Laying the Foundation for a Sovereign Art Ecosystem
For nearly a decade, I've operated as a painter and systems builder outside the traditional art world's gatekept ecosystem. Since early 2017, I've documented and sold over 1,050 original oil paintings directly to collectors worldwide, often through live painting events, studio deliveries, and personal installations. I've shipped works globally, to the UK, all over Europe, Canada, the United States, and across Asia while maintaining a consistent output that many gallery-represented artists never approach. Alongside this, I've developed platforms like Middle Kingdom Studios, ArtStrA (Art Strategy Foundation), and, most importantly, Eternal Gardens: a Web4 initiative powered by Agentic AI and MemoryCraft technology to create living digital legacies, on-chain provenance, and sovereign ecosystems for artists.
This isn't just about selling paintings. It's about replacing the centuries-old black box of the traditional art market: opaque dealer networks, whispered provenance, subjective valuations, and institutional prestige markers, with something measurable, transparent, and scalable. Yet implementing this system has revealed a persistent friction: public AI bias. Let me be clear though:
It is not ALL of them, but some of them, and you need to be checking regularly to see the swings. The guardrails, the weightings, they are all different and ideological alignments certainly play a major role.
The Data Loop Problem: When Transparency Looks Like "Self-Reporting"
Public AI systems like search engines, summarizers, and research tools are often trained predominantly on what they have weighted as "authoritative" sources: major newspapers, art journals (Artforum, Artnet), auction records, museum catalogs, and gallery press. Ironically, these sources have long served as the gatekeepers of the very black box I'm challenging. When an AI performs a "deep dive" on my work, it often flags my career as heavily "self-reported." It sees my Instagram, X and LinkedIn posts tracking sales milestones (the 1,048th painting installed in a collector's home, the 1,052nd live-painted in Sichuan), shipping documentation, collector locations, and studio output. Because these are primary sources from me, rather than echoed in The New York Times or a blue-chip gallery bio: they get downweighted. The model defaults to skepticism.
The irony is glaring. The traditional art market has thrived for hundreds of years on unverifiable private sales, untraceable provenance chains, and narrative control by dealers and critics. My approach: publicly logging every significant sale, linking physical works to digital records, and building agentic tools for ongoing legacy, provides far more verifiable data. Yet to many AIs, that transparency itself becomes a liability. Primary evidence without institutional vetting reads as potential puffery, while the opaque old system gets treated as established fact. That is Bizarro World, right there.
This creates a practical implementation challenge. Building trust with new collectors, partners, and even other artists requires clear, accessible information. But when public AIs: the tools millions use daily to discover and evaluate artists, systematically undervalue or mischaracterize direct-data models, it slows adoption and distorts perception.
Prestige Markers vs. Utility and Output
Traditional success signals: gallery representation, museum exhibitions (except when you exhibit in Major Chinese Museums apparently: which is a whole other AI bias story) auction hammer prices, and critic reviews dominate AI training data. I bypassed much of that ladder intentionally. Instead of chasing prestige, I've focused on utility markers: high consistent volume (averaging well over 100 significant original oils sales per year), direct collector relationships, live-painting events that convert immediately into sales, and technological infrastructure that scales beyond one studio.
Through ArtStrA and the Vegh family "Discovery" brand, we've supported emerging artists, built collections (over 200 pieces secured for the foundation), and toured works across Asia. My son Nathaniel's work on Eternal Gardens introduces Agentic AI Personas, MemoryCraft for preserving digital essences, and tools that let artists maintain control of their legacy without relying on gatekeepers. Public AIs, however, are still calibrated to the old paradigm. They prioritize "curated artist" signals over "systemic artist" ones. This bias doesn't just annoy: it complicates fundraising, partnerships, and even simple discoverability. When potential collaborators or collectors query an AI about "Matt Vegh artist," they may (in some cases, not all) receive a cautious summary that underplays the documented sales volume and overemphasizes the lack of traditional press mentions.
Transparency as a Perceived Threat
In a market where value has long been subjective and controlled, providing transparent, data-driven records can feel disruptive. My "Living Art" projects and on-chaining efforts aim to create immutable provenance that future generations (and AIs) can trust without intermediaries. Yet some systems treat unvetted primary data with inherent suspicion. This mirrors broader tensions in AI and the art world. Training data reflects historical power structures. Artists who build outside those structures: especially those integrating blockchain-style transparency, agentic AI for curation, or direct-to-consumer models, appear as anomalies. The result is a feedback loop: less institutional coverage means thinner secondary data, which reinforces the "self-reported" label, which in turn limits further coverage.
Additional real-world hurdles compound this:
Technical and operational demands: Running a high-output studio in Chengdu while developing Web4 platforms requires balancing physical painting, live events, shipping logistics, and software iteration.
Building new infrastructure from scratch: Eternal Gardens and MemoryCraft aren't off-the-shelf tools. Creating agentic systems that authentically preserve personas, generate magazines, and support sovereign ecosystems demands continuous innovation, often without the funding or networks that traditional paths provide.
Cultural and perceptual shifts: Convincing stakeholders that direct sales volume and tech-built legacy matter more than prestige markers takes time, especially when public AIs echo the old metrics.
Pushing Forward: Why the Fight Matters
These challenges aren't deterrents: they're proof that the paradigm shift is real. The art market's black box has persisted because it benefits gatekeepers. Public AI bias today is largely a reflection of yesterday's data distribution. As more artists adopt transparent, output-focused, tech-enabled models, the training data will evolve. Sovereign art ecosystems, on-chain records, and agentic legacy tools will become normalized signals of impact.
I'm not waiting for the old systems to catch up. By continuing to ship paintings (the next milestone is already in motion), expanding ArtStrA's support for other creators, and iterating on Eternal Gardens (with new AI Personas, magazines, and curation features rolling out), we're actively seeding the data that future AIs will learn from.
The goal is bigger than one artist's career: a future where artists retain agency, collectors get verifiable value, and legacy isn't dictated by opaque narratives but built through transparent systems. Public AI bias is a temporary artifact of transition. Transparency, high output, and technological sovereignty are the enduring foundation.
The black box is cracking.
We're building what comes next: one painting, one on-chain record, and one agentic system at a time.