Tyler Hobbs on Generative Art and NFTs: ‘I see code becoming a standard tool for artists. It’s just way too powerful to ignore’

Tyler Hobbs is a pioneer of the Generative Art movement, producing work using autonomous systems — specifically, computer algorithms.

In June 2021, employing the most complicated algorithm he had ever created, Hobbs made Fidenza, a series of 999 computer-generated images minted as individual NFTs that examine the point where the worlds of digital and analogue — and chaos and order — collide.

The project is based around the concept of ‘flow fields’: images that map the density and velocity of a fluid over space and time. Hobbs gave his computer ‘authority’ to randomly choose and apply layers of digital code to these images, altering elements such as colour, scale and turbulence, and selecting the work’s final appearance. And because he programmed some pieces of code to apply themselves less often than others, each work’s ultimate combination of traits can be ranked in terms of rarity.

Fidenza  quickly became one of the most sought-after NFT drops of 2021. One work alone rocketed in value from 0.58 ETH (then worth around $1,400) to 1,000 ETH (more than $3.3 million at the time) over the course of just two months. More importantly though, Hobbs secured his place as a leading figure of the digital art movement, as well as an artist able to survive the NFT market’s peaks and troughs relatively unscathed.

On 28 February, Christie’s is offering the rare Generative artwork Fidenza #724  as part of the 20th/21st Century: London Evening Sale. Ahead of the auction, Hobbs sat down to explain how the project came to life, and where he thinks the future of art and technology lies.

Where did the idea for Fidenza  come from, and what does the title refer to?

‘Many of the algorithmic and visual elements in Fidenza  can be traced back through my work over the past five years. However, it was also a radically new experiment — designing an algorithm suitable for 999 outputs was a massive increase from my previous largest, which only had 32.

‘The title Fidenza [the name of a town in northern Italy] was specifically chosen not to be significant. I wanted the viewer to interpret the work without preconceptions. I borrowed the strategy of using a town’s name from Franz Kline.’

What was involved in writing the algorithm for Fidenza?

‘The algorithm itself is all hand-coded text and numbers. It’s written in a programming language called JavaScript, which is extremely common, meaning it will be supported for at least several decades, if not longer.

‘It’s hard to mark the start of Fidenza  because it evolved out of algorithms I’ve developed over the years. However, there was a period of roughly three-and-a-half months where I focused on it intensely.

‘I began with an experimentation and exploration phase, where I changed different elements to introduce new concepts and behaviours. Next, I focused on removing buggy areas and making weak outputs less likely. The last step was to adjust the probabilities in order to make the entire output varied and balanced. My goal was to ensure that you could look through all 999 outputs and not get bored.’

Are the parallels to work by artists such as Piet Mondrian and Bridget Riley built into the code?

‘I wasn’t relying on any particular art-historical references. However, I do spend a considerable amount of time studying the work of other artists — especially those with a systematic approach to their work, like Mondrian, Riley, Agnes Martin and Kandinsky in his later period. They have been my biggest influences, and I often notice connections between their work and my own, even if I didn’t introduce them consciously.’

Can you explain how and why you sometimes make analogue counterparts to your digital works?

‘There are many different ways to translate a digital work into the physical realm, and I decide which to use based on what suits the art best. Sometimes I use a projector to make hand-trace drawings, or employ laser-cut stencils.

‘For Fidenza  I created one painting on paper, which is based on output #163. I began by using a plotter — a simple, programmable robot holding a pencil — to create a drawing of it, which I then painted over in gouache. It’s hard to get a robot to paint well: there’s a lot of “touch” that is still extremely difficult to get right.

‘Seeing digital concepts in analogue form can make them more accessible and human. It also prompts us to consider that while we have physical bodies, we spend a lot of time in digital worlds, and in order to explore this reality, art needs to span both.’

Once Fidenza  had been made, did you curate it in any way?

‘There are several ways that I can curate a project’s output, and each one has interesting trade-offs. If the desired output set is small, I will typically curate that myself, usually filtering my favourites. With some other projects, I put the curation in the hands of the community. For Fidenza, however, the 999 images were generated at random with no curation whatsoever.’

Can a computer ever compete with human creativity?

‘I don’t think we have a good definition of “creativity” right now. Like every other form of artistic practice, the act of creation using artificial intelligence and other Generative approaches has particular strengths and capabilities, as well as particular weaknesses and limitations.

‘It’s up to the artist and viewer to determine what’s meaningful and how we can best relate to these new technologies — which are certainly going to dominate our lives in the coming decades.’

What do you predict for the future of art and technology?

‘I see code becoming more and more of a standard tool for artists. It’s just way too powerful to ignore, and those who can wield it effectively can do things others can’t. It also doesn’t have to be all-or-nothing; I think a mixed use of code and traditional practices will create the most interesting outcomes.’