Bitcoin Believers, Guzzling the Flavor-Aide
I haven't mentioned it recently, or at all, but a few months ago I spoke with a Bitcoin guy. He was my first, a tecky by trade who really believed in the Bit.
I asked him what the appeal was. For him, it was to take control of money in ways that separate it from people. ( Things proceeded to weird. )
I asked him what the appeal was. For him, it was to take control of money in ways that separate it from people. ( Things proceeded to weird. )
For all I know, Bitcoin might be as un-hackable as tally sticks. That would be good. Tally sticks, though, are pretty easily created with a stick and a knife, where Bitcoin requires more sophistication just to start. And internet. And electricity.
Bitcoin probably would be easier to bootstrap back to existence than most electronic finance infrastructure....
Good to know.
Gold already exists, goldbugs are nothing new.
True, but moving gold is difficult. There are some providing physical specie security (gold in a remote vault) and secured transaction electronic transfers (people trading possession of remote vaulted gold), but these are proprietary, so the mechanism cannot be verified as to quality.
Perhaps the block chain folks could abandon their "mining" and simply provide specie transaction. This would give gold new mobility as an exchange commodity that would satisfy the security requirements of many.
Then again, how would the technically competent create value without leaving the comfort of their computer dens? Silly of me to consider. ;-)
Reply
I'm getting the possibility of hacks from dystopian sci-fi
There's some risk to generalizing from fictional examples.
any system must be easily understood to be verifiable, but any system that is so easily understood presents its security flaws
The first part is certainly true (especially if by "easily understood" you mean "by people with the technical background to understand that sort of thing"). The second part is not quite true, which is why cryptography works at all; it relies on a family of mathematical functions that are easy to understand but still provably hard to reverse.
True, but moving gold is difficult. There are some providing physical specie security (gold in a remote vault) and secured transaction electronic transfers (people trading possession of remote vaulted gold), but these are proprietary, so the mechanism cannot be verified as to quality.
My first reaction to this comment is to say that moving gold is not that difficult, relative to other valuable commodities you might use to settle transactions (if for some bizarre reason you think it's a good idea to settle transactions in commodities). That's one reason why gold is valuable. (And one reason why nations that hold these nominally-leased gold reserves still maintain the premise that it's not just a number in an account, and that they could go back to settling transactions with international bulk shipments of physical gold.)
But even given that, it's strange how well your comment highlights the advantages of Bitcoin relative to gold. Possession can be verified remotely with confidence, transfers are very quick and don't require any physical security since they can't be intercepted in transit at all. (Though of course that doesn't solve issues with fraud or theft at the endpoints. And even if your settlement can't be intercepted in transit, any physical goods you're paying for still need to be shipped.)
I'm not a goldbug, so I find the idea of settling transactions with valuable physical commodities pretty unappealing. At least, currently there's a financial system that does a reasonable job of satisfying my needs, given the sorts of transactions I want to do. But if I was a goldbug, I'd certainly note that bitcoin has many of the properties that goldbugs like about gold, and some relative advantages that are substantial.
(Or, if I was a more-typical goldbug, I wouldn't be at all interested in using gold for settlement in the present, but (depending on how apocalyptic my preferred apocalypse is) I'd fantasize either about eventually selling the gold bars in my basement to fund my retirement or about eventually exchanging the gold bars for gold coins and exchanging those gold coins for cans of beans.)
Perhaps the block chain folks could abandon their "mining" and simply provide specie transaction.
I'm not sure what you mean by this exactly. I can think of a few interesting things you could mean, but I don't want to respond to a guess.
Then again, how would the technically competent create value without leaving the comfort of their computer dens?
That's a little unfair. People try to make money with the skills and opportunities they have. I feel like criticisms of the tech crowd sometimes get into a theory of value that's a sort of Marxoluddism: The production of value originates from labor, but real labor, the sweaty kind.
