Hacker News Reader

Comment by mgraczyk | Hacker News Reader

Believing that the technology is worthless is just as naive as believing the space should have $2T in capital.

There are tons of interesting things you can do with cryptocurrency and blockchains. Most are financial products, but some are just easier ways to develop software. None of this justifies the insane valuations of most coins, but I have built products that would not have been possible in the pre-ethereum world.

For example, who wouldn't want a system that forces banks to track customer balances in a secure, private, and auditable data structure? This is completely possible with cryptography that was researched as a byproduct of cryptocurrency development. It would be a marginal improvement on the existing banking system, but it would provide real, practical use and would improve confidence in the banking system.

View on HN

I increasingly see this sales pitch as similar to a pyramid scheme schilling vitamin supplements.

Yes, vitamins "have some benefits", but no one wants them. If they cured cancer, sure. But they dont.

Likewise, sure, an international distributed ledger that can compute about as much as a school calculator "has some benefits", but it doesnt cure anything. Least of all "international finance".

Crypto-holders (ie., people in the pyramid scheme) want the debate to be about the technology... so you can't quite see the structure of whats going on.

ie., that there's no demand for the product. No one wants your 100MW blockchain calculator

You're confused about the difference between "something having value" and "something being revolutionary and universally transformative".

People buy vitamins, that is a fact. From walgreens or CVS. I do not buy vitamins, but some people do. I know that most people buying vitamins do not benefit from them. Does it follow that vitamins "have no value"? No, they just have less value than the currently market would suggest.

I buy vitamins. It’s because of a vitamin deficiency identified by my yearly doctor visit. When I am good about taking them, my numbers come back good. When I am not, she reports a deficiency again.

I don’t guarantee my life would be worse if I continued my vitamin D deficiency, but my doctor suggested I fix it and CVS sells me the fix. Certainly not a useless product until the scientists tell me my doctor is a quack.

I'm referring to the marketing of MLM companies around supplements which do heavily imply (or claim) that they have often dramatic and curative benefits.

Customers dont believe this, they dont want supplements from MLMs. But dupes believe it enough that, at least, they buy a warehouse full and start shilling; or otherwise, think that enough other people can be duped.

Now, since there's no actual demand, these dupes are forced to sell to other dupes just the prospect of further selling, not the actual product for its own use. Since there's no actual market, this whole thing collapses ("when the dupes run out") and 95% of people lose money with warehouses full of product they cant sell.

This is, in my view, where crypto is heading. I believe within the year there'll be a lot of impoverished people with wallets full of useless coins: though, of course, we're already there in many cases.

I think this is patently obvious to the 1% of crypto holders who own and run the crypto market. It's a scam.

This has already happened 3 times, and it will keep happening. The dynamic is more complicated than what you're describing.

Crypto isn't a big MLM. Crypto is a family of technologies that make it possible to easily build MLMs.

The analogous thing in traditional MLM is the internet or mail. The difference is that the internet is much more valuable as a technology than crypto. That doesn't mean crypto has 0 value though, it's just that the internet is one of the most valuable things humans have created.

Difference being vitamins can be produced in a profitable way despite limited value. Bitcoin, by design, can not.

What does that even mean? Plenty of people have profited off of bitcoin.

> There are tons of interesting things you can do with cryptocurrency and blockchains.

Interesting? Sure. Useful, economically viable and slowing real problems? After tons of companies spending many billions of dollars in RD, we still have none.

> For example, who wouldn't want a system that forces banks to track customer balances in a secure, private, and auditable data structure?

No one, expect for cryptobros. Tracking balances and solving discrepancies is long solved problem. And it’s not a technical problem. I don’t need super wasteful DB to do that, I need law, procedures and customer service.

The hard part isn't tracking balances. It's reporting, auditing, and security/privacy.

We need laws and procedures, but the barrier to getting operationally intensive things shipped becomes lower with better technology and particularly with standardization. One of the key benefits, if not the main benefit, of blockchain systems is that it forced all the players to agree to a highly detailed and easily testable specification (otherwise their nodes fork). That is an enabling design pattern for inter-bank communication.

I believe the main reason it hasn't materialized is just what you say though. It's not that valuable. But it's clearly not worthless.

> the barrier to getting operationally intensive things shipped becomes lower with better technology and particularly with standardization

100% agree.

> One of the key benefits, if not the main benefit, of blockchain systems is that it forced all the players to agree to a highly detailed and easily testable specification (otherwise their nodes fork)

There’s a lot of, way cheaper and convenient ways of enforcing it.

I disagree, there are not cheaper or more convenient ways. There is overwhelming evidence that building highly standardized financial infrastructure is much cheaper and easier with blockchains than doing it without blockchains.

You can argue that the result of that is mostly scams, because building scams is easy. But clearly crypto made building financial infrastructure cheaper and more convenient. There was absolutely nothing in the world like Uniswap for example, and we were not on a trajectory for something like that to ever be built in traditional finance. If you don't believe that, then I encourage you to look closer at Uniswap. Try to think about it from a purely technical perspective, without the understandable bias earned from its usage for scams, pump and dumps, etc.

