- With generative art there's usually a tension between variety and beauty, but Fidenza achieves an extraordinarily high level of both.
- I like the Artblocks model in general. Pieces are minted on-chain. It means the artist can't curate the output, so the algorithm itself is the art. Artists have been making generative art for decades, but it feels like this is the way it was always meant to be done.
- Artists can make certain outcomes of their generative algorithms more rare than others. The decision of what should be common or rare can be made for commercial reasons (kind of like hologram trading cards), or artistic reasons. After studying all the ways Fidenza runs, it seems like Tyler Hobbs has hewed really closely to the artistic ideal.
Great work by Tyler, love his dedication to not just the work itself but also documenting the process, the technology and last but not least reflecting on the "art" of it: https://tylerxhobbs.com/essays
This was a really popular project. For those who haven’t poked around generative art recently, check out artblocks.io. You buy a piece of art (limited edition) and it creates a unique (randomized) piece for you on the spot. Pretty rad.
So how does this work from a practical level. Is the source code open by default because it works on chain or can you do things like this closed source?
Also, are there good resources to learn generative art?
The source code is visible on-chain - it can be queried from the smart contract, but also it has been etched into the data of the “script upload” transaction. This is how ArtBlocks and other projects can fetch the script and display it. But that doesn’t mean the code is open source! Just like scripts on a website (which are also easy to see), the author decides on how to license the work.
I often suggest GenerativeArtistry as a simple starting point if you already know a bit of JavaScript and want to understand how it works on a basic level[1]. Tyler Hobbs' blog (OP link) also has some great essays on generative art[2], as with the blog of Anders Hoff[3], and also lots of other resources in awesome-creative-coding repo[4].
- With generative art there's usually a tension between variety and beauty, but Fidenza achieves an extraordinarily high level of both.
- I like the Artblocks model in general. Pieces are minted on-chain. It means the artist can't curate the output, so the algorithm itself is the art. Artists have been making generative art for decades, but it feels like this is the way it was always meant to be done.
- Artists can make certain outcomes of their generative algorithms more rare than others. The decision of what should be common or rare can be made for commercial reasons (kind of like hologram trading cards), or artistic reasons. After studying all the ways Fidenza runs, it seems like Tyler Hobbs has hewed really closely to the artistic ideal.
That's also a nice write up on Tyler's process. For those interested he mainly uses Closure:
https://tylerxhobbs.com/process
Unfortunate name collisions in this space. (The concept closure, Google Closure, Clojure.) That’s Clojure he’s using.
(I don’t work for artblocks)
Also, are there good resources to learn generative art?
I often suggest GenerativeArtistry as a simple starting point if you already know a bit of JavaScript and want to understand how it works on a basic level[1]. Tyler Hobbs' blog (OP link) also has some great essays on generative art[2], as with the blog of Anders Hoff[3], and also lots of other resources in awesome-creative-coding repo[4].
[1] - https://generativeartistry.com/
[2] - https://tylerxhobbs.com/essays
[3] - https://inconvergent.net/#writing
[4] - https://github.com/terkelg/awesome-creative-coding