I do think Bitcoin has some real use-value, maybe even comparing with the use-value of gold.
Though how much inherent value gold has is an interesting question. Generally, a bubble involves things acquiring a value not justified by the properties of those things. But gold (and bitcoin) have properties that make it easy for them to sustain a bubble: When the price gets high, it's hard to increase production, hard to counterfeit, easy to split into smaller units, easy to store and transfer. It makes it a bit of a paradox whether the high price of gold is based on its inherent value.
Reply
True; there is, however, a corresponding risk to discounting yet unrealized possibilities as merely fiction. ;-)
More seriously, think of the myriad examples of people who built a secure situation and, simply because they could not fathom a method of breach yesterday, failed to update the security for tomorrow. The Maginot Line comes to mind.
...which is why cryptography works at all....
My point in addressing the security of cryptography is to note the pace at which current computing makes all cryptography less and less secure. That's why I mentioned Gibson; given computing which is here but just not yet that accessible (quantum comes to mind), an uptick in availability would crack current security (if that hasn't already happened; remember that one Vault episode?).
I'm not sure what you mean by this exactly. I can think of a few interesting things you could mean, but I don't want to respond to a guess.
Completely coincidentally, I heard a TED talk about block chain just after posting/responding. The tech is being expanded beyond Bitcoin to decentralized money transactions, music, and a few others which escape me. (Pardon the lack of linkage; I'm at work now, using a shared and unfamiliar computer.) That talk reinforced my original notion of using the block chain not to move Bitcoins, but simply to facilitate more conventional purchases, transfers, etc.
The talk also gave me a better understanding of how robust the block chain might be.
I feel like criticisms of the tech crowd sometimes get into a theory of value that's a sort of Marxoluddism: The production of value originates from labor, but real labor, the sweaty kind.
Which is fair, isn't it? After all, we are all to some extent benefiting from an economy allowed to be more complex thanks to those gooey ergs of ancient sunlight. No propulsion to make the 3,000 mile salad possible, no tech "jobs."
(Scarequotes there to differentiate coin mining to actual, helpful debugging and tech assembly. Setting up, plugging in, and waiting for the coins to *bing!* into existence is as challenging as sitting on a property and waiting for the rent checks to clear.)
And seriously, don't get me started on goldbugs, especially the hoarding type. If the worst does happen, I plan to ignore anyone who offers me gold for anything of current value. A part of me will gleefully deny food for someone awash in gold just because they thought it would *always* hold its value.
Reply
Increases in computing power aren't making cryptography less secure because you generally can get exponential increases in how hard a crypto-system is to break with only linear increases in key size and encryption difficulty. That's not to say it's a non-issue, sometimes software (especially proprietary software) is used for far longer than it's updated. But it's not a particularly hard issue, even for bitcoin in particular.
quantum comes to mind
It might be theoretically possible to build quantum computers that can factor large numbers more efficiently than classical computers. But:
1. The state of the art in 2016 involves quantum computers that can prove with high confidence that 3*5=15.
2. Being algorithmicly more efficient doesn't matter if your quantum computer isn't able to achieve the overall speed and scale it needs to compete with classical computers. In other words, if individual operations are much slower than on classical computers, a quantum computer could be more efficient in some sense, yet still be slower than a classical computer on every problem it's capable of solving.
3. It's still not certain that some of these theoretically-possible improvements in algorithmic efficiency are actually achievable with actual physical quantum computers.
4. There are some cryptographic algorithms that quantum computers are not (even in theory) more efficient than classical computers at cracking, so software developers could respond to breakthroughs in quantum computing by switching to those algorithms.
Still, the "secret quantum computer in the basement" scenario is at least theoretically possible. The implications for Bitcoin are far from the most relevant thing about that scenario!
