At least they're sensitive to the issue. BFL customers who paid in full 11 months ago are pissed to see Ars and Wired (aren't they the same company?) getting free equipment before paying customers.
Hyping a BFL's product with no mention of the number of outstanding orders that haven't been delivered, nor of competitor products. Also, failing to disclose that he was mining on a BFL employee's mining pool (Eclipse Mining Consortium) is connected to BFL-Josh completely destroys any credibility that otherwise might have been. I'm less impressed than ever WRT Robert McMillan's journalistic ethics.
lol yes but this is a product that has within itself the seed of its owners undoing... the more they sell, the more bitcoins are mined and the harder the next is to be mined
my point was that these businesses that produce products with short life cycles are process oriented; can be very rewarding... VC won't come knocking, but that isn't always a bad thing.
But they've already sold it by then. Those who get rich in the gold rush are selling the shovels. Very clever to get in on this, especially because first mover will mean literally everything here.
Spending real resources (electricity/CO2) to produce money, then destroying the money, reducing the eventual money supply of this currency permanently, doesn't seem like the best way to handle any worries of journalistic integrity. They could've donated it just as easily.
Exactly! The only obvious thing to do is something constructive... not a blatant waste. In addition they are increasing the difficulty for others. He might as well flush money down the toilet and let sick orphans watch.
> He might as well flush money down the toilet and let sick orphans watch.
I hope you're engaging in Poe's Law here. Wasting an amount of electricity comparable to leaving a single incandescent light bulb turned on is like flushing money down the toilet while sick orphans watch? Okaaaay then.
The resources were minimal, but the final product has a non-trivial value. You are saying that being proud of throwing away hundreds of dollars is not an affront to all those who could use that money? I am not using Poe's law here.
Bitcoin is either a commodity or a currency, depending on who you talk to. If it's a currency, it has the ability to purchase goods and services. If it's a commodity, then creating it and throwing it away represents a waste of resources... plain and simple.
Your opinion that 'throwing money away hurts nobody' must mean that you disagree with the statement that 'all that is necessary for evil to prevail is for good men to do nothing.' That is essentially what is happening if we take currency/commodity and throw it away. They could have donated it to any charity and escaped the ethical issues. Instead they passed on that ability, there IS value in the money. This isn't some purist Economics debate, this is real life application. Same principle behind the whole idea of 'every dollar spent on war machines is a dollar not spent to clothe and feed the masses.'
It's an interesting exercise. Any party with an interest in manipulating bitcoin can do the exact same thing. No governing authority means no governing authority.
Destroying currency is not wasteful in the sense that destroying an equivalent amount of food is wasteful because the total value of all the BTC is almost completely independent of the number of BTC in existence: if there were half as many BTC in existence, each one would cost twice as much. Consequently, your destroying some BTC enriches everyone else holding BTC to almost exactly the same degree that it impoverishes you.
I anticipate objections to the above on the grounds that "deflationary currencies are bad". The most chartitable interpretation of, "deflationary currencies are bad," is "the downward stickiness of the price of labor": during a recession, unemployment tends to go up in part because most employees are irrationally averse to taking a pay cut when times get rough, and enlightened monetary policy can sometimes negate or ameliorate that particular irrational reaction by employees by inflating the currency, which has the effect of fooling the employees into thinking that their "real" compensation has not decreased. But that is a minor effect compared to the basic "invariant" described above -- especially since almost no employees are being paid in BTC.
But the bitcoins are still in existence. It is impossible to distinguish between a bitcoin that can be spent and one that cannot. Unless you could actively tell the bitcoin servers "I am destroying these coins", nobody can determine which bitcoins are inaccessible. So won't they continue to maintain the value they would if they were still accessible?
The price ostensibly depends on the coins that are available for purchase. People will do some guessing about the price trajectory, but they will mostly consider the current price. Other notions of value exist, but people overwhelming use price.
