Brief intro

The goal of this article is to summarize my thoughts on the generative art landscape from a platforms point of view. Mainly trying to answer the question, where as a platform we can do better for artists & collectors. These observations come through my experiences as an artist, collector and founder of a generative art platform. The key topics we will be covering are as follows:

• Preservation: How well preserved is your (on-chain) art?
• Gatekeeping: Is generative art being gate-kept?
• Cost: Can artists release on-chain without spending a fortune?

We will compare the above in general and also specifically for multiple generative art platforms, more specifically: ArtBlocks, FxHash, GmStudio, Gen.Art and 256ART.

Other platforms won’t be covered as most (all?) of them use ArtBlocks engine under the hood.

As the founder of 256ART, I will try to set aside my bias towards my own platform when looking at each of the above topics, reflecting on how platforms can do better, and in doing so challenging the status quo.

Preservation

Storing generative art on the blockchain is done in four ways. It can be:

• Centralized: a url to a file hosted on a centralized server, if server goes down, file is no longer accessible.
• IPFS: a url to a file hosted on IPFS, as long as one node in the IPFS network is mirroring the content it remains available, most content is only mirrored by one or a handful of nodes.
• On-chain: in the context of generative art this usually means you can reconstruct the art from data available on the blockchain.
• In-chain: all data is entirely available on-chain (including traits, libraries, etc.) and the art can be constructed on-chain as well.

How are platforms handling preservation?

ArtBlocks, GmStudio, Gen.Art and 256ART all put the art scripts on-chain. They also all provide generated off-chain assets for images and traits and rely on off-chain libraries (p5js, threejs). Some of these libraries now have certain versions available on-chain, but this isn't directly being leveraged by any of the platforms at the moment. FxHash is the odd one out here, where the art is not stored on-chain at all, and instead stored on IPFS exclusively.

For ArtBlocks, GmStudio and Gen.Art, you cannot render your artwork directly from chain and instead have to re-construct it with the data available yourself. With 256ART you can render your artwork from chain, but if a library (e.g.: p5js is used) is used, it still depends on the CDN for that library remaining available. With FxHash, as the art is not on-chain, so there is no way to reconstruct your art from chain, instead they leverage IPFS (but the pointer seems to go to their specific gateway at the moment, meaning if their gateway goes down, the IPFS url won’t work).

For ArtBlocks, Gen.Art and 256ART the off-chain assets are hosted on centralized servers and should be seen as an extra service for collectors. GmStudio hosts their off-chain assets (not including libraries) via IPFS (but the pointer also goes to their specific gateway at the moment). Gen.Art mentions they also store the assets on IPFS, but the one series I checked (Mutual) didn't seem to do so.

What does this mean if any of the above platforms stop existing and can no longer host the files (either on their centralized server or via their IPFS gateway)? For all platforms it means any marketplace or interface using the defined standards would no longer show your art. As simple as that. Does this mean your art is lost? No, not necessarily (see above), but it could have a severe impact on the perception of the art and it’s value if it’s no longer easily accessible.

With regards to preservation, one aspect that's often not thought about is the ever-changing nature of technology. If the coding language the art was created in, no longer supports certain functionality the artist coded, the art wouldn't render anymore. If there no longer is a client (in the case of generative art, often times a browser) to support the language the art was coded in, there's no way to display the art live anymore, etc. etc. In short, your generative art is unlikely to last forever. But that doesn't mean we shouldn't try to make it last as long as possible.

Challenging the status quo

It seems to me that there is a need for a platform that does support fully in-chain generative art (including traits and libraries). It's also debatable if there should be an option for artists to update their in-chain art script (allowing the option to update outdated functions, libraries etc. in the smartcontract). All of this will not make the generative art last forever (see above), but at least will preserve it for as long as possible.

Gatekeeping

It's very hard for artists to release their art on-chain without releasing through a platform. There are multiple hurdles, from the difficulty of Solidity development to the cost of on-chain storage on Ethereum and the struggle of getting your work in front of collectors.

Ethereum and Tezos

On the Ethereum blockchain, there currently is no open platform (but multiple teams are currently working on this). This means curated platforms are gatekeeping the generative art that will be released on Ethereum. The cost of release on Ethereum is so much higher compared to Tezos and platforms have to protect and ideally improve their financial bottom line as well.

This leads to the observation that on average the (perceived) quality of the art is higher on Ethereum. But it also means that there is less experimentation (this is very clear if you compare FxHash to the other platforms). ArtBlocks is admiringly one of the platforms on Ethereum that tends to allow for more experimental releases, as they have the financial runway to do so.