(I rate the conspiracy theory that involves e.g. the NSA knowing about secret algorithmic flaws in crypto algorithms that mathematicians elsewhere haven't been able to find (or, at least, publish) as more likely than theories involving secret hardware that's orders of magnitude more capable than hardware elsewhere. And I actually don't rate the former as very likely.)
remember that one Vault episode?
Going to need more than that to jog my memory.
not to move Bitcoins, but simply to facilitate more conventional purchases, transfers, etc.
You need the former for the latter to be viable with a distributed system like Bitcoin. If there's no competition to be the one to append transactions to the ledger, then there's nothing preventing any participant in the network from double-spending. But since Bitcoin is a distributed system, the only thing it has to award the winner of this competition is bitcoin, and unless some people are trading in Bitcoin as a commodity in and of itself, that reward is worthless.
The reward in Bitcoin comes from a mix of transaction fees and newly-created coins. Theoretically, you could just do the former. Bitcoin itself is designed to move in that direction, the amount (as denominated in bitcoin) of the newly-minted reward halves every four years. But there's a bootstrapping problem; it's clear that Bitcoin never would have gotten off the ground without the mining element. Early transaction fees were just too small to ever incentivise anyone to do anything, even if they made really optimistic assumptions about Bitcoin's future value. Even now, with Bitcoin "mining" 75% complete, the reward for appending a block of transactions to the ledger is 6 bitcoin and the average sum of transaction fees for a block is about 0.5.
I think the details of this are one of the most interesting things about Bitcoin as a system. These details were chosen somewhat arbitrarily. They're one of the things that determined whether Bitcoin would ever get off the ground, and I think those details are going to be a big factor in whether Bitcoin is economically viable in the long run. (I think the system is more likely to fail due to economic unsoundness than security flaws.)
(Continued in another comment due to extreme verbosity.)
Reply
we are all to some extent benefiting from an economy allowed to be more complex thanks to those gooey ergs of ancient sunlight
We all stand on the shoulders of giant dinosaurs.
Setting up, plugging in, and waiting for the coins to *bing!* into existence is as challenging as sitting on a property and waiting for the rent checks to clear.
Actually, that description is not even remotely accurate.
Collecting rent is indeed simple. Buy some of a fundamentally-limited resource (e.g. land in a given location) with predictable use-value that far exceeds its maintenance cost, then profit!
Mining bitcoin now is not like that. You need specialized hardware and efficient data-centers just to turn a profit. And that resource, while expensive, is not fundamentally limited, any number of competitors anywhere in the world can eat your lunch. That is to say, bitcoin-mining hardware is neither fundamentally limited, nor does it have use-value that predictably exceeds the cost of running it. The sound advice on trying to make money mining bitcoin is generally "don't".
There was a time when someone could reliably mine a bunch of bitcoin just by installing and running some software on their home computer. That seems like it fits the bill, until you remember that at the time, bitcoin was totally worthless. By the power of hindsight, that looks like easy money! But it turns out that you can still today easily generate all sorts of worthless and useless things for minuscule investments of time and money, but predicting which of those things will be involved in a future financial bubble is quite difficult!
That's not to say this is fair and there's no privilege involved here. It takes a privileged position to be able to do anything financially speculative at all. And if you want to hear about things that might be the subject of a financial bubble, it helps if you're talking to people who have money. Speculation and rent-taking are both financial activities. But in many ways they're on totally opposite ends of the financial spectrum.
A part of me will gleefully deny food for someone awash in gold
Compared to fantasizing about trading gold for beans, it really seems no better (nor more realistic) to fantasize about not trading beans for gold.
Reply
So they never try.
And those that do dedicate the actual amount of time required to bestow upon listeners a mere smidgen of a smudge of understanding prove such complete blowhards that lack the simple skill of communicating with clarity, if they can communicate at all.
So they should never try.
Which leaves me listening to non-experts who can nonetheless communicate arcane minutia in a manner best described as Elephant Braille that never reaches beyond the trunk.
