3 ms·
So you have two unique NFT's claiming ownership of the same piece of art or tweet or whatever. Maybe on different blockchains, maybe not. Now what? Which one is
by dkersten 5y ago
So you have two unique NFT's claiming ownership of the same piece of art or tweet or whatever. Maybe on different blockchains, maybe not. Now what? Which one is the "real" one that conveys said ownership? How do you prove which one is the "real" one?
- mattdesl 5y agoUsually whichever came first, or is still in circulation (hasn’t been burned), or whichever the distributor/artist says is the real one. Similar to a signed and authenticated physical artwork - if a newer copy turns up, which one is the authentic?
- root_axis 5y agoWhen it comes to fake artwork you at least need to be skilled enough to produce a convincing fake, which isn't trivial. In the NFT world anyone can trivially mint an NFT for anything, at any time, so there's no way to prove that "the first one" is the "real" one without relying on a trusted authority outside the blockchain. Thus any need for a blockchain is completely obviated.
- mattdesl 5y agoHence “or whichever the distributor/artist says is the real one.” This does not prove copyright; it merely proves ownership of a token. It still requires a strong social consensus that the token and its addressable hash was indeed distributed by the artist (as in the case of most NFTs that are trading for any significant sum). This is very much like a certificate of authenticity of an artwork as it also requires social consensus (fwiw, a cert is far easier to forge than, say, a CryptoPunk).
- root_axis 5y agoSure. My point being that the blockchain adds nothing useful. Social consensus does all the heavy lifting, the cryptographic signature is an implementation detail, and the blockchain is wholly unnecessary.
- mattdesl 5y agoThe entire point of the distributed ledger is to help secure a stronger social consensus and build a traceable record of value transfer. Forking or copying a CryptoPunk (or Fidenza, to use an art related comparison) will not do much to sway the consensus that has formed around its contract address and transaction history.
- root_axis 5y agoYou don't need the distributed ledger to build a traceable record of value transfer though, the PGP NFT itself can be signed with an amendment documenting each transfer.
- mattdesl 5y agoEventually you will need a public ledger (to support hundreds of thousands of these signatures), and the choice of making it centralized (mutable) or decentralized (very difficult to change). This is how you end up with a public distributed ledger aka blockchain.
- dkersten 5y agoIf the first one wins, that would mean that if an artist produced something, someone else makes an NFT even though they don’t actually own the art, then sometime later the artist wants to create their NFT, they’d be shit out of luck.
- throw_nbvc1234 5y agoExample solution: You assume that the UI/marketplace you're viewing the NFT in doesn't do this for you. If something is sufficiently famous (ie the artist is well know to the UI) it's trivial to validated that an NFT is made by one of those artists and have the equivalent of a twitter blue checkmark. For stuff that's not sufficiently famous, then the question of duplicates is probably not a real problem. Who cares if someone copies a <$1 digital asset. And if the artist ever becomes famous and is back-filled into the UI's known list, then all their previous works gain the blue checkmark. These UI's will probably have a few centralized winners (which is why VC's fund crpyto startups) but there can be a long-tail of decentralized UI's as well. How much you care about authenticity will guide you to which UI's your willing to use.