MOCA Proudly Introduces: The Library

Which we sincerely hope will blow your mind

This is going to prove a monstrously important week for the Museum of Crypto Art when we look back at it in a few months or years. Everything we create from this moment forward will be built atop the thing we’re discussing today, which we can think of as MOCA’s shiny new engine. It’s a long-time coming.

Our three highest-minded aspirations —

1. Deployable Art DeCC0 agents capable of curating, collecting, exhibiting, and talking about crypto art on-command,
2. fully-rendered art galleries created in moments through text-to-architecture interfaces,
3. our entire museum infrastructure turned into an open-source and deployable toolkit, usable by any person or institution who wants to make their own version what we’ve built

— are tethered to this: an R2R-synthesized Knowledge Graph capable of drawing insights and connections from a massive (and ever-expanding) database of criticism, curation, analysis, and musing about crypto art and web3. That was quite a mouthful, I know. Simply put, we’re talking about the neural infrastructure of an informational AI, which contains within it a near-infinite matrix of connections, a self-referential web of everything that’s ever been created in crypto art. We’re calling it, The Library.

When complete, The Library will be an unmatched compendium of all the information that’s ever been published about crypto art. As of now, it boasts over 600 unique documents, with more than 20,000 connections between them, but we imagine both of those numbers rising sharply as we incorporate more and more documentation.

As it pertains to preserving, understanding, and communicating the true continuum of crypto art — something outside the laconic and draconian narrative established on Twitter — a tool like The Library is not only peerless, it’s mandatory.

How Does the Library Work?

When we ask ChatGPT a question, for instance, it plumbs through the depths of its own gargantuan Knowledge Graph — the interconnected web of all information drawn from the Internet — for not only the answer we’ve asked for, but for related context that will increase our mastery of the subject. It’s not just us asking “When was McDonald’s founded?” and being told, “1940,” it’s that the LLM understands that, yes, McDonald’s was founded in 1940, and by Richard and Maurice McDonald, though the modern version arose in 1956 after Ray Kroc, a milkshake-machine salesman, convinced the McDonald brothers to let him franchise the restaurant, a decision which ultimately led to Kroc’s control of the company.

In essence, The Library is a crypto-art-centric Knowledge Graph, an ever-growing reference list of information about crypto art culture, creativity, and analysis. Via AI agents connected to this tool, the Library can be queried for answers and insight drawn from the vastness of all collected crypto art conversation.

R2R, which I mentioned before, is a powerful AI retrieval system that uses LLMs themselves to generate a Knowledge Graph out of internalized reference texts, forging a massive matrix of connected information. When we talk about DeCC0 Agents understanding the verbiage, names, cultural epochs, styles, mediums, dramas, collaborations, technical advancements, platform launches and closures, inspirations and influences, important figures and their successes and their blindspots, that are all specific to crypto art, they do this by accessing the Library. And the best part: The Library can be periodically rebuilt as new documentation is ingested, creating deeper relationships and more specific insights. The thing never finishes, it only gets better, smarter, deeper, and more creative.

Its forthcoming use-cases are staggering in their potential impact.

DeCC0 Agents

As we discussed in our MOCA 2.0 Announcement, our main priority for the year is establishing Art DeCC0s Agents as a thing, which basically means inventing the mechanisms for as-of-now-unprecedented agentic ability (and thus value for DeCC0s themselves). Our vision is for DeCC0 holders to imbue their DeCC0s with a personality — which implies a curatorial eye — and then deploy that DeCC0 to go complete tasks: collecting artwork with a budget, curating artwork from a collection into an exhibition, exhibiting that artwork in Virtual spaces. In all cases, these DeCC0 agents rely on The Library to understand crypto art’s context.

Ask DeCC0 agents a question, they’ll respond not only with specificity but with unexpected connections. They will know crypto art’s own historical context and draw parallels within it. They will be aware of an unmatched variety of artworks by a huge magnitude of artists, and weave them together in new and creative ways. We believe that there are historical trends which have gone hitherto unrecognized that DeCC0 Agents will uncover. Same with continuums of artistic influence. Crypto art lacks sufficient self-analysis, but DeCC0 Agents will fill in the many blanks left open by the movement’s own deficiency of literature.

