Hacker News new | past | comments | ask | show | jobs | submit login

Stores aren't a great example.

Let's take houses, where the buyer and seller both need a title company to validate and transfer ownership. This process is extremely inefficent, and costly. It's pretty reasonable to imagine a public ledger replacing this process.

The real world enforcement of such thing is really the only hurdle.

Back to the point of this article: if someone steals my private key, do they now own my house?

What about liens? Will I as the owner agree to let an unpaid contractor put a lien against my title? Or can they somehow do it without my consent? If so, how do you stop bad actors from putting liens on everything?

The court is the authority on this. Title companies provide the valuable service of ensuring the state of the world matches what the seller claims.

We don’t need blockchain for this, a regular DB is sufficient.

> Back to the point of this article: if someone steals my private key, do they now own my house?

> We don’t need blockchain for this, a regular DB is sufficient.

If someone makes you sign over your car title to them under duress, do they then own your car? In most jurisdictions (let's use the USA for now), of course not. Assuming you have reasonable access to fair treatment by the legal system, that transfer would not be considered valid, and in fact the whole act would likely be a very serious crime. This is true regardless of how the ownership is officially represented: whether a blockchain, or a government database, or a pink slip with names written on it, etc.

Exactly, which is why an immutable record isn't as great as it sounds for these purposes. If everything is moved to the blockchain but that blockchain requires the exact same government override access to handle common problems, what's the point?

In a hypothetical blockchain ownership society, we'd probably end up paying extra insurance on our ownership as well as extra fees to a 3rd party to retain custody of our ownership tokens.

Insurance scams would be a laugh too.

"All my assets got transferred to a random address! You should probably reimburse me because I was clearly hacked."

If the block chain is not the source of truth, and an external body can enforce changes on it, how is it in any way better than a database under control of the state?

Sufficiently powerful external bodies can always use force to do things regardless of what any particular database says. This isn't a problem with any particular type of database.

Exactly, so what is gained by switching to blockchain if it can't solve this issue?

The point of blockchain isn't that the government can't steal your house; the (well, a) point of blockchain is that the government can't pretend you never owned your house in the first place. It establishes reliable records; that doesn't magically give it the ability to enforce those records.

It's absurd to expect any software change to somehow physically prevent people from taking your things. This "vulnerability" isn't worth talking about when evaluating software.

It is almost impossible for someone to physically steal my house. Certainly improbable. And it would be easy to recover with the help of the police.

It is by comparison easy for someone to steal crypto keys, or for them to be lost.

If we add a solution that allows us to fix the stolen or lost keys issue, then we eliminate the primary values of the block chain, decentralized trust and immutability.

So now we’re using a really complex database (potentially with POW energy use) for something that seems equally well solved by simpler tech.

So why would we want to use the block chain for this?

Right, so just use a normal database then. Shard and/or replicate it if you must. It's orders of magnitude more energy efficient, easier to mutate, maintain, and upgrade, and seems to solve the problem at least as adequately.

So the eternal question remains, what problem is blockchain solving here?

...Lack of blockchain adoption?

Enforcement and source of truth are separate entities/problems.

The blockchain can be the source of truth, but that truth might not be enforced.

The difference between it, and a centrally controlled entity is the truth cannot be altered (at least in an easy way) on a blockchain.

First of all, blockchain isn't the only technology that prevents alteration. any regular database can be run in append-only mode and can be published.

The only difference between blockchain and any other database in append-only mode is that there is no access controls or identity verification on the blockchain, so fraud is way easier (and that's why over $1 billion in crypto is stolen every year)

"no access controls" -- trustless.

Pretty big difference.

If it's officially represented by a blockchain, there's no power on earth capable of fixing it. Even if the government came and threw them in jail and confiscated their accounts, you couldn't get legal title to your car back.

Unless you include a backdoor in the blockchain, but that would make the whole thing moot. You could issue a new title and mark the old one as legally invalid, but again, that leaves you open to all the problems that the blockchain was supposed to solve.

There can be mechanisims for recovery. Social recovery wallets, multi-sign wallets, DAOs all can be part of the solution to loss of keys.

The question really is: is this all necessary? And are inefficencies/issues/costs with current centralized gov systems worth replacing?

There are undoubtedly technical solutions for resyncing the block chain to reality. But once you need them, you’re effectively saying the block chain is not the source of truth, and it’s governed by the same inefficient system that managed titles before.

At this point it is effectively a complicated solution to a simple problem - record all bits about property ownership in a publicly available place. It is unclear to me what properties the block chain has that makes it superior to a database with backups.

So I would argue that it’s not necessary and makes a complicated system more complicated, not less so.

What would really help is digitization of the records, which is a political and bureaucratic problem more than a technical one. There are undoubtedly better/worse implementations, but a block chain solution feels like it’s probably on the “worse” end of the spectrum.

It seems the unstated proposal here is that cryptography will reduce the burden necessary to validate a title. Maybe.

