HN Theater @HNTheaterMonth

Oh damn, that's a long comment, let me try to answer point by point.

> The core things that blockchain stuff tries to solve are not actually the hard problems.

What are the hard problems though? Depending on where you stand you might see different things, but to me the hard problems are that banks throughout the world have a hard time trusting and interoperating, things are slow (payments have to go through banks correspondences and central banks and the BIS and that route takes a long time even if some of the stages are RTGSs), errors are way too common (basically audits happen all the time, are manual, and consequences of errors are real issues that people have to deal with).

> The hard part of online commerce is not that customers can reverse credit card transactions. The hard part of contract law is not automatic payments. The hard part about property law is not atomic transactions of money and deed transfers.

What are the hard parts from your point of view? Is your point of view that the system is perfect and doesn't need the technical improvements proposed by blockchain? I'm not sure I follow.

> It's not trivial to track money, but the stated goals of cryptocurrencies is to make it hard or impossible to trace, and to compel.

I feel like that's too much of a strong statement considering that some cryptocurrencies, which I've worked on, were built specifically to follow regulations and make fraud hard.

> The explicit goals of cryptocurrencies is to get around AML & KYC laws. This is what Pintrest was talking about.

Again, there are many cryptocurrencies with very different goals and this statement can't apply to all. The statement is at least very wrong for some well-known cryptocurrencies.

> We as a society, as a people, don't even want people to be able to send money any way they want. E.g. we don't want people to be able to send money to ISIS. There's a "baby and bathwater" argument that can be made here, but since cryptocurrency people don't even WANT to prevent this use case, it's not what they're making, or selling.

So, ignoring what I said previously that different cryptocurrencies have different goal, and so that statement doesn't apply to the field in general, there's still something to be said about that. How much privacy do we deserve? Where is the line where privacy is too strong that it is damaging to society? There's a very similar argument with end-to-end encryption in messaging applications and governments trying to fight it. For example: https://www.gov.uk/government/publications/international-sta...

> Like airline security, it's not effective because it's perfect, but because there's too much risk that one of the layers will catch you.

BTW I don't think airline security is a good example. Adam ruins everything has a good episode on that: https://www.youtube.com/watch?v=-LDzOi1dyAA (tl;dw: it's security theater)

> Reversibility: People don't want this. One of the main reasons to use a credit card AT ALL is that people want reversibility. If that's all you wanted then you can probably just accept debit cards.

Credit cards are mostly a US thing btw, I never knew about them before moving here, and even in the US reversibility is not always a thing. For example, you can't reverse transactions done on venmo or zelle. But even then, cryptocurrencies don't preclude reversing transactions, there's nothing in the technology that would prevent that, especially if it used as a settlement layer for virtual assert service providers (exchanges, banks, etc.)

> And as for smart contracts: All smart contract systems need a way for a court to interpret the contract. The contract may not be legal, it may have been signed under duress, etc.. So all smart contract systems need this escape hatch.

I'm not sure I understand your point. A smart contract has its code published in clear so by using it you already choose to trust and obey the code.

> The most ridiculous one is "land ownership registry". You know who ultimately decides who own a piece of land? The government of the country it's in.

Agree, linking real objects to the blockchain is hard, and projects that tout that usecase are often scams.

> And note that all this is in addition to technical problems like PoW/PoS, electricity use, scalability, transaction speed, 51% attacks, etc... Even assuming all those are solved, the problem is the the very GOAL is a dystopia.

PoW/electricity use are things of the past for the technology. No new cryptocurrency rely on these. Scalability and transaction speed are things that are being solved (I believe Algorand is aiming for 20k transaction/s for the end of the year). There are other consensusless protocols that pretty much scale linearly based on the number of machines you throw on the problem (see fastpay or At2).

BTW if you have any question, happy to continue this conversation or clarify anything.

⬐ knorker

> What are the hard problems though?

I mentioned this. E.g. "The hard part is codifying your intention into legalese, and conflict resolution".

