• By Amy Castor and David Gerard
  • This is part three of the first draft of an early history chapter for our planned NFT book.
  • We welcome corrections and nitpicking, so pick away!
  • Don’t forget to subscribe to our Patreons. Amy’s is here and David’s is here. We need your support for projects like this!
  • As an incentive, we’ve put the full interview with Adam McBride, mentioned here, on our Patreons for subscribers. It’s a great 3,000-word historical rundown.

Curio Cards: the first art NFTs on Ethereum

After the 2017 crypto bubble popped, the early attempts at NFTs faded into obscurity. But many were rediscovered in the 2021 NFT craze.

Buyers would scour the Ethereum blockchain for old NFTs, pick them up for a song, and then, once the discovery was widely broadcast, flip them for a handsome profit. This often required extra coding to get the old NFTs’ smart contract code to work on the OpenSea NFT marketplace — but there was always a “community member” willing to put in the work.  

In early 2021, most people considered CryptoPunks to be the first Ethereum-based art NFTs as we know them — but then Curio Cards was unearthed. Curio Cards was originally launched on May 9, 2017 — predating CryptoPunks by six weeks. 

Etheria, which launched only a few months after Ethereum’s mainnet went live in 2015, predated Curio Cards. But Etheria was a virtual world, and its NFTs were for plots of virtual land — not art. Still, even Etheria land would resell for big money in 2021.

Curio is a 30-card series featuring work from seven artists: Cryptograffiti, Cryptopop, Daniel Friedman, Marisol Vengas, Phneep, Robek World, and Thoros of Myr. It’s actually a 31-card series if you count the misprinted card #17 — called “17b” — as part of the series.

Like all NFT projects, rarity was the big selling point. Curio Card created between 111 and 2,000 cards of each image, for a total of 29,700 cards. The images are of everyday objects, like apples and acorns, mixed in with the occasional corporate logo altered to mention bitcoin. 

The idea of Curio was to put artists’ work onto a blockchain and display it in an online art gallery, so the artists could sell their work without a middleman taking a cut of the profits — 100% of primary sales went to the artists. This didn’t add up to much, though. Before 2021, the cards only sold for between 25 cents and one dollar, and even that was a difficult sale to make at the time. 

Bitcoin street artist Cryptograffiti contributed three pieces to the collection. As an activist, he used to make stickers related to bitcoin and banking and roam around San Francisco sticking them on bank walls and ATMs to show his disdain for modern finance. Curio Card #12 is a 1990s MasterCard logo with the words “MineBitcoin” in place of “MasterCard.”

Now-famous NFT artist XCOPY, who was unknown at the time, missed being included in the collection by a hair.

“You could put in your Google form to be part of Curio Cards, and he got it in too late and was denied,” Adam McBride, an “NFT Archeologist” who spends countless hours per day researching early NFT projects, told us. “If XCOPY was part of the Curio Cards, it would have been a whole different story.” The collection would have been even more coveted today. 

The initial founders of Curio Cards were Thomas Hunt, a bitcoin Youtuber known for the Mad Bitcoins show, and software developer Travis Uhrig. The two met at a bitcoin meetup in San Francisco. Hunt’s love of baseball trading cards gave him the inspiration for putting art on the blockchain: “I thought, why not combine my love for cryptocurrency with my love for collecting and make something fun?” [Bitcoin.com]

Hunt and Uhrig brought in former child actor Rhett Creighton to help with developing the smart contract — the bits of computer code that would run on the blockchain. NFT coding standards, such as ERC-721 and ERC-1155, did not exist yet, so every detail had to be coded by hand, and each card manually put onto the Ethereum blockchain.

All three founders were big fans of Looney’s Rare Pepe project, which preceded Curio Cards. [Show me the Crypto]

As a kid, Creighton starred in Crocodile Dundee II. As an adult, he studied physics and nuclear engineering at MIT. Rhett Creighton was his stage name, a shortened version of his full name, John Everett Creighton IV. In 2017, he also used the name Everett Forth, before reverting to Rhett Creighton.

Creighton was known for his entrepreneurial spirit. He used Amazon’s Mechanical Turk, a crowdsourcing website for businesses to hire remotely located “crowd workers,” to game the initial allocation of the cryptocurrency Stellar. [Inside Bitcoins, Everett Forth]

Creighton also started his own cryptocurrencies. The privacy-enhanced cryptocurrency Zcash originally had a “founder’s reward” that paid a portion of all freshly-created crypto-coins to the developers. In November 2016, Creighton removed the founders’ reward — he considered it inequitable — and called the resulting cryptocurrency ZClassic. Unfortunately, without a founder’s reward, the developers had no way of making money, so the currency didn’t go anywhere. Creighton went on to launch another cryptocurrency, Whalecoin, in November 2017, based on Ethereum, with a 33% “development fund.” He quickly abandoned this also. He then started Bitcoin Private by forking ZClassic, and abandoned that as well.  