Had Bitcoin Guy simply pointed out that coin miners do more than gobble up electricity in their coiny pursuits (he actually declared that the act of electrical consumption was enough to justify coins); that they instead actually help maintain the ledger of transactions as an integral part of BC functionality-as you, Sir, just did- I might have given him a bit more credit, along with all the others dedicated to such a pursuit.
I'll chalk up the whole encounter with BCG as yet another "You could have mentioned that earlier" moments in life.
(Oh, and the Vault guy to which I earlier referred was the dude claiming to be trained by various secret services to be the next secret leaker. Has to be one of the more bizarre interviews KMO ever managed to record.)
Reply
Anyways, using proof-of-work to ensure eventual consistency in a distributed system with no central authority is the innovation of Bitcoin. It's a great piece of computer science, the algorithm would be pretty exciting to me even it turned out to have no practical application. The way it works is pretty counter-intuitive, though.
The thing is that the specific work miners are doing does nothing to ensure the security of the system. They do verify that individual transactions follow the rules, but that's easy. The hard part is keeping the system in a consistent state. The system is hard to get out of a consistent state if winning the privilege of appending to the block of transactions is hard. But it can't be too hard, or the system will grind to a halt. The rules specify that the difficulty of the problem is adjusted based on how quickly the previous set of problems were solved (this happens approximately every two weeks), to keep the average speed at which blocks of transactions are processed approximately constant.
You might view this as a prisoner's dilemma, where:
Cooperate = Everyone goes slow so the difficulty stays low
Defect = Work as quickly as possible to increase your share of the reward, even though that will increase the difficulty for everyone
The problem is that the cooperate scenario becomes increasingly vulnerable to lone attackers with the computing power to just steamroll everyone. That outcome is catastrophic.
(In Bitcoin, the longest chain of transactions is the true state of the world. Once a transaction is a block or two down from the most recent in the ledger, you can assume it's astronomically unlikely that block will be discarded. Except if an attacker has way more computing power than the rest of the network, then they can publish a whole bunch of blocks of transactions at once and catch up. That would allow them to double-spend.)
So Bitcoin encourages everyone to throw those levers to defect right away. The network's honest actors are cooperating to defect so that attackers can't defect alone. (This is a property that makes Bitcoin work as a trustless system, but if your response to that is, "Hmmm, that seems to be a serious disadvantage of trustless systems," you'd be correct!) The result is that Bitcoin miners are doing huge amounts of computational work with the indirect benefit that the system is secure from attackers. The serious drawback of that scenario is that Bitcoin miners are doing huge amounts of computational work with no direct benefit, it's just all solving meaningless cryptographic puzzles.
(There are other indirect effects that are more mixed. For example, it's put more funding behind research for specialized hardware to solve these particular cryptographic puzzles, but maybe that has led to some improvement in general methods for creating specialized computer chips. And it's pushed forward the cryptography arms race in some ways, with a mix of benefits and drawbacks for the security of things other than Bitcoin.)
Unsurprisingly, it has occurred to people that you could do a blockchain-like thing where the direct effect of the competition of miners is also useful work. It's tricky, though, you need to find a computational problem where:
1. Solving the problem is hard
2. Verifying a solution is easy
3. The difficulty is easy to adjust
4. Solutions are useful
The system doesn't work without any of the first three, and Bitcoin doesn't achieve the fourth.
(That's probably far too much of a ramble, but seriously, I love this stuff. No idea if this will be a historically-significant development in the history of money, but it deserves more than a footnote in the history of computer science. Well, at least in a textbook about cryptography or the design of distributed systems.)
Reply
the dude claiming to be trained by various secret services to be the next secret leaker
That might be C-Realm Vault 56? At least, that's related to the "secret quantum computers" stuff you mentioned earlier. (I searched for some keywords, found C-Realm 375: Synthetic Slaves and Snowden Bots, and guessed from that post that it might be a vault episode published close to that one.)
Reply
Reply
Leave a comment