Hacker News Reader

Comment by kyoob | Hacker News Reader

Good on Tyler for pressing on the advantages of Web 3.0 for artists. I didn't really find MA's answers satisfactory.

There are few actual good use cases for NFTs in ways that actually matter, buuuut... even in the case that a highly talented but unknown artist hits it big one day, there's nothing that is going to prevent the current owner from selling the art without going back to the NFT to compensate the artist. That's a pipe dream outside of the purely digital space.

Why do people always use art as an example? Couldn't you do it with something like cars (where the car doesn't start until you prove you have the NFT)? I'm completely unfamiliar the with the space

Would the NFT-locked car require a live blockchain connection and refuse to open its doors if it hasn’t been able to sync its node after a certain time? So you’d get locked out of your car if mobile Internet goes down?

If it doesn’t do this, then you can steal the car by selling the NFT, simultaneously disabling the car’s Internet connection, and driving off. Now the NFT is held by one person but the physical car is held by another.

Would likely require a form of NFT that's a little different and probably more complex than the current forn. Something about having the car key also be a cryptographic key, and transfers of the NFT invalidating keys or requiring rotation.

What is it with NFTs and solving theoretical problems in extremely complex ways. Car thefts at at 1/3 of what they were in the 90s, and the existing mechanisms for buying/selling cars work just fine. Why do we need to introduce car DRM with brittle, distributed access control systems? You're more likely to get your tires or cat stolen - no NFT is going to prevent that.

because then a16z can get a cut of it

You could shape the NFT in the form of a metal rod with notches on it. When placed inside a slot and turned, the car could validate that the right pattern of notches exist on the NFT and turn on.

If you want to sell the car, simply hand off the physical NFT to the new owner.

Well crypto allows for trustless digital transfers. Like selling a house halfway across the world, without having to fly over to hand off the keys.

So the person halfway around the world can immediately start enjoying their house?

I assume it would be something like, the door lock is tied to the current owner as defined by the blockchain. No need for intermediaries, and no physical key.

So we're back to the original question: without an Internet connection, this blockchain-controlled door won't open, even if you're the legitimate owner.

Or if it does open without Internet, then you can buy the house NFT, but the seller disabled the Internet connection in the house so the door doesn't know about the transaction. Now you're locked out even though you theoretically own the house on the blockchain.

Turns out the blockchain doesn't solve ownership of physical objects. You still need intermediaries like the police and courts to decide who owns the house.

Or perhaps it's looking towards a future where internet access is more reliable, as it has been getting for the past few decades

Do you really want to be locked out of your house or car if the Internet is down?

We've had the electrical power grid for over a century, but power cuts and brownouts are still a thing.

Designing vital systems to assume 100% reliability of shared systems is dangerous.

I agree. It's not only that it's dangerous (let's say the internet always worked, someone could still ship a bug to the protocol, or too many nodes could fail, or it could get hacked), it's just not even better. Technologists often seem to not start with problems. What is the problem being solved by a blockchain based lock? What pain point exists that would no longer exist? Having to carry keys?

Keys are already decentralized and fully controlled by me. I can copy it and distribute it freely. Putting an intermediary (a blockchain) between myself and access to my house just seems silly.

Perhaps keys are valid for 24 hours. So if internet goes down, as long as you were validated sometime within the past 24 hours, you can enter. This also means that it takes 24 hours for ownership to change, which should be fine when selling a house. This also gives the city 24 hours to deploy countermeasures (like starlink receivers) in case the internet outage is expected to last longer

This is the classic definition of a solution looking for a problem though.

I don't know about "no need for intermediaries" because of course you need someone who creates and maintains this blockchain, since blockchains are not present in nature (perhaps that is a decentralized group of people but the blockchain and its maintainers are then the intermediary between the two parties) so let's focus on the no physical key part.

If the primary benefit of blockchain is that only the registered owner of the house can get access to it, on that blockchain, that's a pretty weak feature and actually pretty shit. I would like my family to also get access to my house, as well as perhaps other people like houseguests or a maid etc. Keys seem to work pretty well, nobody is pounding down the door to buy "smart locks" as far as I can tell.

I'm not saying blockchain won't provide transformative utility, it might, but it probably won't look like what we imagine it will (existing paradigms, but on a blockchain).

Sorry I meant centralized intermediaries, which is what is often meant when referring to blockchain tech. Some other examples of centralized intermediaries are banks, ticket resellers like ticketmaster, deed notaries, etc.

And as for sharing keys with family members, perhaps just list them as sub-members of the owning member on the blockchain. When the owner changes, all previous sub-members lose access too. This can probably be done using a smart contract

You can certainly do it. But then the problem being solved is "keys are bad". Typically in startup lore you want a product to be 10x better than the existing solution to gain mainstream adoption. I have a hard time picturing smart locks to your doors on the blockchain as being 10x better than keys.

Yes.

But real estate law does not.

I think the point of an NFT would be that everybody else can also see that you control the token, which doesn't really seem relevant to starting a car. You could just ("just") have a privately controlled public/private key pair, and the previous owner could tell the car that your public key should now be the only one able to open the car; no need to announce that to the world.

(I also know nothing of the space, just engaging to build my own understanding.)

I clipped the 3 minutes where Tyler pressed Marc on a specific Web3 use case: https://twitter.com/liron/status/1537186589486460928

Reply on news.ycombinator.com