There are basically two uses of bitcoin: a medium of exchange and a store of value. My "invariant" above can be rephrased to say that the total value of all bitcoin is approximately equal to the total amount of value people ("investors") want to store in it. (Entities, e.g., Silk Road, that use BTC as a medium of exchange can immediately convert the BTC they receive from their customers into dollars if they do not want to hold BTC.) So, if an investor has $10,000 he wants to store in bitcoin, it is almost completely immaterial to that investor whether he can buy 100 BTC or 10 BTC with that $10,000; all he really cares about is how well his $10,000 investment will hold its value, which is sort of independent of the current price, since for example the fact that 1 BTC costs about $100 today is pretty good evidence that they will cost about $100 a month from now -- and if 1 BTC cost $1000 today, that would be pretty good evidence that they will cost about $1000 a month from now.
The point is that withholding BTC from sale increases the price of BTC whether or not the reason for the withholding is so that one can sell them a month from now or because one has destroyed them (i.e., destroyed the relevant private key).
That is an example of the "Fundamental law of microeconomics," which says that in an efficient market, the price is determined by supply and demand -- "supply" meaning the number of BTC for sale today, not the total number of BTC in existence.
Thought experiment: how would the price of BTC be determined if no one knew how many BTC are in existence? Would people just give up on trying to buy or sell BTC because that figure is unknown?
P.S. Comments like these where I answer a technical question tend to stay at score 1. Since I am making an effort to increase my average comment score (currently 2.0) I plan to stop answering technical questions in comment threads unless this comment gets upvoted at least to 2 or 3.
Pretty awesome watching the technological innovation around Bitcoin. I personally don't have any, but just from following the progress and learning, I am starting to dabble with technologies (FPGAs, for example) in my research that I completely did not anticipate getting familiar with. Great hardware opportunities today for those traditionally limited to software (like me!).
This sort of thing has actually been pioneered by high frequency traders. Utilizing FPGA's & ASIC's is pretty standard fare for high frequency trading and algorithmic trading. What's novel about bitcoin is that it has been driven largely by end consumers. Thus, it is easier to see the technical landscape around it evolve.
If I understand their graph, they are generating about 0.29 BTC / day with this. Current BTC exchange is $118, so $34.22 revenue / day. At 44W, they'll spend ~1kW/day on power so only around $0.10 or so depending on where their power is coming from. So their break even point is somewhere around 8 days if I have everything correct.
It seems that if these devices can be produced at scale, the marginal cost of mining 1 BTC will be only about $0.30 for now, thereby possibly pushing down the market price for BTC significantly.
Bitcoin are mined at a (relatively) fixed rate, around 3600 per day right now. Mining has little impact on the market price (10,000s of bitcoin are crossing Mt Gox each day).
What will happen is that the 0.29 BTC / day will go down as more people plug in their asic miners.
>So in just two weeks, those lucky enough to have snagged one of these rigs will have paid off the initial investment. Everything after that is gravy.
That's the biggest joke ever. The difficulty will go up substantially once thousands of people get their ASIC miners. It will take a lot longer than two weeks to pay itself off.
> Meanwhile, with more and more computers joining the network, mining difficulty is quickly getting harder. So the amount of money these machines can make per day is slowly declining.
The value in USD of BTC comes from the supply and demand of people exchanging it for USD.
So the value is derived from usage, not the cost of what it took to mine. How much does it cost the Fed to mint a USD? The gap in cost/value is called seigniorage.
Is it hard to re-engineer these once they are out? Why is in this case supply/demand artificially shortened and not just regulated over the selling price
They were one of the two companies which managed to produce dedicated integrated circuits - so called ASICs.
ICs are blazingly fast, because they have only one purpose - to compute SHA2 / to mine bitcoins. There is no overhead that processors & GPUs had.
A production of the IC costs high hundreds of thousands of dollars, and requires a ton of domain-specific knowledge. But once they're produced, each batch is quite inexpensive, and is capable of quite fast.
Of course there could be other startups doing the same thing, but you'd need an around a million dollar budget. Not easy.
Marketing (for example, BFL equipment is much prettier than Avalon).
What kind of proprietary knowledge do they have, that there are these long waiting lines for the machine?
None. But they do have capital, which they got by having a fairly good reputation from making cheap FPGA miners which they could do because they found a way to buy obsolete Altera FPGAs really cheap.
What is they key issue that makes these boxes so hard to replicate?
People who know what they're doing don't design Bitcoin mining equipment. There are probably a lot of different reasons why, mostly due to very high financial risk. The only people designing this stuff are incompetent noobs, with predictable results.
