The idea is to stick these chips into household applications (e.g. USB chargers), give the appliances out for free, stick the electricity bill on you, and give 75% of the Bitcoin revenue back to the parent company.
Yep, a free USB charger that you amortize over its entire use, at many multiples of what it should cost. It's great for people who love playing a guaranteed losing shell game between buying electronics and electricity.
I'm becoming more and more convinced that it's impossible to profitably mine bitcoins without screwing someone else in the process. The trick is to screw them as long as possible without them knowing.
The scenario you describe is malicious because people don't expect a charger to have "features" which consume large amounts of power without benefit, but it does make me wonder if such tech could be incorporated into electrical items whose express intent is to turn electricity into heat.
Say, a clothes iron, or washing machine/dishwasher water heater (where not already plumbed for hot water), etc.
Since computation ultimately dumps essentially 100% of the consumed energy as heat, why not use it for that?
Obvious arguments against are the cost of the initial hardware, and the engineering difficulty of making something which can actually operate whilst generating kilowatts of heat (i.e. not cooking your cores in the process)
The idea has been around for a long while. Google "data furnace".
The crux seems to be Moore's Law. We buy heaters and use them for decades, but any compute device has a useful lifetime of a few years before it's surpassed. If we ever see Moore's Law come to an end, expect to see a lot of highly-distributed computing.
Under these conditions, it could still work in irons or especially electric kettles. Those are so crappy nowadays, you have to buy new every two-three years anyway.
Unless the device is on almost all the time, it won't compete with datacenters, even on price.
Using compute for heat is a dodgy proposition even where heat is needed at least half the year. The minimal uptime of an iron or kettle really puts the kibosh on the economics.
> I'm becoming more and more convinced that it's impossible to profitably mine bitcoins
I suspect that's been the case for a few years - at each time I investigated, in the pre-ASIC days starting from 2011, the electricity cost was roughly the same as the Bitcoin value, and the only people who seemed to be able to mine profitably were ones with "free" electricity (e.g. students in halls of residence).
Ironically this is the free market working as expected, that the sale value should collapse to exactly the marginal cost of production. Profit can only be maintained by barriers to entry or market differentiation, and there's a constant threat of oversupply with no way to cartelise to reduce oversupply, OPEC-style.
The difference is that for most goods, the marginal cost of production falls with increasing economies of scale. Bitcoin is the exact opposite: The block difficulty will continue to escalate as more miners enter the market. It's like running on a treadmill that keeps getting faster and faster. Oh, and instead of making you healthier, your treadmill pumps more and more soot into the air around you.
Yes. A bitcoin is basically a token that says that someone, somewhere has wasted a (quite large) amount of electricity on your behalf. It's an incredibly pure form of conspicuous consumption.
(Likewise with gold you know there's a pile of acidic tailings and a river full of arsenic to back it up, and with diamond mining quite a lot of real human suffering)
Though it's never wasting. You can phrase it better by saying that:
A bitcoin is basically a token that says that someone, somewhere has spent a (quite large) amount of electricity on your behalf to make sure that your bitcoins are secure and safe.
(The more computing power used, the harder it is to counterfeit transactions.)
The actual computation done is not computationally useful. It only "secures" your bitcoins in the sense that an attacker has to put more economic input into their attack to match the economic input of the mining network. The act of mining does not even check whether the transaction is valid - that is done by the host that decides which transactions to include and the rest of the network deciding to propagate a mined block. Mining just churns through a single hashing algorithm to deliberately expend power. You could also secure your network by introducing trust and actually securing the trusted nodes, without having to create an ecological travesty.
Except those computations aren't actually doing anything useful. They're not protein folding, they're not analyzing data, they're just computing for the sake of computing.
Mining operation KnC Group actually had to publish their earnings statement for a BTC-based fund they created in Sweden. Turns out, they have been losing money throughout 2014.
"it's impossible to profitably mine bitcoins without screwing someone else in the process"
That is not true. Most mining is done professionally in data centers these days, by companies that all pay for their own electricity without screwing anyone:
And so on. They certainly seem to be making a profit; it is much easier to do so nowadays (~5% monthly increases of the difficulty level) compared to last year (30-40% monthly increases).
I think you just landed on the "Free" version of future electronic devices. Imagine a terrible reality, not unlike today's videogame market, where every device, charger, lamp, TV, and box fan comes in "Free" and "Pay" versions - most manufacturers will have a mail back program, so you can return a broken "Free" item for a brand-new "Free" item. The "Free" items will mine Bitcoins and return all profit to the manufacturer.
