Yeah, that's another thing. Basically a group of at least 50% of the bitcoin holders have to get together and agree on new rules. This is because each peer validates the transactions.

I think though that the chain of transactions can be split up pretty easily, its already built into the protocol.

Really, the whole protocol behind bitcoin is really sleek technically. It is unfortunate that other poor decisions were made.

It is interesting to think about what could have been done besides an absolute cap on the number of bitcoins.

Deflation/inflation is a function of demand as well as supply. It's hard to think of a rule that bitcoin clients could follow that would make inflation always be just a few % per year, as the US attempts to do for the dollar. Depending on population size change (I.e. demand), a 0% increase in bitcoin supply could be inflationary and a 10% increase in supply could be deflationary.

You can argue that 0% per year increase in bitcoin supply is not the right number, and I don't necessarily disagree-- but what is the right number? Bitcoin clients only affect the bitcoin supply, not demand...

I mean as we speak bitcoin is probably inflating, not deflating...

_________________
That which can be destroyed by the truth should be.
--
My (sadly neglected, but not forgotten) project: http://dailyjoseki.com

These are interesting points.

To some extent, the "who's responsible?" question has always been something that people ask about anything that's open source. In general it's hard to conceive of it working, but a lot of the internet runs on it. I have a good friend who maintains that Wikipedia shouldn't work as well as it does and his arguments are persuasive...

Presumably most of the people running clients are only involved passively because they're using the distributed network. So even if their computers are validating transactions, they're not making the decisions. I don't think you need to reach a 50% consensus with the users. In theory, you just need consensus amongst contributors and then you gradually roll out software updates for the clients and phase out the old rules. Since there's money involved some people would still take issue with this, particularly if they stood to lose something. So making major changes could be unwise. It's not perfect, but I think that's how it would have to work in practice.

To be honest, my understanding of both economics and cryptography is relatively basic, so even though I did some reading about bitcoin, I won't presume to critique the design. I still view it as a proof of concept and if it fails because of various constraints, it will be an opportunity to learn and make the next attempt better.

One thing I'm wondering about is, when mining becomes less profitable (as SpongeBob says) and the network is big enough, couldn't you just start to take advantage of the idle cycles of all the computers that are running the client? What's not really clear to me is just how intensive the processing for validating a transaction is. Possibly some of the design decisions were predicated on assumptions that processors would become x times more efficient and renewable (but not free) energy would become more prevalent. What do other people think?

Another thing I was wondering about is the deflation issue. On the Wikipedia page someone wrote:
So it seems to me you're not going to run out of divisible units any time soon, although that number still does seem a bit small for a truly global currency. But my question is, if there's very slow deflation, to the point where it's still better to invest money elsewhere than hoard it, to what extent does that matter?

Secondly, if a lot of early adopters choose to speculatively hold onto the coins and limit supply, there must be some point where the value of a bitcoin gets so high that the system becomes unworkable, people start to abandon it, demand collapses and the experiment fails. But before that happens, if we assume some people are rational (always dangerous), won't the smart people sell or spend their coins and push the system back to some sort of equilibrium? I imagine that it would then depend on the initial conditions and rules of the system as to where that equilibrium lies, and it may lie at a point where the tendency of the system is to spiral out of control until it collapses. Though it's probably hard to model this and find out without putting it into the 'real world'.

This might not be the right way to think about these things, but my background is in chemistry, so I guess I tend to think in those terms. Does anyone have any thoughts on this?

_________________
David

Go Game Guru: Learn Go | How to Get Better at Go | Go Game Shop | Go Boards | Baduk TV

The block difficulty scales with the number of hashes per second the network of miners maintains. Faster computers just mean the difficulty goes up. If the cryptography behind it breaks, we'll have a lot more problems than just bitcoin not working.

Jts will have to respond to the rest of your comment.

_________________
That which can be destroyed by the truth should be.
--
My (sadly neglected, but not forgotten) project: http://dailyjoseki.com

Ahh, I see, thanks for clarifying that Daniel. In that case, I don't really understand how that won't become a problem . I could be wrong though. Whoever designed this thing was pretty hardcore.

_________________
David

Go Game Guru: Learn Go | How to Get Better at Go | Go Game Shop | Go Boards | Baduk TV

I guess technically it scales with how long it took to create the last 2000 blocks, but it amounts to the same thing. The difficulty slope is exponential, we'll not run out of difficulty any time soon, and if in the distant future it looks like we might, the hashing algorithm can be changed by 51% of the miners switching over.

_________________
That which can be destroyed by the truth should be.
--
My (sadly neglected, but not forgotten) project: http://dailyjoseki.com

Let's say I have a lot of resources and happen to own more than half of the Bitcoin network nodes in the world (I have lots of computers and perhaps a botnet or two), what then?

Sure.

Money has more than one function. Acting as a store of value is one. People want to put money in the bank or in a hole in the ground and have it retain its value. Inflation works against that. If your money buys less tomorrow, why save it today?

Another function is as a medium of exchange. People want to buy things with money. Deflation works against that. If your money buys more tomorrow, why spend it today?

I would like to say that the medium of exchange function is more basic than the store of value function, but the Yap stones tell against that. However, in modern economies, the value of money is realized when you spend it. The threat to spend it may also have value. (Relation to aji and thickness is interesting, but not relevant to the current discussion. )

In the modern era we have learned a fair amount about money. One thing that we learned is Gresham's Law: "bad" money drives "good" money out of circulation. Good and bad refer to how well money stores value. Multiple forms of money in the same region is not a prominent feature of life today. But from what I hear, before the Federal Reserve Act of 1913, there were some 50,000 different currencies circulating in the U. S. These were not only bank notes, businesses could issue money that they promised to honor for purchases. Of course, bank notes could become worthless if the issuing bank went out of business. If you were unsure about a bank, you might not accept its note at face value, but discount it. You will tend to save money that you trust to retain its value and spend other money. Bad money drives good money out of circulation.

