Milady Maker is a collection of 10,000 generative pfpNFT’s in a neochibi aesthetic with randomized cosmetics inspired by 00’s Tokyo street fashion. Under the vision of digital artist Milady Sonora, Milady’s are designed to be fashion-minded and genuinely good social media avatars, and an invitation into their personal aesthetic world.

Sample Milady

I helped Milady Sonora conceptualize and design the Milady Maker project into something suitable for a pfpNFT, a practical execution on my thoughts on a profile-first pfp project described here. This write-up will walkthrough the thought process behind our design.

Milady Sonora is a digital painter who has shown works in galleries both physically and digitally through Remilia Collective under this psuedonym.

She’s developed her distinctive take on the Chibi Anime style in a portraiture series over the years, one which invokes a calming, ethereal, white-hearted angelicism in a blend of Western and Japanese aesthetics.

Takashi Murakami's "Chibi Squishy"

This is style we’ve taken to calling “neochibi”, being a contemporary take on the chibi aesthetics once explored by artists like Takashi Murakami, a part of a larger wave of “avant-anime” art being explored by digital artists such as Mara Barl, 情緒 and Rednbou.

Milady Maker as a project aimed to both explore and evolve the field of pfpNFTs, fixing certain problems identified in earlier designs while approaching the novel space with an experimental ethos.

The development of a generative NFT is something that cannot be separated from its financialization; its future aftermarket is an intimate motivation at every aspect of the design, inseparable from the uneven rarities that drive speculative interest in the art, and the art itself is generate by the process of it being purchased, rendered into and transferred as a financially valued asset.

It’s odd to us that this aspect has been rejected and shied away from by generative NFT artists in this space, apparently ashamed of the inherent commercialization of their work. We find this attitude demure and misguided, and one that prevents real exploration of the medium, because it ignores half the conceptual interest in a generative project.

Large series generative NFT do not necessarily need to follow the holy trinity of magic value creation they’ve inadvertently stumbled into, but they do because it produces genuine markets:

(1) Virtual scarcity induced by an artificial total on-chain supply,

(2) Variable scarcity induced by uneven rarity distributions o create variable scarcity,

(3) Underpriced (no-price discovery) sale on an unknown/lottery basis.

Each of these elements together produce a market where buyers can get in at a low price, have a chance to obtain highly rare NFTs that can be worth more than their purchase; virtual conditions similar to a collectable card game, but innovative (and profitable) for its transparency and efficiency.

This is a novel financial sector that happens to share confluence with artists as the lead directors of what is realistically an investment asset. It is only natural as artists we would want to experiment with and explore every element that underpinning this medium; which means recognizing the power our design has to control and define a microcosmic financial market. It’s an exciting opportunity.

The design decisions described in this document can be understood as motivated by this excitement on one hand; and on the other, by being in a position to intervention as aesthetes. The last motivation driving the decision to make Milady is that it’s not just easy to us, it’s a lot of fun, and we’re addicted to having fun.

An interesting, unplanned development in generative NFT projects is buyers using the works as profile pictures on online communities. In retrospect, this makes sense, as the field operates almost entirely by word-of-mouth from digital networks. Unfortunately this is not something that most of the earlier generative NFT projects have designed with in mind.

Milady’s were designed to be great profile pictures from the start, with a higher vertical ratio so that a square crop will be a perfect zoom to convey emotional expressiveness, while still retaining the space to show off full accessories and the scoring overlay.

The design of the accessories also intentionally lent high-value rarities to larger or more distinctly colored pieces, with the intention that a valuable Milady can be read as such even as a thumbnail.

This is also why an NFT header add-in is planned as the gen 2 production, to cement our recognition of the unique relationship between generative NFT’s and online spaces - perhaps as an early, organic structure for Ethereum based identities!

Assets were carefully curated to ensure good outfits despite being randomly distributed. The Chain Algorithm was also added to lend the generated works more aesthetic coherency, something that seemed missing in earlier generative NFT projects which followed more straight-forward RNG.

A major motivating factor for this design was also the desire for relative value to be intuitive; without knowing the ‘meta’ of the Milady market, one can ascertain higher or lower value Milady’s at a glance by appraising their drip.

This is achieved understanding the direct relationship between rarity and secondary market value; distribution rates were decided by first assessing each accessory by its “drip”, making the highest drip accessories most rare and thus most valuable.

With Milady Maker we were interested in approaching the generative art space from an angle of sartorialism by using the Milady as a template for fashion accessorization.

