The camera without a lens that takes “photos” thanks to AI
[ad_1]
Paragraphics it is a very special camera. It doesn’t have a sensor or even a lensbut when you press the shutter button is able to generate an image anyway. It does so thanks togenerative artificial intelligence: instead of recording the light data of the scene in front, Paragraphica composes a paragraph of text (slang a prompt) including geolocation and other parameters such as weather and date.
Then he feeds it to Stable Diffusion, which takes care of the rest and recreates a “photo”, interpreting the information provided in its own way.
Artistic operation
Paragraphica is an invention of the Danish artist Bjorn Karmann, who created it to explore “the role of artificial intelligence in a period of tremendous creative tension”. It is obviously not an attempt to subvert the market of “human” photography, but a experiment who wants to investigate the use of artificial intelligence as a tool for the perception of reality. The physical version of Paragraphica is a unique piece, but Karmann has also made one software version of the camera, which can be used directly from the artist’s site (remember to turn on location services before trying it).
The case
Does Generative AI Infringe Copyright? The 2 reasons why Getty sues Stable Diffusion
by Emanuele Capone
How does it work
The physical version of Paragraphica is a minicomputer based on a Raspberry Pi 4 and housed in a body in the shape of a traditional 3D printed camera.
On top of the device, Karmann has included three wheels that allow you to adjust some parameters.
The first sets the radius within which the AI can search for images and references with respect to the transmitted GPS position. The second rule is amount of digital noise to be inserted in the picture, while the third rule the precision with which the algorithm will follow the indications provided in the prompt: for the artist this last setting is a digital metaphor for focusing in a traditional camera. The display on the back of the camera finally shows first the generated text prompt and then the image obtained with Stable Diffusion.
With the eyes of AI
“As AI-based speech models become more and more conscious,” writes the artist, pandering to the controversial school of thought that chatGPT and co. would show glimpses of humanity, “we will have a very limited idea of how they see the world. [Paragraphica] it allows us to perceive the world around us in a way that is not limited to visual perception alone. Thanks to localization and visual synthesis algorithms, it offers a deeper look into the essence of a world through the experience of other intelligences”.
Artificial intelligence
Who created the photo of the Pope with the duvet?
by Pier Luigi Pisa
The results, some examples of which can be seen in the “photos” of the article together with the prompt that generated them, are as interesting as devoid of any “photographic” value.
Which is not, moreover, the goal of a conceptual artwork like that of Karmann. The images created by Paragraphica are therefore not photos, but a surrogate for all the human photographic artifacts that can be traced back to a specific place, and for which the generative algorithm has found a reference in the data on which it has been trained.
Tech Test
Nikon Z8 and the meaning of professional cameras in the Midjourney era
by Andrea Nepori
derived simulacra
In exploring the concept of “consciousness” of the machine, Karmann has done nothing but outline its limits, showing that any production by the machine can only be a simulacrum of a cumulative human experience. And if it is true that in many cases, for some types of text, these simulacra are difficult to distinguish from human production, photography remains the litmus test of what the machine cannot yet replace in any way: the role of the physical presence in the process of perception of reality.
However already the technical image, understood as a product of reproduction tools such as cameras, it is a simulacrum. And the fact that it represents reality is basically just a convention. It is precisely for this reason that Karmann’s work is interesting: because he proposes himself as a derivative of a function of reality, exploring the ethereal and still undefined world of simulacra of simulacra produced by generative artificial intelligences.
Artificial intelligence
AI-generated music
by Ernesto Assante
“More generally, it is not in its role as a reflection, as a mirror, as a counterpart of the real, as a representative form that the image is interesting”, wrote Jean Beaudrillard about photographic and cinematographic images, with a definition that fits even better the images created today with artificial intelligence. “But that’s when it begins to contaminate reality and to model it when it no longer conforms to reality except to deform it better, when it thins reality to its advantage, when it anticipates it in such a way that it no longer has time to produce itself as such.
Artificial intelligence
The AI had a problem with its hands. Now that he’s fixed it, we’re in trouble
by Pier Luigi Pisa
[ad_2]
Source link
