On Generative Practice: A Working Definition

This essay is part of the Working Definitions series from DAP LUXE Cultural Futures Lab. The series documents the maison's relationship to inherited tools, vocabulary, and methods. On Generative Practice applies that framework to AI tools.

Source

Generative AI as it exists in 2026 is a young instrument. The technical lineage runs through deep learning research in the 1980s and 1990s, neural network breakthroughs in the early 2010s, and the diffusion model and transformer architectures that became consumer-facing tools between 2022 and 2026. The instruments are not from any one cultural tradition. They are recombinant tools built on the global commons of human image, language, and pattern.

The instruments are not neutral. The training data that shapes what generative tools can produce was assembled from sources that overrepresent some cultures and underrepresent others. The aesthetic defaults that emerge from these tools tend to favor the cultural patterns that dominate the training data. Working with generative tools without naming this means accepting the defaults.

Colonial export

The dominant frame for generative AI in 2026 treats the technology as inevitable, scaleable, and culturally neutral. That frame benefits the institutions that hold the largest training datasets and the largest compute, most of which are US and European technology firms. The frame positions AI as the next industrial revolution, framing usage as either embrace or resistance, with no room for measured engagement.

That binary framing is itself a colonial export. The fashion industry has seen versions of this framing before. The maison concept was framed the same way: French institutional weight as the global standard for luxury, with parallel African and diasporic traditions positioned as either imitating or rejecting the standard, with no third position available.

DAP LUXE works in the third position by default. The maison uses inherited vocabulary (atelier, maison, house) while reframing what those terms refer to. Sankofa Forward applies the same move to generative tools.

Stewardship

DAP LUXE Cultural Futures Lab uses generative tools for one purpose: to extend the maison's existing visual vocabulary into speculative territory that human photography cannot reach. The vocabulary is the maison's own: the tapestry method, the Royalty Collection's conversion of historical artworks into wearable artifacts, the African royal textile lineages documented in the Blyden Blueprint, the brushed-gold and obsidian palette, the Didot and EB Garamond typography.

The generative tools do not produce the vocabulary. The vocabulary produces the prompts. The Lab supplies what the maison already knows; the tools render what the maison wants to see next.

Every work produced under Sankofa Forward carries an AI-GENERATED · 2026 label. The label is not a disclaimer. It is a credit line, the same way the Royalty Collection essays credit the source artworks, the same way Grounding Together photography credits BashLNCS.

The Lab does three things to make this practice legible:

1. Labels every work. No ambiguity about authorship. Every tile in the Sankofa Forward gallery, every lightbox view, every external citation carries the label.
2. Files the method. This essay is the canonical document of the Lab's generative practice. When tools change, the essay updates. The maison's audience can always see what the Lab is using, what the Lab is not using, and why.
3. Pays human collaborators in parallel. The Royalty Collection is made at the atelier. Grounding Together is photographed by BashLNCS. Currency of Culture is produced in partnership with Black Dollar. Ancestral Futurism in Practice is exhibited with human curators in human institutions. Sankofa Forward does not replace any of this; it extends the vocabulary into speculative territory while the human-anchored work continues to anchor the maison.

What this is not

Sankofa Forward is not the maison's argument for AI as a creative medium replacing human creativity. The maison takes no position on that broader debate, because the maison's interest is not in winning a debate but in doing the work.

Sankofa Forward is not a substitute for the maison's human-anchored programs. The Lab's human-led work outweighs its generative work by every measure: number of garments produced, hours of human labor invested, dollars circulated to human collaborators, exhibition footprint in physical space.

Sankofa Forward is not a technology proof of concept. The Lab is not interested in demonstrating what the tools can do. The Lab is interested in extending what the maison already does into speculative dimensions.

What this is

Sankofa Forward is the maison's deliberate position on generative practice in 2026. The position is: engage explicitly, label transparently, document the method, keep human-anchored work primary, treat the tools as instruments rather than authors.

The Lab opens this program because the alternative was worse. The alternative was to either pretend the tools do not exist (which the maison's audience would see through immediately) or to use the tools without naming them (which would compromise the maison's credibility with the audiences that matter most: museums, curators, serious technology partners, and the community that has earned the right to know how the maison works).

The third position is available. The Lab takes it.

Related references

• On the Atelier: A Working Definition · Working Definitions I
• On the Maison: A Working Definition · Working Definitions II, forthcoming Q3 2026
• On the House: A Working Definition · Working Definitions III, forthcoming Q4 2026
• Sankofa Forward · the Lab program documenting this practice

Filed: May 16, 2026 · Owner: DAP LUXE Cultural Futures Lab · Status: Strategic position document. Updates as tools and the maison's practice evolve.