Another good analogy is nuclear weapons. Purely as a weapons technology, they made warfare cheaper and easier. The end result of the existence of nuclear weapons may be the opposite (countries spending more to keep up with MAD). But as a technology, they clearly enable specific things that weren't possible before (blowing up cities). Crypto is like that.

> I need law

Those bills can get big. On-chain is way cheaper.

Laws are big and complex for 2 reasons: 1. Corruption 2. Most importantly, because real world is complex

Blockchain doesn’t make world simpler. And “cheaper” is just laughable, as whole idea of blockchain and its security is about being as expensive as possible.

>For example, who wouldn't want a system that forces banks to track customer balances in a secure, private, and auditable data structure?

Crypto doesn't force banks to do that.

Also, if it was a real problem, there would be legislation for it since RSA became a thing.

And if it was, having a 3rd-party timestamping server and regular cryptographic signatures would be enough to solve it. No bank in their right mind would burn terawatts of power to prevent an error that has a 0.00001% chance of happening and costs virtually nothing to fix.

Technology from the cryptocurrency world can be used to force banks to do that.

RSA alone isn't enough. How do you use RSA to build this? You can do auditable sums (prove that private balances sum to a public total), but I can't think of a way to do things like proving that a particular debit happened.

> costs virtually nothing to fix.

There is very high system-wide cost to the status quo. Part of the reason that bank errors are so inexpensive to fix is because transfers take 2-5 days.

>I can't think of a way to do things like proving that a particular debit happened.

Prove to whom? Banks have multiple layers of redundancy (including internal logs, paper/PDF statements, etc). So in case of any software error or a single-employee malice, it will be just fine.

>Part of the reason that bank errors are so inexpensive to fix is because transfers take 2-5 days.

That's a feature, not a bug. It gives you time to stop a transaction initiated by someone who stole your identity. With blockchain, you lose keys - you lose coins.

It doesn't have to be that way though. There's nothing inherent about the timescale of 2-5 days that prevents the same level of security and reversibility with shorter transactions. The problem is the lack of coherence, lack of information, and lack of fast technology. It's easy to build a reversible version of BTC with programmable timeouts for reversal, and the timeouts can then be tuned based on transaction size, risk tolerance of the sender, and willingness of a third party to underwrite the transaction risk.

Credit cards basically do this. Part of the reason they are able to is because they gather more data and rely on more sophisticated technology than the banking system. With blockchains, you have another small tool to improve the situation.

> cryptography that was researched as a byproduct of cryptocurrency development

Oh please. What cryptographic advances were made possible by cryptocurrency? Name one. Merkle trees and hash chains aren't exactly rocket science and they predate cryptocurrency by decades.

My understanding is that the SNARK/STARK and VDF mathematicians are being paid by cryptobros.

Not going to defend cryptocurrencies, but there's been huge explosion in formal methods research funded by the money chasing smart contracts and the like.

VDFs, zk-STARK/SNARK, Verkle trees, coinjoin, taproot, ... There are hundreds and some are already used outside of cryptocurrencies. In 20 years many of them will be. As a person with a background in distributed systems I'm particularly excited about VDFs and compact cryptographic datastructures because they made developing secure distributed systems much easier.

Genuinely interested in what products you've developed that weren't possible pre-crypto.

I built an ICO platform that allowed companies to raise funds and I did not escrow the money or get approval from any financial institution. We operated outside the US. It would have been impossible to do anything remotely similar in terms of UX without crypto. Best alternative would have cost $XX millions to set up and would require users to wire money to an escrow entity. Nobody would have used that. We built ours with 2 engineers and very little capital in less than a month.

So, you’ve built a platform to help scam people (that’s what ICO were) in a way that doesn’t involve financial institutions (that give people at least some protection).

Yeah, that’s what current uses of blockchain are all about.

Regardless of where you stand on the value of financial products, undeniably the technology was enabling. I'm not a luddite, so I put a nonzero value on all new enabling technology even if in practice the technology is currently not used productively.

I’m all for selling shovels during gold rush. But I’m less into selling guns during gang wars.

> that doesn’t involve financial institutions (that give people at least some protection).

What protection? They take the wire. Charge fees. That’s pretty much it. Corp clients don’t have Reg E.

Your UX is easier because you punted the hard part to the crypto on-ramps.

Yeah that's the point.

But this is sort of like saying Moderna has it easy because they punted the hard part of deploying an mRNA vaccine to global shipping and distribution chains that they did not build.

Is that an actual problem that needs solving? How often do real banks make errors in customer account balances? That's never happened to me or anyone I know.

One time, in 1995, Glendale Federal made an error for $-3200 on my account. Confused me with a similar named customer in Turlock. Took two weeks to settle out and another two weeks to reverse the overdraft fees. $3200 was a lot of money for me at the time.

Edit: error rate of 1 per 20 years seems very good to me.

Don’t many banks reorder transactions to collect overdraft fees and such?

And crypto miners front-run or reorder transactions to collect MEV.

Reply on news.ycombinator.com