A new statue is on display at the Museum of Science and Technology in Stockholm. From afar, it looks barely noticeable, a human-shaped stainless steel whirlwind clutching an orb. It almost looks like an oversized rewards trophy. It is only after investigation that the singularity of the five-foot work becomes apparent: it was produced by a fusion of Generative AI and precision manufacturing.
It’s called The Impossible Statue, a title that, given the underlying concept and laborious design process, hardly seems hyperbolic. The work came about through a collaboration between The AI Framework, a consulting firm, and sandvikan engineering company specializing in metal cutting, which pushed the project as a way to display its computer-programmed manufacturing.
Although this project is centered around art, the manufacturing methods involved in creating The Impossible Statue differed little from other tasks, said Nadine Crauwels, President of Sandvik Machining Solutions: “By using all of our capabilities, we can significantly improve manufacturing efficiency, reduce waste and ensure the highest quality.
The Impossible Statue (2023). Photo courtesy of Swedish National Museum of Science and Technology.
First, they trained AI models on the work of five sculptors weaving together the counter post of Michelangelo, the musculature of Rodin, the naturalism of Käthe Kollwitz, the movement of Takamura Kotaro (a great disciple of Rodin) and the audacity of the figures of Augusta Savage. The Sandvik team essentially selected the most desirable attributes of some of the most famous sculptors of the last five centuries, then repeatedly generated images using a combination of Steady broadcastDALL-E and Midjourney until they are satisfied.
The result is an androgynous figure composed of moving shades of steel that holds a globe, one that counteracts the laws of gravity, and whose lower half is entangled in a corrugated sheet of metal. Are the invisible hands of the aforementioned sculptors obvious? Somewhat. The musculature is certainly bold and the fabric looks like a contemporary reworking of a Renaissance binding. But again, the darker social commentaries advanced by the work of Kollwitz And Savage seems absent in a statue full of promise.
Sandvik worker checking the statue for accuracy. Photo courtesy of Sandvik.
After choosing a design, Sandvik converted the 2D image into a 3D model using depth estimation software and human pose estimation, a computer task that identifies different parts of the body human in a scene. The company then thoroughly tested the manufacturing process in a series of digital simulations, halving the amount of steel used in the process. The statue he produced was designed was comprised of nine million polygons, 17 separate pieces of steel, and differed from the digital design by less than 0.03mm.
“The Impossible Statue is a great example of what the combination of modern technology and human intelligence can create,” Peter Skogh, Director of Science and Technology Museum, said in a statement. “Our mission is to create a broader understanding of the possibilities of technology and to inspire the next generation. This project ticks all the boxes for us.
Follow Artnet News on Facebook:
Want to stay one step ahead of the art world? Subscribe to our newsletter to receive breaking news, revealing interviews and incisive reviews that move the conversation forward.