>But they do have capital, which they got by having a fairly good reputation from making cheap FPGA miners
The capital they have is from taking pre-orders and not delivering. They've also burned a lot of that capital spamming the internet with ads to sell more pre-orders. If they ever had a good reputation they have squandered it by letting their loud-mouthed-pompous-ass spokesperson loose on the forums.
How are people getting such huge variances in the wattage drawn for the 5GH/s miners? Wired measured it to be 43W-44W, Ars measured it as 50W [1], while David from Coding In My Sleep [2] measured it to be 30W.
Likely because they're using different revisions of the hardware. BFL didn't meet their power usage promises and have been going through several design changes over the past few weeks to try to get the power down more.
Likely not worth it actually. In 6 months they'll be obsolete or barely make as much money. The entire premise of the protocol is that the difficulty keeps going up basically. As so many of the machines enter the workforce things even back out to where there's no distinct advantage. If they can keep R&D going to make better ones they can keep selling them regardless of how the bitcoin market is doing as far as difficulty goes.
Plus if they mined all the coins themselves it would lower confidence in the currency (especially if it meant they aproached 51% of the mining capacity) which would entail that the price of bitcoin would go down. Possibly before they got their moneys worth. It is a lot safer to make money selling them to lots of people.
Only the first few miners will be profitable - if there will be tens of thousands of them already running, the costs will be higher than the returns. (Probably - nobody knows for sure.)
Not always the right way to go. The truth is, you can do either and make a lot of money. Depends on what kind of risk you want to take on, what kind of capital you have etc.
Ask Barrick Gold or Newmont etc. It's very context dependent. Barrick is worth $20 billion, and have made $8.5 billion in profit the last three years (the gold market tends to produce wild swings of prosperity and poverty years).
If you have little startup capital and can't buy a great mine and can't afford to prospect for one (or don't want to roll that dice)... then maybe sell shovels.
I'm NOT saying they did, but someone in their position could take the preorder money, build the boxes, start running them en masse, tell customers that orders are delayed in manufacturing, have them wait, and finally deliver once they have extracted a bit of value.
57 comments
[ 2.9 ms ] story [ 69.1 ms ] threadThe solution of effectively destroying the earned money is pretty funny, given that that's an occasionally expressed concern.
I wonder if the typo was intentional.
Hyping a BFL's product with no mention of the number of outstanding orders that haven't been delivered, nor of competitor products. Also, failing to disclose that he was mining on a BFL employee's mining pool (Eclipse Mining Consortium) is connected to BFL-Josh completely destroys any credibility that otherwise might have been. I'm less impressed than ever WRT Robert McMillan's journalistic ethics.
[EDIT] They linked to the wrong one. The one they are using does cost $274. Not bad ROI.
my point was that these businesses that produce products with short life cycles are process oriented; can be very rewarding... VC won't come knocking, but that isn't always a bad thing.
More people made money selling shovels. The people that got the richest owned the mines.
I hope you're engaging in Poe's Law here. Wasting an amount of electricity comparable to leaving a single incandescent light bulb turned on is like flushing money down the toilet while sick orphans watch? Okaaaay then.
Your opinion that 'throwing money away hurts nobody' must mean that you disagree with the statement that 'all that is necessary for evil to prevail is for good men to do nothing.' That is essentially what is happening if we take currency/commodity and throw it away. They could have donated it to any charity and escaped the ethical issues. Instead they passed on that ability, there IS value in the money. This isn't some purist Economics debate, this is real life application. Same principle behind the whole idea of 'every dollar spent on war machines is a dollar not spent to clothe and feed the masses.'
A unit of money has no value itself. Destroying it is exactly equivalent to distributing it evenly among all other money-holders.
The energy/resource waste of creation is real, though.
I anticipate objections to the above on the grounds that "deflationary currencies are bad". The most chartitable interpretation of, "deflationary currencies are bad," is "the downward stickiness of the price of labor": during a recession, unemployment tends to go up in part because most employees are irrationally averse to taking a pay cut when times get rough, and enlightened monetary policy can sometimes negate or ameliorate that particular irrational reaction by employees by inflating the currency, which has the effect of fooling the employees into thinking that their "real" compensation has not decreased. But that is a minor effect compared to the basic "invariant" described above -- especially since almost no employees are being paid in BTC.
