the DO MACHINES DREAM
OF NFT?
DO MACHINES DREAM OF NFT?

A CRYPTO ART COLLECTION ON TEZOS

機器會夢見NFT嗎

At the onset of GAN, the technology of Non-Fungible Token (NFT) was not born yet. Nowadays, along with the popularization of blockchain and cryptocurrency, there is an increasing number of artists creating and turning works into NFTs with GANs. Most of the exhibits in this exhibition are from Hic et Nunc (now TEIA), a Tezos NFT platform. If machines are approximating humans in the process of learning, will they also discover that their works are on sale in the form of NFTs, and have some thoughts to ponder over at midnight?

Do machines dream of NFT? This question is left for everyone stepping into the exhibition, as well as the ever-smarter mobile phones, computers, and electronic devices around us, to answer.

Forum Ticket

Encounter (Plasma Red: Passport)

Encounter (Plasma Red: Passport)

Encounter (Plasma Blue: Day 1 Ticket)

Encounter (Plasma Blue: Day 1 Ticket)

Encounter (Plasma Green: Day 2 Ticket)

Encounter (Plasma Green: Day 2 Ticket)

Unveiling

Since Ian Goodfellow proposed the Generative Adversarial Network (GAN) in 2014, this technology has attracted a lot of attention in fields like game design, astronomical research and advertising image, and visual art is no exception.

A GAN is a Deep Learning in Machine Learning, which is an essential topic of Artificial Intelligence that focuses on the neural network simulation to identify, produce, and improve information. Of the lengthy term “Generative Adversarial Network,” “Adversarial” is the core feature: setting up the duo models of “generator” and “discriminator” for both to fight and grow with each other.

First of all, the generator acquires materials from the latent space and creates seemingly genuine products. Secondly, the discriminator identifies these products as true or false on the basis of objects in reality or targets in the database. If a product is identified as true, the product passes the test, and the information is verified. Otherwise, the generator will generate information again until they cheat the discriminator. The relation between the generator and discriminator is like the one between counterfeit money makers and criminal affairs bureau or that between art forgers and art forensic investigators. Even if we do not treat it as such, their interaction can be understood as that between artists and critics, which improves information or artworks in various aspects.

Applying this logic and technology mentioned above to visual art, viewers can not only see something or someone that does not exist, but also appreciate the repeated process of production and correction. By distinguishing true from false, grasping the variable and the constant, and imaging the links among all and the possibilities in appearance change, we play with “ways of viewing” yet again. Besides, if machines can paint, what do they “think of” and “see”? This question may be exaggerated. After all, aren’t all these products creation of humans? These controversies may be chances to the reinterpretation of human-machine relationship.

At the onset of GAN, the technology of Non-Fungible Token (NFT) was not born yet. Nowadays, along with the popularization of blockchain and cryptocurrency, there is an increasing number of artists creating and turning works into NFTs with GANs. Most of the exhibits in this exhibition are from Hic et Nunc (now TEIA), a Tezos NFT platform. If machines are approximating humans in the process of learning, will they also discover that their works are on sale in the form of NFTs, and have some thoughts to ponder over at midnight?

Artist