This week’s featured collector is Safetytrance
Safetytrance focuses their collection on pfps. Take a look at lazy.com/safetytrance
Last week’s poll on collecting AI art produced the most polarized result we’ve seen: 50% said “not for me, full stop,” while 33% said they’re already collecting it. Only 17% landed in the middle with “watching, not buying yet,” and the two nuanced options — collecting only work where AI is genuinely the medium, and worrying about provenance — both drew zero. That’s a striking shape. Our readers aren’t weighing tradeoffs on this one; they’ve largely decided, and they’ve decided in opposite directions. A third of this audience is actively buying AI work while half won’t touch it, with almost nobody occupying the careful middle ground that Christiane Paul and Refik Anadol spent the whole article mapping out. The zeros are the tell. We expected the “only where AI is truly the medium” option to perform — it’s the distinction the experts draw, and the one that separates prompt output from serious practice. Nobody picked it. Which suggests that for our readers, the medium question isn’t the deciding factor. Either you’re in on AI as a legitimate material or you’re not, and the craft distinction doesn’t move anyone across that line. Given that this same audience has repeatedly told us concept matters more than medium and that connecting with the artist matters most of all, the split may be less about AI’s aesthetic legitimacy than about whether there’s a person on the other side you can actually know.
Eighteen Months of Data on What Actually Moves NFT Floors
Every so often someone in the NFT space publishes something rigorous enough that it reframes an entire debate. This week Sergito, who leads MeebCo, published an eighteen-month retrospective on legacy NFT collections — indexed floor data, token launch outcomes, market cap comparisons — and it’s the most useful thing we’ve read all year on the question of what actually creates value for an NFT.
The headline numbers set the stage. Since MeebCo closed its Meebits acquisition in February 2025, Meebits are down 10% in ETH terms and CryptoPunks are down 9%. The median major collection is down 49%. And the four collections that launched tokens since late 2024 have watched their NFTs lose between 60% and 85% since those tokens arrived.
That last figure is the essay’s cleanest verdict, and it deserves unpacking.
Tokens: brilliant for the company, brutal for the NFT. The four token launches all followed the same arc — floor pumps into the claim, then bleeds. Pudgy Penguins are down 85% since PENGU. Azuki down 82% since ANIME. Doodles down 62% since DOOD. Moonbirds compressed the whole cycle into ten months: BIRB was announced in October, the floor spiked to that day, and by the January launch it had already fallen 68%. The market didn’t even wait for the token to exist.
Sergito is refreshingly clear-eyed about why teams keep doing it. Launching a memecoin is, in his words, the best business an NFT company has ever shipped — treasury allocation, permanent liquidity pool fees, and an airdrop that converts years of community loyalty into a single liquidity event. On pure business terms it’s brilliant. And that, he argues, is exactly the problem: the token is the clearest evidence we have that the company doing well and the NFTs doing well are two different things.
The fairest part of his analysis is that he counts both sides. Holders did receive airdrops worth real money. But on launch day, the NFTs that earned those tokens were worth on average 7x what they’re worth today — and the tokens themselves now trade 77% to 97% below launch. The only way to capture the headline value was to sell fast, which is a strange pitch for a long-term holder. An airdrop goes to every holder at once, which means every holder can sell at once. Today’s prices are what remains after everyone tried. And the holder who suffered most is the one who believed most: never sold the token, kept the NFT, watched both halves collapse.
Brand-building doesn’t reach the floor. This is the finding that should provoke the most argument. Every major project ran the same playbook — blind boxes, plushies, trading cards, huge social followings, partnerships with some of the largest companies in the world. And after years of it across the whole cohort, floors repriced as though none of it happened.
His explanation is structural, not cynical: without an embedded reward mechanism, brand success simply doesn’t reach the floor. Nothing connects ongoing revenue to the NFT unless someone deliberately builds that connection. And even if it existed, consider what would flow through it — the global toy industry grows about 4% a year. That’s a fine business for an established company, a strange one for a venture-backed startup, and a stranger one for holders pricing NFTs as growth assets. The line that lands hardest: the NFT doesn’t own the brand. The company owns the brand.
The exception proves the rule. VeeFriends is the only major 10k+ collection green over the past year, up 18% in market cap. Two differences stand out: no token, and a founder who personally underwrites community belief — Gary Vaynerchuk pushing the IP well beyond the crypto podcast circuit, and buying back VeeFriends NFTs with his own wallets to onboard new holders. Skin in the game, visibly.
The mechanism that does work. The essay’s central concept is what Sergito calls the attention-to-actual-value flywheel: attention becomes volume, volume becomes fees, and fees get programmatically reinvested into the collection rather than extracted from it. Legacy collections broke both ends of that loop years ago — royalty enforcement died in the marketplace wars, so volume pays the collection nothing, and Web2 revenue flows to the company and never returns. All that attention, monetized by everyone except the thing generating it.
Chimpers are the one legacy collection that went back and fixed it, migrating to a royalty-enforced contract and routing royalties under their own floor. Combined with a smaller 5,555 supply and adoption of the $CHIMPSTR flywheel, they’re up 121% since February 2025 — the best-performing legacy collection on his board.
MeebCo’s own data supports it. Their floor’s two best stretches both tell the same story. The acquisition itself sent it up 74% in twelve days. Then in September, TokenWorks launched MeebitStrategy ($MEEBSTR) — a token MeebCo didn’t create and doesn’t control, but which generated fees they directed into buying floor Meebits for treasury. The floor jumped 38% in nine days to its high of the year. He’s candid that much of it was given back in the autumn. But the mechanism is the point: the one time the revenue-to-NFT loop got closed, even partially, the market answered immediately.
Longtime readers will recognize TokenWorks — we covered their Fake World Assets protocol two weeks ago and their PunkStrategy flywheel before that. Seeing their strategy-token framework show up as the single clearest positive datapoint in an independent eighteen-month study is a notable validation of the mechanism, whatever one thinks of the broader genre.
And then there’s doing nothing. Punks are down 9% while the median major lost half its value. Sergito’s read: doing nothing isn’t really a strategy so much as a consensus. Punks are art, they’re finished, and the market long ago agreed they shouldn’t be touched. They’re the cleanest evidence available that provenance alone can carry an asset — with NODE Foundation balancing restraint on IP development against genuine stewardship of what makes Punks Punks.
His distinction between Punks and Meebits is worth quoting for anyone thinking about legacy collections generally: both were created by the same artists, so the Larva Labs provenance is intrinsic to each. But Punks were designed to be finished, and that’s part of their power. Meebits shipped as 3D characters with rigs and models, built from day one to be used, ported, and played with. Nobody expects Punks to be developed, and nobody should. Meebits were designed for it.
The test worth borrowing. The question Sergito applies to every initiative is one collectors could apply to any project they hold: does it feed the NFT or feed off it, and does it beat doing nothing?
That’s a demanding standard, and by his own data most of the last two years of NFT strategy fails it. Tokens fed off the NFT. Brand-building fed off the attention the NFT generated. Only two things passed — closing the revenue loop back into the asset, and, for a genuinely finished work, simply not interfering. It’s a sobering framework, but a clarifying one, and it’s built on evidence rather than narrative. That alone makes it worth your time.
This post is based on Sergito’s eighteen-month retrospective on legacy NFT collections: https://x.com/sergitosergito/article/2086843333553357061
Poll: What creates NFT value?
We ❤️ Feedback
We would love to hear from you as we continue to build out new features for Lazy! Love the site? Have an idea on how we can improve it? Drop us a line at info@lazy.com