Let's say the house is in a town that has 100 houses and the current title verification process is that you go to the town hall, which is open from noon to 2PM on Thursdays and every other Tuesday, and you ask Ethyl, who is 71 and has cataracts, to go check the title and verify the owner is who you say it is. She takes the title document from you and then goes to an original master and checks the names. There was a flood in the town hall, so houses 21-40 had their original documents damaged in the box but there's an attestation process that applies in that case. Also, a bylaw requires that Ethyl only deals with a licensed title verification agent, of which there is one in the town, and he doesn't work Tuesdays.

This is burdensome! Both Ethyl and the verification agent are probably rent seekers!

It seems like your pitch is "if instead of this, the town webpage had a cryptographically signed register that simply notes who owns everything, then you could just go on the webpage and verify it yourself."

That would be simpler! But not because of blockchain. It's also the case that in a non-blockchain world, the town could put the ownership on a traditional database online, and give owners PINs that can be used to do the title verification. Or just list all the owners publicly by name and make the town webmaster the central trusted authority that the changes are not fraudulent. Or they could even make the ledger cryptographically signed exactly like a blockchain but with the contents centrally stored rather than distributed/public.

So it's still hard for me to think how you can separate the blockchain pitch "processes would be more efficient if we designed them to be efficient", blockchain or not.

Other problem is ownership transfer isn't even necessarily the bottleneck to liquidity of real estate. Most real estate is still occupied by people using the property for some reason, either for a place of business or because the live there. Moving their business or family to a new piece of land and someone else onto theirs requires changing school districts, updating bank records, work records, canceling utilities, physically moving all of the objects, packing, unpacking. All of that is much more cumbersome than transferring the legal ownership rights to the lot. Housing stock will never have high liquidity regardless of how you record ownership. If you want liquidity while exposing yourself to real estate as an investment vehicle, that's why REITs, MBSs, and CDOs exist. You don't want to trade physical houses.

While I agree real estate transactions are ripe for innovation, the problem is that real estate law and title can be complicated, with legal encumbrances on a property and title that are not digitized in any way. Some (such as traditional rights of access) may not even be written down. A title company is supposed to help with all that.

Making these sorts of encumbarances be written to a digitized ledger to have legal would definitely be a good innovation, but a blockchain solution isn't necessary or even desirable for a ledger. A government run database is just fine.

right, but there's no reason for that ledger to be on the blockchain - since property rights are always enforced by the government we should just have a public ledger that is akin to the ledger the gov records about driver's licenses.

Incidentally, the organization which keeps a central record of personal and commercial drivers' licenses in the United States (AAMVA, American Association of Motor Vehicle Administrators) is nominally non-governmental, although it works closely with state and federal jurisdictions as well as with 3rd parties. And it gets by fine with a conventional DB, although they have shown some interest in blockchain tech IIRC

A centralized goverment is one mechanism of enforcement yes.

No reason there can't be others. One that doesn't require blind faith/trust.

Arguably police are decentralized mechanisms of enforcement, no reason to formalize that responsibilty in a DAO.

It still requires blind faith and trust, but now you're trusting a combination of computer code that cannot apply common sense to transactions, and human programmers and bad actors, who can through accident or malice do a great job of fucking up your day.

And, since the government still would retain rights to seizing property under eminent domain or force the sale of a property for other reasons... you'd still need to put your blind faith and trust in the government to not seize your property. Essentially you've just increased the numbers of actors you need to all explicitly trust.

In what way is "these papers signed by these people in person, countersigned by notaries, postmarked on this date, after a thorough identity and title verification process conducted by these people, copied and hand delivered to bank, realtor, broker, seller, buyer, city, and state" blind faith

But "I found a message on the internet with a secret number attached" is not blind faith?

How do you enforce ownership without the state? If I take your house from you, what incentive do I have to care what the blockchain says?

I can already picture cities full of forever inhabitable houses because somebody died unexpectedly or lost his private key.

I get to your last line and all I can think: "Wow, it's so simple, implementation is the only detail we haven't figured out yet."

Just because there isn't a solution/implementation currently, doesn't mean there won't be.

No, real estate is even worse.

Surely you need a backdoor for courts to fix problems? And if you have a backdoor, then what's the point of the whole fucking blockchain?

I say this having bought a house, and knowing the BS paperwork and costs involved.

You don't need blockchain to establish such a system. Australia has been using the Torrens title system [1] since the 1800s, which is essentially a (semi) public register of titles and transfers.

[1] https://en.wikipedia.org/wiki/Torrens_title

New Zealand too, and it really is as good as it sounds. Transacting property can be done very quickly, safely and cheaply, and no title insurance is required.

The register is of course just a normal relational database.

One key to the success, though, is a large and mature body of legislation and processes to ensure that the register really is always correct.

Government run public digital database might be much simpler and efficient solution. No point involving blockchain for something where it is not needed.

How would you validate that the transactions related to the house on the blockchain is correct? What do you do if you inherit the house from someone who neglected to leave you the private key? You can come up with solutions to all the problems and you’ll have reinvented the entire existing process, which is inefficient and costly for a reason.

The process of transferring real estate ownership is messy and expensive for all kinds of reasons that aren't solved by a distributed append-only database.

If it is unclear to start with which titles are valid, then filling the ledger will be an issue, i.e. the effort is in creating a clean register and not with the register "mechanics".

Btw. there are also countries with very well maintained registers where this is not a problem at all at present.

Applications are open for YC Summer 2023

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact