Their method of decentralization seems pretty costly though, and that cost will not abate over time, since the costliness is what makes it work. If widely adopted, I don't look forward to the day when a large percentage of all computing resources in the world is spent meaninglessly churning bits in the name of a decentralized system.
Their method of decentralization seems pretty costly though, and that cost will not abate over time, since the costliness is what makes it work. If widely adopted, I don't look forward to the day when a large percentage of all computing resources in the world is spent meaninglessly churning bits in the name of a decentralized system.
Every market has friction and costs associated with its security. This is fundamentally no different than a bank hiring a security guard. If someone comes up with a solution to this problem which lacks this cost, it would of course be preferable, but no one has yet done so.
So... if I want to settle a $40b transaction, like a big merger, in <24h with bitcoins, assuming the level of fragmentation you could expect if bitcoins became ubiquitous in everyday transactions, what kind of server farm are we talking about?
It could take years before the block chain is long enough to dissuade a double-spend attempt for a transaction that large, where you have a single party single-handedly accounting for a significant fraction of spending on the network. So for a single $40b transaction, perhaps bitcoin is infeasible (or at least must be combined with traditional mechanisms like reputation, courts, etc.), but for a million independent $40k transactions, they can all be confirmed in a matter of hours for only a few hundred thousand dollars worth of computation, total.
The amount exchanged in a transaction doesn't directly affect the time required to settle it in Bitcoin. Bitcoin is not so much electronic currency as it is a distributed accounting system. As such, the monetary value of the transactions is irrelevant.
The thing that matters is the size of the transaction in bytes. This size is essentially proportional to how many cryptographic key pairs are involved in the transaction. If your $40b are associated to a single key pair, then it doesn't matter whether you send $40b or $1. If they're distributed over many different key pairs, then it does matter - but then again, you could have $1 distributed over just as many different key pairs.
Agreed, and since it seems that the price of bitcoins will converge to the average cost of computing bitcoins, won't the ones with an economic advantage will either be those with efficient hardware/software or those with free computing resources (botnet operators)?
the price of bitcoins will converge to the average cost of computing bitcoins
Actually, it's the other way around. Since rate of supply is fixed by the protocol, the cost of producing that supply has no influence on the price.
But the price does have an influence on how much the suppliers are willing to pay. Hence total bitcoin computation power follows price (with a delay due to time to acquire new hardware or repurpose existing hardware).
