Everything with form is unreal 凡所有相皆是虚妄
AI generated Image, 2022
The AI Mountain project is a collaboration between Nvidia landscape generator and me. Through the generating process, I draw basic diagrams of the "two mountains" then choose the material types of them. After that, the generator will create different landscape images based on the diagrams and materials. I test not only reasonable combination like both triangles are mountains which eventually makes reasonable images but also unreasonable one like the upper triangle is sky but lower one is fog which create something visually real but logically unreal.
AI Mountain 项目是 Nvidia 景观生成器和我之间的合作。通过生成过程，我画出了“两座山”的基本形态，并选择了它们的材质类型。之后 Gaugan2 生成器将根据形式和材料创建不同的风景图像。我在两个三角形以及地面上测试了几种材料的组合，例如山脉、岩石、森林、海洋和雾。有合理的组合，就像两个三角形材料都是山，最终形成合理的图像。还有一些不合理的测试，比如上三角是天空，下三角是雾，创造了一些视觉上真实但逻辑上不真实的东西。
At the presentation level, we can observe the following points through juxtaposition. Specific construction of meaning is constructed by specific composition, which transcends the limitations of time and space and becomes a generalized image language. At the same time, an image as a "type" is created. On top of the effective composition, the meaning can be switched only by changing the material.
The world that human interpreted and experienced could be massively collaged both in forms and materials. There will not be a real "thing" behind the photos. Everything is from reality but being fake at the same time, indicating somewhere but nowhere, something but nothing. A hyperrealism is established through image production. Meanwhile, to some extent, the AI creates something unique. Even if this kind of uniqueness is based on the big data from the real world. Who produces the uniqueness of the work? Algorithms and humans create double randomness, but this randomness is then incorporated into the framework and rules of the algorithm.
Are we still keen to get "there" if the production of landscape images are completely turn into digital? Will there be nostalgia for the virtual landscape?
The prototype of AI Mountain may come from my memory of the mountains and rivers in my hometown. It may also come from my impressions when I travel around the world, the wilderness of Scotland, the broken stones in Atlas, the endless snow of Arctic. I see their shadows in the randomness of AI as they awake a certain part of my memory. What is the difference between all this wonder and the generative mechanism of AI, that everything in the world existed in a certain way before my eyes, and returned as a shattered memory after I left? Everything is connected, and the sights recorded by a sensor will eventually become the threads of my memory, and in this sense, it becomes a shared nostalgia.
Though there is still a gap between "human want AI to show something" and "AI want to show human something", it seems the gap is becoming more and more blurry in the near future. Now we have to understanding the meaning that AI offering to us.
A new mirage is emerging on the horizon.
虽然“人类想让 AI 展示一些东西”和“AI 想展示给人类一些东西”之间还有差距，但在不久的将来，这种差距似乎会变得越来越模糊。现在我们必须了解人工智能为我们提供的意义。