Hacker News Reader

> How would an NFT be better than the above database in any way?

It wouldn't depend on the benevolence of a third-party absorbing the cost of running the index.

It would be geographically and culturally agnostic. Zoom out here and notice how little of the World's population is involved: https://map.bikewise.org/#0/84/-77

It would be globally unique and would not require the cooperation of manufacturers. Neither of my two bikes has a serial number because they're 'too old'. Other bikes have conflicting serial numbers.

It would eliminate the privacy and security risks of maintaining a central database of PII and associated valuable assets.

> It would be geographically and culturally agnostic

Who's to say everyone is using the same NFTs for this? We already see a variety of different platforms for NFT art.

> It would eliminate the privacy and security risks of maintaining a central database of PII and associated valuable assets.

How are there now fewer privacy and security risks when the entire blockchain is completely public? Where's the privacy?

I still am confused by this though. If the police/government don't acknowledge this then you can't really get it back without a fight.

Wouldn't you rather just have it run by an entity the legal system can point to?

If you don't trust the legal system...then I really doubt that having a third-party website is going to help you with a bike thief in anyway.

No instead it would foist that externality on the entire world in the form of carbon emissions and depend on the highly uncertain future value of a cryptocurrency.

And it would still require a third party to benevolently run an app to search and make sense of the data from the ledger, and also to verify that the data inserted was actually real and not poisoning the db.

How would you prove to a blockchain that a particular serial number really is owned by you, or was lost, or all the other things that can happen that don’t occur on the blockchain? Aren’t you sort of back to needing the cooperation of lots of different entities globally, and some kind of authority equivalent to the third party database maintainer to ensure the data makes sense?

An NFT is inherently decentralized, maintained by multiple independent parties, and more or less tamper-proof. That database is only as resilient as the organisation running it (or maybe less, plenty of organisations have lost their databases before), and presumably there are corruptible humans with write access to it.

The certificate transparency project has the same attributes you mention to my knowledge.

There is the problem of correct data, but I don't understand how an NFT manages to make data correct. How do I prove to the NFT that I own the bike with SN 123? How do I prove it was stolen?

I'd love if we replaced the bikeindex with a certificate transparency project-esque system and each large bike manufacturer ran an instance.

That also seems like it would be vastly easier to create than having people use the bike-stolen-registry-NFT you're talking about, and as a bonus it would be far less wasteful.

It seems like it would optimally require the manufacturer to participate in creating a token representing that serial number and assigning it to the person who bought the bike who would have to transfer it to the next owner and so forth.

And so when I sell my bicycle to someone (and write them a receipt) but forget to do the weird token transfer thing, who owns the bike?

In the eyes of the law, the guy holding the bike and the receipt owns it. What the NFT says is irrelevant.

So how are NFTs useful again?

If they had been out long enough to be pervasive you could require all online marketplaces ebay,offer up, facebooks etc which are commonly used to traffic stolen goods to upload an NFT as part of the transfer whose description matches the item being sold which is transferred with the good.

It would be like requiring every good sold to have a trivially electronically verifiably accurate original receipt that only someone who acquired it illegitimately would lack.

Certificates still have the problem of who your trusted root is, and don't (by themselves) solve provenance issues - if two people have certificates for the same bicycle then there's no way to choose between them, whereas an NFT inherently has an auditable chain of custody going all the way back to the original issuer (presumably the bike manufacturer in this case) and digitally signed at each stage.

I'd love it if these industries/communities adopted log-structured distributed datastores with digital signing to track provenance of things without wasteful "mining". But the fact is that they've shown no willingness to do so, and it's a lot harder to bootstrap a system like this - if your system relies on a trusted root then you need the trust before you can start to develop the system at all, but users have no reason to trust the system until it's all fully running and proven.

It's another example of the title problem. You can think of it in a similar to how VINs work for cars, or ESNs work for phones. In those cases, there's a centralized database that tracks whether things are stolen or not. It's possible that NFTs could be a decent way to handle supply chain auditing of this variety, considering it's just rebirthing a process that already exists but in a more analogue form.

The advantage over something like bikeindex is that you could make the transfer process a lot more seamless. You could build things like escrowing into the protocol, so when someone sells a bike to you on craigslist, you have a nice clean transactional system which adds some nice anti-fraud measures and protections for you without necessarily requiring you to buy it from a store or marketplace to gain those protections.