Explaining the art behind Forbes’ AI 50 design

By Nick Sheeran

Wa hen in charge of creating art for the fourth year Forbes AI 50 list, it immediately occurred to me that we should use artificial intelligence to generate our score. AI is making significant progress not only in commercial applications but also in the visual arts. His artistic endeavors are being auctioned at Sotheby’s, creating NFT collections and accelerating traditional media production processes. What does this mean for the future of art and its reception in the general public?

In most cases for artificially intelligent visual work, such as the main art for AI 50, General Adversarial Network, or GAN, one “trains” on a large set of data and compares individual relationships between each instance of data to understand what belongs and what belongs. does not seem. GAN is slowly learning to filter noise in this data, discovering similarities and finally gaining the ability to recreate the received material or determine whether a new input suits it or not. It’s the same machine learning process that goes into automating credit approval or public health diagnostics, and the main differences are the end goal and the data set used. Once trained, GAN knowledge is contained in a vector matrix, called “latent space”. The art you watch consists of 4 separate “latent space walks,” rehearsed using two ready-made datasets provided by Runway ML, and two hand-crafted Forbes staff. We “walked” through the matrix, and each step resulted in a video frame.

I find this process enjoyable because each individual GAN ​​output makes it less interesting than their array. My goal is to engage the audience in art in a dynamic format that relies on time and systematic relationships beyond the static perfection of a framed painting or sculpture. Instead of objects made to create an aesthetic experience and retain value, I like to think of art as the current result of artistic practices, living, breathing, often focused not only on form but also on exploring the world. It’s more than giving the viewer a moment to feel something, even though it’s great. It is also intended to stimulate thinking, influence thinking, and ultimately influence change. This activity is dynamic, deliberately vague, poorly defined, loose, and calls for random twists and turns. It creates turbulence and criticizes optimistically, inviting you to do the same.

There is one more thing to understand in the turbulence of this dynamic: the so-called author of a work of art of artificial intelligence is no longer a lone creator. AI, plus the sources of its training material, is its (often unpredictable) partner. Designers and artists have long debated the idea of ​​program co-creation, see Sol Lewitt Wall drawings or Conditional Design Manifesto. But the impending widespread utility of artificial intelligence will bring this spirit of collaboration further into the mainstream. Everyone and not one can be a creator, and that’s great! It comes at a time that is historically crucial for other reasons that carry the need for a spirit of cooperation, such as climate change and the reshaping of peaceful globalism facing a new Cold War. The idea of ​​domination must disappear: it is time to go wild again, reunite and return to the traditions of common and ecological reciprocity that we have lost sight of. Although artificial intelligence is high-tech and can certainly be used for evil, it has the potential to revive organic relationships that are key to a sustainable future. By no means do I want to convey that I recommend for the mother to be inactive.

Let’s take a moment to explore the idea of ​​sophistication in artificial intelligence because it refers to kitsch and ingenuity. Researchers measure the accuracy of the GAN using a metric called the Fréchet initial distance or FID, which basically quantifies the accuracy of the GAN output relative to the data it is trained with. If you want to make a GAN that generates, say, leaf blowers, the lower the FID, the more realistic the leaf blower. If we tried to do that in art, we would immediately come up with kitsch; it is just a reckless repetition of something previous. It’s the difficulty of creating something meaningful with artificial intelligence – if it’s too precise, it’s pointless, and if it’s too ambiguous, it’s pointless again (and yes, swirling images that make buildings look like dogs, Van Goghs or cheese balls almost always fall off to the first extreme of this spectrum of nonsense).

Click here for full AI 50 list coverage for 2022.

Not to mention, it’s pretty hard to find 5,000 images of something, and even harder to curate those images in a way that doesn’t betray restrictions and fundamental discrimination, the apparatus for creating images of humanity. For example, for one component of the illustration we used a series of images capturing well-designed industrial products and I wonder if this will only result in a kitschy shepherd that reinforces the heuristics of the field, as opposed to researching the universal structures that govern them. , and deeply wrapping the world in an instant. It is crucial that, since the public accepts GANs, we avoid going astray. Look around you, what is the meaning built into the built environment? How do you communicate with the cumulative expression of society?

Ultimately, we will forever continue to merge disciplines into others and discover the intersectionality of our existence. Art will become a tool, and tools will become art, and hopefully, at some point soon, decentralization will shift the responsibility of creation and design from a few technocratic to many cooperative. When that happens, it will be more important to us than ever to define what is essential to sustaining our lives. I would like to think that they will be filled with individuality, curiosity, responsibility and optimism.

MORE THAN AI 50 2022

MORE THAN FORBESAI 50 2022: the largest AI companies in North America shaping the future
MORE THAN FORBESEmoji worth $ 2 billion: Hugging Face wants to be the starting point for the machine learning revolution
MORE THAN FORBESAI Upstart Waabi is adding self-driving veterans to the race to commercialize robotic trucks
MORE THAN FORBESMashgin estimates the value of $ 1.5 billion with a self-checkout system with AI

Leave a Comment