‘reproduction’ of “The Horse in Motion” that was used as the base. On the other hand, it is a double-edged sword, because if it is not done properly, a horse running with more than four legs will be generated, which is a horror.
I hope you understand….
By the way, can we not equate this algorithm with our daily experience and learning? We improve and grow every day by doing almost the same thing over and over again, although we don’t grow just by doing exactly the same thing.
Sometimes we are greatly influenced by the concentration of experiences, and sometimes our days are thin and unchanging. When the repetition reaches a certain point, internal changes somehow occur suddenly.
In carefully arranging the 256 image data sets this time, I wondered how to generate a more ‘sophisticated image’ from the ‘noise’. With the added condition of not reducing the size of each frame, it seemed impossible, and in fact, I gave up several times along the way. However, when I reached the 13th image, I finally realized this;
I could generate a more ‘sophisticated image’ by limiting the arrangement of the images to a grid!
Isn’t this so-called “limitation of possibility space,” which starts from the point where the data set is limited to 256 images? This is surely one of the driving forces behind our ‘growth’, a kind of algorithm that is equally available to all beings in the universe, I wonder.
|Series||疾走するアルゴリズム – The Algorithm in Motion|
There are no comments