But again, I would invite you to talk to some contract lawyers and see what the hard problems are in their space. And carefully listen to how they solve them, and if your cryptosolution maybe not only doesn't solve the problem, but actually makes it worse.

Especially internationally, where you maybe hire a service, but then one of the countries involved has export restrictions, or immigration requirements, or whatever else.

These blockchain things are explicitly made to "go around" all government regulations (that's why they can say it'll be cheaper and more efficient). But the companies still have to obey the law, or they'll go to jail. So no that's not the problem.

> Is your point of view that the system is perfect and doesn't need the technical improvements proposed by blockchain?

I have not seen a solution based on blockchain that solves a hard problem. Nor have I seen one that doesn't make a hard problem even harder.

E.g. when I bought a house I had to prove to the bank and various involved parties where the money came from. And then when transferring I had two people from the bank walk me through it, to make sure this 6 digit cash transfer was legit.

It was annoying, yes. But I understand why it's there. To say "oh if this had been bitcoin you could just have transferred and it would be there in an hour". Or "had this been a smart contract it would have triggered deed transfer". Ok, sure. But what if there were some legal dispute and the seller was about to lose the house? That I would essentially be buying stolen goods?

What if the house had known deficiencies that the seller provably knew about but lied about during selling? If this had been Bitcoin they could simply decline to pay me. They'd launder it in Monero, and sucks to be me.

It's easy to set yourself up to be judgement proof if your assets cannot be seized.

> Credit cards are mostly a US thing btw, I never knew about them before moving here, and even in the US reversibility is not always a thing

I've lived in two countries, neither of which are the US, and credit card extra protection is certainly not US-only.

But that was still not my point. I'm saying reversibility is a feature. That mainstream finance can deliver. And it's good, so cryptocurrencies should too. They mostly do the opposite on purpose, which shows that they are not actually making what people actually want in a currency.

> Again, there are many cryptocurrencies with very different goals and this statement can't apply to all. The statement is at least very wrong for some well-known cryptocurrencies.

zcash, I believe, makes lip service to AML/KYC, but to me as a non-ML-investigator it smells a bit loophole-y, and one that may close.

> BTW I don't think airline security is a good example.

Adam ruins everything (and I've seen it) is entertaining and informative and all, but while yes there are aspects that are security theatre, turns out many of them do add up. The hijackings of the 70s largely went away. But also, Adam's not even saying airline security is security theatre! He even enumerates four added security features that DO stop terrorists. He's talking about the TSA. And then Bruce Schneier says more relevant things.

So the sum of airline security is: good old fashioned police work, infiltration of organizations, wiretaps, anomaly detection, behavior checking (Israel does a lot of this, and the have a good track record despite the obvious targeting), TSA, passengers, air marshals, reinforced cockpit doors, etc..

So if you want to be successful, then you have to get past ALL that.

And while the TSA doesn't do much, any suicidal idiot could just get on a plane with a fullauto in his carryon. At least now he has to weigh the risk of being embarrassedly found in the security check.

So maybe he probably switches to a handgun instead of the AK47. So he's less dangerous.

> cryptocurrencies don't preclude reversing transactions, there's nothing in the technology that would prevent that

Sure. But every time someone tries to sell the benefits of cryptocurrencies they will list this as a feature. And again this is my core point: All of the features they list, are actually bugs. And if you invent a cryptocurrency without the bugs, then it doesn't actually solve anything for anyone.

> I'm not sure I understand your point. A smart contract has its code published in clear so by using it you already choose to trust and obey the code.

It's not your choice to obey an illegal contract. And it shouldn't be. That's the point of contract legality.

And the hard part of writing a contract is not the enforcement, but codifying your inner intent into a bugfree contract, signed in a legal manner.

If you hold a gun to my head and force me to sign a smart contract, and I can prove it, how can a court invalidate the contract if you refuse to sign the invalidation block?

> PoW/electricity use are things of the past for the technology

And the present. I won't celebrate prematurely.