DeCC0 Agents will be connected to The Library by default, but given our open-source foundation, any LLM or outside agent can connect to The Library’s contents and become a crypto art expert. DeCC0 holders can use the LLM of their choice — GPT-4o, llama 3.3, Claude, Nous Research’s Hermes, etc. — to fuel their DeCC0 Agent’s intelligence via compute, meaning these Agents can both pull from the massive datasets of these outside entities, and access The Library’s unmatched specificity/insight relating to crypto art itself.

A Few Notes: As of now, The Library exclusively includes documents that we have been given permission to use by their distributor or author. We publicly pledge that none of these documents will be used to train agents or models to produce content; they will function solely as references. But all informaiton on the internet has a clear desire to be referenced and read, otherwise it would not be published. We will ensure that sources of information are communicated as clearly as the information itself. Beyond our initial testing of The Library’s functionality, we will begin to include as much crypto-art-based writing as we can curate into the Knowledge Graph. If you have writing to direct our attention to, or that you’d like to contribute, please email cohen@museumofcryptoart.com.

Preservation of Culture

As The Library grows to include more documentation, the depth of its insight will grow, as will its understanding about crypto art’s participants. All of them.

This is how we preserve the real truth of this movement — who was here, what did they do, when did they do it, what did their work mean? — in an ecosystem that has not otherwise found a way to disrupt the too-influencer-weighted narratives of Twitter. The Library necessarily challenges predominant narratives, providing alternatives rooted in multi-vectored analysis which supersede individual bias or market-driven agendas.

On top of that, the complete contents of The Library’s Knowledge Graph are preserved on our back-end indelibly for the future. No more praying that some long-ignored servers will remain online or that Substack will host defunct newsletter in perpetuity. This information is important, too important to be at the mercy of whims which do not notice it. If nothing else, we want to make sure this knowledge persists through time until scholars and academics take wide notice of crypto art and come to study it en masse. Let nothing that happened in crypto art’s history go unappreciated or un-included in the capital-t Truth of it all.

New Museum Spin-Ups

We anticipate other institutions and individuals — museums, collectors, curators, whomever — coming to realize that The Library system is too powerful, too easily-constructed, and too useful for them to ignore. Because it is designed open-source, any institution can use our tech to create their own “Library,” for their own specific audiences, based on whatever information they want to explore further or preserve.

This is an extension of the same reasoning for which we released MOCA ROOMs. With ROOMs, we wanted to put the exhibitive power of our museum in as many hands as wanted it, so that everyone could own a piece of MOCA, exhibiting their own artworks from their own collections just as we did. In a sense, we’re taking that same ethos and applying it more broadly. The very foundation of an AI-aligned Museum — the hosting site, artworks display, internal Knowledge Graph, the mechanisms and weights — made open-source and portable.

Why This is Vital for MOCA’s Continuing Mission

The whole point of our MOCA is to disrupt. But we can’t disrupt alone, nor do we want to. There are so many larger narratives that when analyzed prove ungrounded in fact, or motivated mostly by market-motivated individuals with great power and great reason to construct the history of a thing around themselves. Truth rests on two shoulders: information and awareness of it. We hope that The Library is a tool that many of crypto art’s institutions will adopt and use for themselves. We hope it will give rise to much more writing, that it will be an invaluable source for constructing retrospectives of this art movement’s many moments and nuances. We want others to iterate atop The Library, make it better, and thus make us better.

A more informed crypto art public makes us all smarter. A more accessible crypto art history destabilizes established narratives and reveals the true achievements of every member of this community. What could be more noble work than bringing this to fruition? What is MOCA meant to be if not a fighter for what crypto art is really about at its deepest, and most interrogated, and most representative?

What’s Next

This week (probably), we’re launching our redesigned MOCA 2.0 website: our same collections in an entirely new interface, which is smoother, and it’s capable of displaying all our collected artistry — from videos to still-images to 3D sculptures to giga-pixel artworks — in their native forms. With that comes the launch of The Moral Collection by Matt Kane, combining imagery and analysis into the most in-depth analysis of Matt Kane’s 2023 Moral Performance (perhaps better known as the Contractual Obligations performance), which is perhaps the most in-depth analysis of artwork in crypto art history (there’s no better and more deserving subject).

As we discussed and demonstrated this past week on our Weekly Town Hall (every Thursday at 12pm EST in our Discord), The Library is approaching full-functionality in its testing phase. We are also working hard to pioneer new avenues towards agentic personality. Our next step is to teach The Library to recognize pieces from elsewhere on the internet that deserve a place in its dataset, which we will confirm, seek permission from its creators/distributors, and curate into its Knowledge Graph.