There are basically two uses of bitcoin: a medium of exchange and a store of value. My "invariant" above can be rephrased to say that the total value of all bitcoin is approximately equal to the total amount of value people ("investors") want to store in it. (Entities, e.g., Silk Road, that use BTC as a medium of exchange can immediately convert the BTC they receive from their customers into dollars if they do not want to hold BTC.) So, if an investor has $10,000 he wants to store in bitcoin, it is almost completely immaterial to that investor whether he can buy 100 BTC or 10 BTC with that $10,000; all he really cares about is how well his $10,000 investment will hold its value, which is sort of independent of the current price, since for example the fact that 1 BTC costs about $100 today is pretty good evidence that they will cost about $100 a month from now -- and if 1 BTC cost $1000 today, that would be pretty good evidence that they will cost about $1000 a month from now.
The point is that withholding BTC from sale increases the price of BTC whether or not the reason for the withholding is so that one can sell them a month from now or because one has destroyed them (i.e., destroyed the relevant private key).
That is an example of the "Fundamental law of microeconomics," which says that in an efficient market, the price is determined by supply and demand -- "supply" meaning the number of BTC for sale today, not the total number of BTC in existence.
Thought experiment: how would the price of BTC be determined if no one knew how many BTC are in existence? Would people just give up on trying to buy or sell BTC because that figure is unknown?
P.S. Comments like these where I answer a technical question tend to stay at score 1. Since I am making an effort to increase my average comment score (currently 2.0) I plan to stop answering technical questions in comment threads unless this comment gets upvoted at least to 2 or 3.
It seems that if these devices can be produced at scale, the marginal cost of mining 1 BTC will be only about $0.30 for now, thereby possibly pushing down the market price for BTC significantly.
What will happen is that the 0.29 BTC / day will go down as more people plug in their asic miners.
>So in just two weeks, those lucky enough to have snagged one of these rigs will have paid off the initial investment. Everything after that is gravy.
That's the biggest joke ever. The difficulty will go up substantially once thousands of people get their ASIC miners. It will take a lot longer than two weeks to pay itself off.
> Meanwhile, with more and more computers joining the network, mining difficulty is quickly getting harder. So the amount of money these machines can make per day is slowly declining.
What kind of propitiatory knowledge do they have that there are these long waiting lines for the machine?
What is they key issue that makes these boxes so hard to replicate?
ICs are blazingly fast, because they have only one purpose - to compute SHA2 / to mine bitcoins. There is no overhead that processors & GPUs had.
A production of the IC costs high hundreds of thousands of dollars, and requires a ton of domain-specific knowledge. But once they're produced, each batch is quite inexpensive, and is capable of quite fast.
Of course there could be other startups doing the same thing, but you'd need an around a million dollar budget. Not easy.
Marketing (for example, BFL equipment is much prettier than Avalon).
What kind of proprietary knowledge do they have, that there are these long waiting lines for the machine?
None. But they do have capital, which they got by having a fairly good reputation from making cheap FPGA miners which they could do because they found a way to buy obsolete Altera FPGAs really cheap.
What is they key issue that makes these boxes so hard to replicate?
People who know what they're doing don't design Bitcoin mining equipment. There are probably a lot of different reasons why, mostly due to very high financial risk. The only people designing this stuff are incompetent noobs, with predictable results.
Why? Is it not profitable enough?
The capital they have is from taking pre-orders and not delivering. They've also burned a lot of that capital spamming the internet with ads to sell more pre-orders. If they ever had a good reputation they have squandered it by letting their loud-mouthed-pompous-ass spokesperson loose on the forums.
[1] http://arstechnica.com/gadgets/2013/05/weve-got-a-butterfly-...
[2] http://codinginmysleep.com/bfl-jalapeno-unboxing-and-demo/ (see 7:33)
As for why they sold them at all, I'd assume they didn't have the capital to build them on their own.
Ask Barrick Gold or Newmont etc. It's very context dependent. Barrick is worth $20 billion, and have made $8.5 billion in profit the last three years (the gold market tends to produce wild swings of prosperity and poverty years).
If you have little startup capital and can't buy a great mine and can't afford to prospect for one (or don't want to roll that dice)... then maybe sell shovels.
2. Other miners would reject their blocks if they "51%-attacked" the network. (Nobody would trust Bitcoin if one entity could control the blockchain.)