Bitcoin is not backed by a bank or government, but by a community of people who use it. No bank is going to go out of business, no country is going to be conquered. As long as there are people who are willing to accept bitcoin in payment, it will have value. There is no inflation with bitcoin. In that sense, bitcoin is good money, by comparison with every form of government money today. By Gresham's law, other money should drive bitcoin out of circulation. But if it is driven out of circulation, since there is no guarantee that anyone will accept bitcoin, it can lose its value almost overnight. In that sense it is bad money.

I suppose that there may be some value where these countervailing forces are equal, but I do not think that there is a stable equilibrium. If enough people decide to save their bitcoins at the same time, the market could dry up. If so, how does it start up again? Perhaps there will always be enough people who will want to buy bitcoins for money laundering purposes or drug transactions that the market will not dry up. Anyway, it is an interesting social experiment.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Just as transaction fees will ultimately have to be integrated into the protocol (correct me if I am wrong here), it would in principle be possible to do away with the upper limit of bitcoins, wouldn't it?

Also, are there any educated guesses about how high those transaction fees are going to be? Someone has to make profit with validating those transactions, just as other payment systems like PayPal have to be profitable. And if transaction fees would be comparable to those of other payment systems, wouldn't Bitcoin then lose one of its major benefits?

_________________
Stay out of my territory! (W. White, aka Heisenberg)

I have to double check, but when I say 50% I mean 50% of the bitcoin wealth. Not 50% of the participants.



A small number of people own 50% of the bitcoins. They own a lot. Releasing more bitcoins would devalue their holdings, so they will never decide to do that. It is one of the difficulties that arises when having an unregulated system and nobody is interested in acquiring goods. They are still only interested in the value against the dollar (or insert real currency here) and not acquiring goods or investment with bitcoins.

I think there is a misunderstanding about what "money" is or is not.

An essential feature that distinguishes money from just another commodity is that it is generally accepted and used.

Let's settle this qestion. Have any of you ever used or received a "bitcoin"? I mean in a financial transaction, buying or selling something, not buying or selling bitcoins themselves for (real) money.

If not, then bitcoins are just a commodity.

Commodities have uses. The only use for bitcoins is as a medium of exchange. Bitcoins are just as much money as Monopoly money or airline miles.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

51% of the miners set the rules, not 51% of the wealth.



It would be an incredibly expensive proposition to produce more than half of the hashes on the network (would probably cost on the order of 35,000-50,000 video cards, computers to run them, and electricity). It would also be incredibly stupid, because it would probably cause bitcoin to drop in value, and then you wouldn't be able to recoup your investment.

Unless you wanted to kill bitcoin.

_________________
That which can be destroyed by the truth should be.
--
My (sadly neglected, but not forgotten) project: http://dailyjoseki.com

I'm going to narrowly try to answer these two. These answers aren't necessarily meant to be convincing, just to give you an idea of what mainstream economists think about these questions.

Deflation:

Equilibrium prices:

I would like to quote this comment as it highlights my biggest concern about bitcoin.

Do you want to create a new online currency? Fine. However, keeping to yourself or distributing to the very early adopters the majority of the "wealth" is a certain way to kill it.
I'm saying this assuming good faith from the creators, even though I highly suspect that this was planed from the beginning as a "feature".

Not only that, but the identity of the creator is pretty much an unknown.
https://en.bitcoin.it/wiki/Satoshi_Nakamoto

... and the usage (number of clients connected to the networt) is declining significantly, see the current chart.

_________________
Stay out of my territory! (W. White, aka Heisenberg)

It's unclear how meaningful those charts are. I start up my client only rarely. Trading doesn't require running a client.

I would like to hear what jts has to say about my comment earlier (what kind of automated rule would cause a small amount of inflation?)...

_________________
That which can be destroyed by the truth should be.
--
My (sadly neglected, but not forgotten) project: http://dailyjoseki.com

It depends what you mean by "automated." Your earlier question takes as a premise "bitcoin clients can only effect supply, not demand", so I took that to be the context: something like "what automated rule could produce a small amount of inflation, despite only taking account supply, not demand?" And the answer is "none". After all, inflation is affected not only by the physical monetary base, but also by how fast people are spending/lending it as opposed to hoarding it.

On the other hand, bitcoin could have an automated rule like "if inflation isn't accelerating, produce bitcoin at a faster rate; if inflation is accelerating, produce bitcoin at a slower rate." But that would require you to build a model of inflation into the bitcoin protocol, if I understand correctly what sort of automated rule you're looking for, and that would be a little bit difficult. Not difficult by the standards of science fiction, but difficult for the here and now.

From Wikipedia:



Bitcoin blocks are created at a fairly steady rate. It is the halving of the award that makes the total number of bitcoins constant, and is severely deflationary. If you follow Milton Friedman's formula, you would increase the award by a certain percentage, such as 1/512 every 2016 blocks. That should be mildly inflationary, at a similar rate to most developed economies.

Edit: You are really "debasing" bitcoin by increasing the award, but debasement tends to have inflationary effects. Also, increasing the award rewards current miners instead of rewarding early miners, which is what halving the award does. That's what you want for a virtual gold standard.

Edit 2: I changed the increment to yield around a 5% per year increase in the award. About half that should accommodate economic growth, about half that should be inflationary.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.