[AlexandrB]

> If I have a fake BAYC, OpenSea can determine that very quickly and take it done, regardless of how many intermediate off-platform transactions there have been.

What qualifies as "fake"? How many modifications would I have to make to the original images for them to count as new art? Because this is a fundamental question in copyright law RE: derivative works and NFTs don't seem to help with this at all.

[berberous]

This is again a black and white answer.

There are many fraudsters who, quite literally, make fake NFT that look exactly the same. The decentralized chain of title is very useful to quickly and definitely distinguish the originals from the fakes.

Just because blockchain cannot solve all problems, does not mean it has not solved the above problem.

I agree that an artist could come along and make a derivative artwork that is right on the line of what is infringing, and in that case, you would solve it with traditional methods, e.g. an IP lawsuit for a judge to decide. Nobody said NFTs can solve all IP disputes.

But they can definitely determine the provenance of an item, even after a complex round of intermediate off-platform secondary transactions.

[BobbyJo]

This completely glosses over the fact that this only works for things on-chain. So you own an NFT, what does that actually mean? Does it mean you verifiably own the thing the NFT points to? No. It simply means you verifiably own the pointer. The link between the pointer and the off-chain thing is still completely vulnerable and unverifiable.

I think the real world equivalent is having a signed legal contract, but no courts or police. Feel free to buy and sell the contract, as I'm sure that has value as art in and of itself (which is how I see NFTs). But if you wish to lay claim to the items under contract, glhf.

[berberous]

I think this is a problem that’s easier to solve than you think.

Let’s say an artist creates a minting process, which contains a click through contract granting you legal rights to the image purchased, the right to make derivatives, etc., which rights transfer always to the holder of the NFT.

You now have a legal contract granting the NFT holder all IP license rights to that image. If there are 300 intermediate secondary transactions, and then I buy the NFT, make a derivative to exploit commercially, and get sued, my defense in court is simple: (1) here’s the governing IP license terms and (2) I own the NFT which is easily proved on chain. That IP license, by the way, could even be embedded on chain.

I can think of lots of examples like the above, where you can leverage some offchain system (like an IP license) with an on chain one, so that you have real world rights to an item but also benefit from easy and auditable transferability.

Again, just because NFTs do not solve all problems (eg you still need an IP license) does not mean that combining the two is not a real and useful improvement.

[BobbyJo]

> Again, just because NFTs do not solve all problems (eg you still need an IP license) does not mean that combining the two is not a real and useful improvement.

Totally agree. I think that's where the divide lies with tech savvy people and why HN tends to be anti-crypto. It is indeed a useful piece of a stack for solving a problem, but honestly it's a pretty minor piece. A majority of the cost and attention that goes into IP is enforcement, not contract tracing. It's definitely a part of the problem, especially for smaller players, but it's not the whole pie that the less tech savvy seem to think it is.

[Joeboy]

I'm not sure if I'm understanding you or not.

I can see that if there's an NFT minted in eg 2020 and no record of the associated work existing prior to that, that provides a bit of evidence in favour of the (earliest) NFT holder being the legitimate IP owner. Is that what you're saying, or something more than that?

[berberous]

That’s basically it.

For a well known and successful project, like BAYC or Cryptopunks, the authentic and original mint contracts are well known. If in 100 years someone offers to sell me one, which has traded hands 3500 times, it is extremely easy to ignore the noise of the 3500 intermediate transactions, pull up the source contract, and confirm it’s the authentic original.

This of course works best for “well known” NFTs, but it is a legitimate improvement over how Sotheby’s would have to authenticate a newly discovered alleged Monet, as in that case, the secondary transactions are relevant and need to be investigated as it’s impossible to just jump to the source and know with certainty that this is a legitimate Monet.

For projects that are not well known, you do not have the benefit of having a “well known” source contract, but yes, with everything on chain, you could at least do as you suggest and say, well, this appears to be the earliest example on chain anyone has been able to find.

So while there is of course still some uncertainties or some need to refer to external knowledge or sources off chain, being able to totally track a clear decentralized chain of title is very helpful, and solves a lot of problems, even though it does not solve all problems.

[Joeboy]

Thank you! My initial idea of what NFTs were was that they were like a free copyright registry, and as far as that goes, I don't really see a problem (although it seems like a rather over-engineered solution).