Meanwhile "Pay" items will still be widely available from numerous online vendors. These devices will require you to enter your personal Bitcoin address, so they can credit you with their mining.
And thus the world will continue to be dominated by Whales, so named for the size of their bank accounts. The rich will be able to keep the entirety of their revenue, while the poor must subsist not only on the charity of the State, but also the charity of the Corporation.
There would no doubt be plenty hacking to disable them, or modify to feed the Bitcoins back to the user.
It certainly raises some interesting problems with ownership... if I can modify my hardware/software to disable unwanted features, this certainly fits in that category.
A USB charger that phoned home? If it's free, I'd certainly take it and modify it so it doesn't, in a similar fashion to the way a Lenovo laptop with Superfish would get its OS wiped and reinstalled.
The question of ownership on this topic isn't as new as you'd think.
If you get a free sports bag from your local training center, is it ethically right for you to use a black marker and erase their printed logo?
I think most people would say that once the training center has given you the sports bag, even if their intention was for you to advertise their brand by walking around with it, it's no longer their property and they cannot control it.
Yes, because most people will not erase their printed logo and they get advertisement. If they notice that most people erase the logo, then they could just stop giving away the bag.
In some ways it's like the adblock debate - and there are JavaScript bitcoin miners (https://news.ycombinator.com/item?id=2566365 ) - uselessly slow now, but the concept still holds.
I've collected a few "ad supported" free items (keychains, bags, writing boards, etc.) from various trade shows and conventions, and of those that are still usable, amusingly enough the logo was the first thing to come off from normal wear and tear.
Have you encountered the new breed of ad-supported free flash drives? I recall from a discussion on HN that apparently some people are putting a hub inside said flash drives and a fake HID device that will occasionally type in "Win+R http://companysite.com <enter>".
No, (un?)fortunately I haven't. I do have a few small ones which I don't use for anything critical, but they show up only as the usual mass-storage devices.
Their hope was for you to advertise for them, they certainly couldn't have counted on it and they know that up front so it's perfectly ethical to cross it out and even they wouldn't disagree.
Its crap like this that's going to force new federal regulations on us, that will probably be over-reaching and pegged to a bill full of questionable things. Its disheartening to see our energy usage drop because of the move to CFL's over incandescent, targeting vampire power appliances, moving from CRT to LCD, moving towards energy efficient chips, etc only to lose a lot of these gains because of some bitcoin stupidity.
uTorrent just did this by running a background miner bundled with the software. Considering how much extra coal we're going to burn for this, I think its going to become a regulation issue sooner than later. The industry's greed is begging for it.
I'd also be worried about a fire risk from these things. How hot do they get? How much ventilation do they need? How well do they handle a hot day in the American South without A/C? What kind of QA can we expect here considering these will be disposable tchotski's.
Have there been precedents of profiteering by hiding _something_ invisible to the end-user but with an ulterior purpose into normal consumer electronics?
Lenovo's Superfish? Prior to that, the Sony rootkit CDs? Pre-installed adware in general? Amazon's thing with the always-on microphone? The Samsung TVs that report all filenames on connected devices back to a server?
I miss the good old days of "If you aren't paying for a product, you are the product." Now, the new paradigm is: "By using our stuff, we own your data. All of it."
How do Global Fortune 500 companies think bundling malware is OK still after all the bad press the last few years?
You could try and stay under that 0.3W load, or only start using real power whle under load, but I expect the EU will move to forbid that, too, if an Intel-scale company starts doing it.
That will probably kill this idea world-wide, as economies of scale make it cheaper to have only one type of charger (economies of scale similarly exported lead-free soldering from the EU to the world and car safety and economics guidelines from California to the USA to the world)
Then it was in the context of botnets and chinese miners -- but the theory is that dishonest mining (stolen electricity) will necessarily drive our honest mining which will spell doom for Bitcoin.
Whoever is paying the electric bill. You see lots of people talking about mining from apartments with included utilities, dorm rooms, their work, etc. It sounds like 21 Inc. wants to take this route on a massive, distributed scale.
Wasn't there also a company that pre-sold mining rigs, used the pre-sale money to fund the manufacturing, and then used the equipment to mine for a few months before shipping, at which point the equipment could no longer profitably mine?
21 is best understood as an attempt by the VC community to break the Chinese Bitcoin mining monopoly. 21 doesn't appear to be working on Lightning Network like layer 2 technologies that would make the Bitcoin micropayments for server bandwidth viable on the blockchain. In fact as useful financial applications of on chain transactions start to materialize, even P2P payments may start to struggle to be included in blocks.