The thematic “Cores” were chosen around a loose narrative of throwback Tokyo street fashion, as seemed suited to Milady’s fusion neochibi aesthetic. Fruits magazine was a primary reference.

FRUiTS was a very influential fashion photomagazine documenting subcultural street fashion in Tokyo through the late 90s and 00s.

We chose to place the emphasis in rarity generation on cosmetic accessories rather than the base “racial” traits of the avatar, achieved by having even distribution for most facial expressions/hair/race, etc.

The initial plan was to have an even distribution of skin tone - pale, peach, brown and black - with extra-rare ‘species’ - elf, vampire, etc.

However, noticing that this was a standard in other humanoid pfpNFT projects, and that it resulted in a racist price floor with the black-skinned NFT’s being valued lower compared to tokens of otherwise same rarity (see: Cryptopunks, Meebits). Rarity is a lever we can pull to predictably control future price speculation, so we decided to buck the trend and make black Milady’s a rare trait, ensuring a high value for these visually distinct Milady’s. And they do simply look good, accentuating the Japanese ganguro aesthetic.

We began by defining the 5 Cores to be covered under our thematic accessories:

• Gyarou

• Harajuku

• Prep

• Hypebeast

• Lolita

Each of these are distinct, unmistakable style tribes in Tokyo’s street fashion scene, but were also chosen for the eclectic style clashing produced by the random generation resulting in interesting fits that ‘work’.

Every hat and shirt accessory is defined under one of the five cores.

An algorithm was introduced to give a slightly higher likelihood for the second of the two primary accessories (hat/shirt) to be of the same core as the first, to weight the scale in favor of aesthetic coherency.

We gathered up a moodboard of inspiration for each Core, mainly of street fashion shots, and then had Milady Sonora simply draw up five to ten fits on top of a base Milady template under each core.

All the items were prepared on separate layers, so it was trivial to then pool together the complete set of accessories.

A curation process was then applied with the final list of assets based on the random generation that was produced, and items that clashed too strongly - looking good only in its own perfect fit, but not in other contexts crossed with other cores - were thrown out.

Next we individually went through each accessory and gave it a 1-5 grade by its drip; a gut feeling: is it eye-catching, is it cool and is it either expensive or take a lot of confidence to pull of?

This drip grading was then used to apply the rarity distribution for each accessory, such that the most dripped cosmetics = most rare = most valuable Milady.

A drip grade is tied to every accessory and trait, but weighted in the final sum according to the visual prominence of the slot: a rare earring is worth less than a rare hat.

Finally, a small bonus is given to Milady’s whose shirt and hat fit the same core, to honor their coherent fit.

The Drip Grading is designed to accomplish two goals:

(1) Rarity can be immediately and quickly assessed based on the actual drip of the Milady without needing to refer to any rarity chart

(2) The highly valued Milady’s which will ultimately come to represent the project in the aftermarket are rightfully those which look best

A Drip Rank is applied based on the Drip Score, ranking Milady’s into buckets from SS - S - A - B - C - Normal. These ranks are borrowed from standard Japanese video- and cardgame ranking conventions in spirit with Milady’s sartorial citations.

The Normal rank is the lowest grade, one which we originally had demarcated as “D”, but during testing, our hearts melted for the drip-less Milady’s at the bottom of the rarity ranking. We realized they held their own charm, simple and sweet, trying their best but not try-hard, and couldn’t bring ourselves to label them D. Instead we dropped the rank from their visual score card, and gave them the trait “Normal”.

In doing this, we also hoped to produce an interesting counter-valuation at the bottom of the pool. A speculative market for the most normal Milady, inverse to the valuation of high drip.

Since minting has begun, we have seen some speculative action on these lines:

The final step in Milady generation is the visual score card, comprised of 4 elements:

• Hat Core

• Shirt Core

• Drip Grade

• Drip Rank

The design was done in citation to Japanese trading cards, and a fun way to clearly categorize the Milady traits that otherwise define them in the database.

Beyond the actual generation of Milady as an NFT collection, we have large plans as the world’s favorite avant net art collective tasked with “saving the internet” as a holy jihad, which Milady’s sale just seeded.

New community tooling and spaces are being developed around the Milady community to bolster it, but Milady as an NFT functionally serves as identification card for our cult of 10,000: authenticating individuals for new works, new network experiments, new shows, and as elite aesthetes self-elected as surfers of the new wave. Watch this space =)