Creighton suggested that Curio make its own version of Ethereum for the cards. Uhrig pointed out this would leave his digital trading card idea “orphaned” on its own chain, so they stayed on the public Ethereum blockchain. [Start with NFTs]

Each card included a link to a copy of each image on IPFS in the contract. IPFS is a distributed file system that works a bit like BitTorrent — as long as there’s at least one copy out there to seed a particular file, you can access it. According to McBride, Curio Cards was the first to use IPFS, and the links were still active four years later when the cards were rediscovered. As with Rare Pepes, Curio cards ended up being a collection of individual fungible tokens, each tracked by its own smart contract with its own total supply. [Bitcoin.com]

Curio Cards had a detailed business plan, with complicated smart contract coding. The creators wanted to create “FOMO” — fear of missing out — to ramp up interest in the cards. When a card launched, it sold for 25 cents in ether. The price would incrementally go up to a dollar over the following days.  The first ten cards (at least) had a distribution of 100,000 — but after 2,000 were bought, the rest were automatically burned. For later cards, the artist chose how many were created. As a result, every card had multiple smart contracts.

The cards could only be bought on the Curio Cards website — there were no NFT standards and no NFT marketplaces. The smart contracts were linked as “vending machines” from the website. You can create an NFT collection today in fifteen minutes that works to the standards, and be sure they’ll work in an NFT marketplace.

As Creighton was creating the contract for card #17, something went wrong, so he had to reissue the card. The misprint became “17b.”

“I don’t think anyone knows exactly what happened with 17b, but it was briefly released,” McBride said. “One person who is actually in the Discord was able to get, I don’t know how many 17b’s, before they realized their mistake and turned off the contract.”

Hunt and Uhrig tried to run Curio Cards like a start-up business. “I believe blockchain-based collectibles will be very big in the future,” Hunt predicted in a 2017 interview. Uhrig’s role was to pitch the idea to venture capitalists. But the idea never caught on. Nobody cared, and Uhrig could barely give the virtual cards away on a Discord channel. “It was a fun thing at the time, but kinda failed in a lot of ways getting any traction,” he said in an interview.  

​​Curio Cards lay dormant until March 2021, when @DieAping on Twitter (whose account has since been deleted) uncovered the old Curio “vending machines” on Github. These were early versions of the Curio smart contracts, still with active Curio Card contracts in them — though the links on the website to the smart contracts had been removed. [Tweet]

“For a while, people were just doing OTC trades through the Curio Discord, and that is just kind of trusting people,” says McBride.

In 2021, whenever an old NFT project was unearthed, you had to move quickly to get the tokens onto OpenSea. Some code would quickly be written to serve as a “wrapper” to interface the existing contracts to OpenSea.

The first rushed third-party wrapper created to get Curio Cards to work on OpenSea had a fatal bug: when anxious Curio Card holders sent their tokens to the contract, the tokens became forever stuck — “permawrapped.” [FAQ: Wrapping Old Cards]

As a result, Card #26 became the rarest card — only 105 were still tradeable — and the biggest bottleneck to getting a full set.  

In October 2021, a complete set of 31 Curio cards, including the 17b misprint, sold for more than $1.2 million in ether at the Christie’s Post-War to Present auction. Taylor Gerring, an early contributor to the Ethereum project, was the buyer. 

Despite early frustrations, the project turned out to be a success after all. [Tweet]

Feature image: Curio Card #23, “The Barbarian” by Robek World, depicting Curio founder Rhett Creighton.

3 thoughts on “The early history of NFTs, part 3: Curio Cards 

  1. “The first ten cards (at least) had a distribution of 100,000 — but after 2,000 were bought, the rest were automatically burned.” I don’t understand how this differs from printing 2000. Is it 2000 total sales across the collection?

    1. Each card had its own limited supply. (Distribution is the wrong word. I’ll fix this.) And to create FOMO, they would have a print 100,000 cards. After 2,000 cards were bought up, the smart contract would automatically burn the remaining 98,000. I don’t quite understand how this incentivized people to buy the tokens quickly, but it seemed they were experimenting to see what worked.

  2. There’s a good lesson in basic economics here. Utopians always try to cut out the middleman, resulting in no sales.

    Real artists and real authors and real manufacturers learned a long time ago that agents and publishers and dealers and salesmen are NOT superfluous.

    Artists know how to draw. Sellers know how to sell. Those are completely distinct skill sets and distinct personalities, and the system works best when each is able to use his own specific skills to the max.

    The digital world allows easy self-publishing, which works for a few stars who already have a buzzed-up audience. For most creators it’s a fake advantage and a cruel prank.

Leave a Reply