21 faces one of the most profound information security problem faced by any company. They propose to put their companies capital assets into the homes of consumers and then extract value from those assets in a manner adversarial to the consumers interests. End users/attackers can do everything from blackhole the network traffic of the device to modify the firmware to target the hashing power to a new mining pool. Virtually any weakness in the engineering has the potential to be exploited at vast scale. This should terrify them.
> They propose to put their companies capital assets into the homes of consumers and then extract value from those assets in a manner adversarial to the consumers interests. End users/attackers can do everything from blackhole the network traffic of the device to modify the firmware to target the hashing power to a new mining pool. Virtually any weakness in the engineering has the potential to be exploited at vast scale. This should terrify them.
Mark my words. If this or similar idea ever takes hold and begins to spread, I am going to start an open source project to hack those devices and redirect their hashing power where consumers want in as easy and streamlined way as possible so that everyone could do it.
I'm tired of companies extracting value from customers in adversarial manner, literally telling their users "fuck you" in their faces.
Embedding cryptographic processors in consumer devices to do massively parallel, low-communication work on behalf of a third party was once called "chinese lotto" or "chinese lottery".
I would say that 21 is an attempt to extract money from the VC community by pitching Bitcoin as the "next big thing". Given that this announcement doesn't seem to have official Intel press release backing, it's a little shady. I'm also skeptical about them getting to produce a small-run ASIC on Intel's incredibly valuable 22nm FinFET production line.
Actually it's even more interesting than "Intel makes a bitcoin mining chip," which really shouldn't be surprising given how much application-specific accelerators have been making noise, and how many of these chips are already on the market. Multibillion dollar companies love to hedge their bets; it'd be somewhat more surprising if nobody at Intel were working on one.
No, the more interesting bit is that it looks like some stealth startup has a close enough relationship to Intel to get them to mint chips. And that's impressive; Intel's silicon manufacturing processes have historically been a black box to outsiders.
It'll be a lot more interesting once more details about this pour out. This could be the start of something huge...
This is presumably from the same source as the recent spate of "mining from toasters" articles started by FT's Alphaville [1]), but with a lot more detail.
The Vimeo clip discussed in the coindesk article shows a live demo from Oct 2014 of a bandwidth auctioning protocol, (the article links to the actual transactions eg [2])
Why would anyone agree to pay for 100% of the power consumption for an ASIC that only offers 25% of the returns? Consumers may not know anything about crypto-currencies but they understand basic math.
When consumers are offered 5 dollars worth of bitcoin as compensation for a 20 dollar increase on their power bill they'll simply conclude "i guess that bitcoin thing really is a scam"
Surely the idea is that the return is more than 4x the cost of the power consumed, so that it's still a net gain for the buyer? It seems unlikely to me that this would actually work out in the long term, but wouldn't that be what they're going for?
That's a good question. For normal miners there's a similar question that's answered by a lack of unlimited capital. That wouldn't seem to apply here, since they're paying for the hardware themselves up-front and trying to recoup costs later. I can't figure out the answer here.
Minus the cost of marketing, distribution, product engineering, customer support, security engineering and lawyers (e.g. for dealing with the legal and security implications of managing customer wallets).
Additionally, I estimate that it's much more pragmatic for 21 to build a centralized ASIC farm at a location where power is relatively cheap with predictable costs as opposed to a geographically distributed ASIC network that will be subject to regional utility price variance.
Why bother with all that when you could literally just mint money at your leisure?
I think they're betting on typical consumer being unaware of their electricity use. If their chargers and a router suddenly start using 4x the power, they won't notice because it still will be dwarfed by the fridge and washing machine, but scaled up this could bring some amount of money to the company. I'm not sure if that amount of money will be big enough to bother though.
Sounds like a weird variant on a penny shaving scam in that case. Scam a few cents each from a lot of people....
It's also interesting to consider the fact that the efficiency of these devices varies a lot anyway, often because manufacturers try to save money. You might have charger A which wastes 0.1W and charger B which wastes 1W but costs a few cents less to manufacture. Is there a moral difference between charger B and a hypothetical charger C which wastes 0.1W and consumes 0.9W to mine bitcoins?
And the whole thing about customers collecting some of the currency for micropayments seems like a trick to incentivize gullible to prefer devices with this "feature".
RE morality of waste, I must say, you have me here. I need to give it some serious thought. But it feels to me that:
- if you can avoid waste, you should (that would make producers intentionally making wasteful but cheaper products shady, but I do think consumer market is basically flooded by crap and we could use some higher-quality stuff)
- the whole concept of that bitcoin-mining device is malicious, explicitly designed to be taking money from users behind their back, and I think the intent makes this evil
Just because you don't notice that I'm stealing hundredths of a cent from you doesn't make it not stealing.