I do also want to note, that I don't think curation by platforms is necessarily a bad thing as it does help collectors make more informed choices, but there has to be an easier accessible alternative as well for artists to release their art and collectors to collect that art.

How are platforms (not) gatekeeping generative art?

On Tezos there is FxHash, which is fantastic as each artist can release art through their open platform. The only form of gatekeeping there, could be the verification badge for artists (and I think it’s definitely the good kind of gatekeeping). As it stands however, it's not possible to release art in- or on-chain on FxHash (see preservation).

If you're an artist that wants to release art on-chain on Ethereum and you don't get accepted by a platform? You're somewhat out of luck. If you want to release your art in-chain anywhere, even more so. Maybe the artist will try to release by themselves, or they’ll release on FxHash instead, but that might not have been their intent at all.

Another form of gate-keeping, but on the collector’s side, is the use of memberships. Whilst memberships may seem like a good way to reward early believers in the platform (and in the time of royalties could provide a sustainable stream of revenue for platforms), they also gatekeep access to generative art. The two platforms not having a membership model seem to have done better compared to platforms using the membership model. I don't think this is per-se because of the membership model, but it's definitely worth noting.

Challenging the status quo

Allowing artists to release their art on- or in-chain on an open platform on Ethereum is much needed. Ideally, the open platform will give as much control as possible to the artist (e.g.: different mint mechanics etc.). This will allow artists an easier avenue to release on- or in-chain art on the Ethereum blockchain, without having to go through the hurdles of being accepted by a platform.

It might also be sensible to seriously revamp or completely move away from the membership model in its current implementations.

Cost of releasing generative art

I think one of the main reasons we haven't seen an open platform on Ethereum yet, is because it's very costly to release on-chain generative art which is the expected standard on Ethereum. Most platforms on Ethereum will cover the cost of smartcontract deployment and on-chain storage (which depending on gas, price of eth and size of the script usually ranges from a couple hundred USD to a couple thousand, the way it's currently done). ArtBlocks is the exception here, where artists are required to pay for their own deployment & storage. The cost of releasing on- or in-chain art is also one of the factors why for so long there have only been curated platforms on Ethereum.

Can we make it cheaper for artists to release their art?

There are ways to lower the cost of on-chain storage drastically. With 256ART we were the first platform to store the art script on-chain in contract bytecode, which is roughly 4 times cheaper to how it was stored before (this is also why 256ART could from the start support upcoming and more experimental generative artists, as we could have a lower mint price due to our costs being lower). Since the launch of 256ART, we have seen all platforms one by one change their contracts to also incorporate storing on-chain data in bytecode. Which has been amazing to see. Whilst, four times saving in smartcontract storage cost is a big difference, it is still not cheap, especially for upcoming artists.

Right now, the cheapest way for artists to release, without having to go through an application process of one of the platforms that covers all costs, is FxHash. The problem here is that the art is not on-chain at all. FxHash does however have on-chain storage on their roadmap, which should make it a great option for releasing on-chain art, if an artist wants to preserve their work on Tezos.

Can an open platform succeed on Ethereum?

With on-chain storage being so prohibitively expensive on Ethereum, does that mean it's impossible for an open platform to succeed on Ethereum? Not at all in my opinion. Especially considering progress in technology is being made each day. To give you a direct example: recently I have been working on an even further improved version of on-chain storage, specifically for fully in-chain generative art, which could lower the cost on average another two to four times.

Challenging the status quo

Artists should have the opportunity to release their art on- or in-chain in an as cost-efficient way as possible. Platforms not fully covering the deployment costs for artists should strive to make the financial burden for releasing on- or in-chain art as low as possible, allowing more opportunities to upcoming artists who may not have the financial means to release their art otherwise in a way that it can be preserved as long as possible.

Conclusion

There are multiple points where platforms can and should improve:

• Add support for fully in-chain generative art, allowing for the best preservation of the art.
• Work towards open platforms, without gatekeeping neither artists nor collectors.
• Keep innovating to lower the costs of releasing on- or in-chain generative art.

You might have seen it coming, but this is exactly what we're working on with 256ART right now:

An open platform, without any gatekeeping where artists can release fully in-chain generative art in the most cost-efficient way and collectors won’t be gatekept with memberships.

At the time of writing a severely overhauled smartcontract set up is finished, in line with the above mission. We’re now working on the design, front-end and back-end.