I just came across a super interesting NFT project design called Slonks, with a somewhat crazy concept—it's basically designed to self-vanish.

First, the background. Slonks reconstruct CryptoPunks using an on-chain neural network model, with each Slonk corresponding to a punk, rendered in 24×24 pixels with a 222-color palette. The problem is, the model isn't perfect; the pixel difference between the rendered image and the original punk is called the slop value, ranging from 0 to 576. This setup sounds very technical at first glance, but it’s actually the entry point to the entire economic logic.

The project has two paths. The first is voiding: holders send their Slonk into a void contract, minting a corresponding amount of $SLOP tokens based on the slop value. For example, slop=15 mints 15 $SLOP, slop=400 mints 400 $SLOP. The total supply cap of $SLOP is 5,760,000. The second is merge: two Slonks of the same level can be combined, with the donor permanently burned, and the survivor upgraded to the next merge level, recalculating slop with a new embedding.

The key here is that after merging, the slop value will increase significantly, with the rise depending on the morphological difference between the two source punks. For example, merging a normal male punk with a female punk, which has the greatest shape difference, causes the slop to spike. An example is #608普通男性slop=15配#645, a normal female with slop=3; after merging, slop=161, an increase of +146. When both are voided, only 18 $SLOP can be minted, but after merging, this NFT becomes #608單隻就能鑄161 $SLOP。代價是永久失去#645.

Deployed on May 1, five days ago, the total supply dropped from 10,012 to 9,505, meaning over 5.06% of Slonks have been permanently eliminated. And this is before voiding even started; the $SLOP contract hasn't been deployed yet, so all burns are purely merge actions. Players have been strategically consuming Slonks based on slop values even before $SLOP launch. The merge rate is accelerating: on day 3, 175 were burned during community testing, and by day 6, it surged to 124.

This is where it gets really interesting. Slonks have two conflicting valuation logics. The first is based on the rarity of the source punk. In CryptoPunks, 9 Alien, 24 Ape, and 88 Zombie are recognized as rare. If your Slonk’s source is Alien or Ape, no matter how it’s rendered, the visual recognition of that green face or ape face cannot be changed. Low slop is the core value of these NFTs because it preserves the rarity premium.

The second valuation is based on the amount of $SLOP that can be minted through voiding. The higher the slop, the more $SLOP minted. A normal male punk Slonk with slop=400 can mint 400 $SLOP, while an Alien with slop=15 can only mint 15 $SLOP. In terms of $SLOP price, high-slop normal punks are worth much more than low-slop Alien punks.

Here’s the problem: if you hold a low-slop Slonk whose source is Alien, merging will destroy both values. First, the new embedding will no longer faithfully replicate the Alien’s green face, visually losing the recognition of rarity. Second, it permanently burns a donor NFT. You get a Slonk with higher slop, but its visual recognition is gone, and the rarity premium in the CryptoPunks community disappears. What you gain is only the potential to mint 100 more $SLOP through future voiding, but the $SLOP price remains uncertain.

Conversely, low-slop Slonks with ordinary sources are undervalued. They lack source rarity and the $SLOP quantity value. Merging is expected to be positive for them—transforming them into high-slop NFTs is the most valuable outcome.

Extrapolating to the entire collection, a Darwinian evolution emerges. Slonks with poor model reconstruction are the survivors—since slop is a gold mine for voiding, no one merges them, so they persist. Well-reconstructed Slonks split into two tiers: those with rare sources survive because merging destroys their visual recognition; ordinary source Slonks are recycled as donors and merged away.

Ultimately, this NFT collection will evolve into two coexisting types. One is high-slop, perfect failures—models render punks into grotesque forms, but they have high voiding value. The other is low-slop, rare faithful reconstructions—sources like Alien, Ape, or Zombie, retaining visual recognition and high secondary market value. The middle-tier Slonks keep disappearing. Ordinary source, low-slop NFTs have no valuation advantage—they’re natural donors for merging, contributing to the two survivor types, and themselves exit permanently.

The 5% disappearance rate over five days will continue to rise as the merge tools become more widespread and $SLOP launch approaches. Slonks are fundamentally designed as NFTs that actively disappear. Their rarity isn’t fixed at the launch with a set number like 10,000; it’s a collective result of every merge decision made by the market. Each merge is a market vote, permanently removing one NFT from existence in exchange for an upgrade of another.

This project’s logic is truly counterintuitive. Upgrading costs grow geometrically, but rewards do not. It’s the opposite of traditional games: Slonks upgrade by throwing similar ones into a furnace. The merge level isn’t recorded by how many times you’ve won, but by how many same-level NFTs you’ve burned. The level is a mark of how many similar NFTs have been destroyed.