Charger C is just a different circuit board with different chips on it. It's the marketing framing it as being free, however, that's makes it a dishonest proposition.
Bitcoin is an environmental disaster. What a horrendous waste burning electricity and real, non-renewable natural resources to generate these tokens. For all people complain about the legacy financial system, bitcoin is a real zero-sum game where the more miners join, the harder the block difficulty gets, and the more coal gets burnt.
As long as you capture the heat, it's not a waste.
Also, how much energy does the legacy finance system waste? All the energy spent to build banks, armored trucks, secured networks, etc.
A couple years back they spent a couple hundred million to build a cable connecting New York and London for financial data because it was 5 milliseconds faster than what they had before.
First of all, I don't think we should be talking about waste in terms of money, but energy used. Surely, that NY-London cable was energetically expensive, but a big chunk on that money was probably created out of arbitrary whatever-they-do-on-Wall-Street and used to pay off people and various non-energy-related costs (e.g. land laws).
Secondly, as wasteful as our current financial system is, it doesn't seem to have that strong growth factor built in. Making and maintaining bitcoins depends on ever increasing energy use; energy, which is wasted in the most literal sense - burned out on nonsense computations just to show that you had enough of it to burn.
I think we should treat the energy used on financial system as upkeep - unavoidable cost we have to pay because as a species, we suck at coordinating, but that we'd do best to minimize.
First, in terms of raw energy, I'm pretty sure finance has bitcoin beat.
Second, the energy used by bitcoin doesn't need to increase. The cost of all the energy used will tend towards the price of the subsidy+fees, and the subsidy is being phased out. In a few decades the subsidy will be almost gone.
If bitcoin completely replaced finance, would it really use more energy than saved by finance stopping?
And you didn't address the point about capturing the heat.
> First, in terms of raw energy, I'm pretty sure finance has bitcoin beat.
This is a painfully disingenuous response. It's obvious that the system servicing billions of people and businesses around the world would dwarf the energy tab of a system that services a relative handful of niche users. Why even bother with such a clearly flawed comparison?
> The cost of all the energy used will tend towards the price of the subsidy+fees, and the subsidy is being phased out. In a few decades the subsidy will be almost gone.
Yes, the subsidy will drop but the energy costs will continue to rise and the fees will obviously have to rise to compensate for the absence of subsidies (otherwise the miners will be losing money).
> If bitcoin completely replaced finance, would it really use more energy than saved by finance stopping?
Finance will never "stop". Demand for consumer financial services (like banks, loans, credit & debit cards, etc) would remain pretty much the same, even with bitcoin at the core of finance. The cost of "mining" dollars is negligible compared to the ancillary financial services that people actually care about, so replacing dollars with bitcoin simply adds the extra energy of bitcoin mining into the mix. If bitcoin were cheaper than the legacy system, then why is it that all bitcoin companies prefer to use "offchain transactions" (aka non-bitcoin transactions) whenever possible? The answer is that the blockchain is incredibly inefficient compared to the legacy financial system. Payment gateways like bitpay that have no choice but to interact with the blockchain have spent lots of money building trusted node networks in order to roll their own double-spend detection heuristics because relying on bitcoin's inherent security is too slow for the needs of an actual business. The only problem bitcoin solves is "decentralized" which isn't a problem for most people.
1. The energy "wastage" being complained about is not dependant on bitcoin usage, it's dependant on the subsidy and fees. If bitcoin would be used worldwide, the energy used wouldn't increase linearly with usage. You seem to be assuming it would.
2. Why will energy costs continue to rise? If it becomes less profitable to mine, less mining happens, and less energy is used. This has happened in the past, when the price dropped.
3. Why are companies using off-chain transactions? It's faster for same-company transfers. In a sense, many banks are using "off dollar" transactions as well, not actual dollars. Sending bitcoin is still faster than sending money.
There are plenty of problems bitcoin solves. Just look at all the big-name companies using it, like NASDAQ, for example.
I'm talking about energy, not costs. Money is arbitrary, joules are not, and you can't plug your computer to a stash of dollars and have it run. Generating energy comes at a cost to ecology, to workers and the supply is not unlimited.
RE capturing heat - heat is a low-grade form of energy and electricity is high-grade. You can't power a machine from heat at efficiency levels of electricity. And if you want an electric heater, buy a heat pump - these run at 500% (!) electric to thermal conversion efficiency (by using 1kW of electricity to move 5kW of heat around).
I'm talking about energy as well. In order to project how much energy bitcoin will use, you need to understand the incentives. Could you clarify what you meant by increasing energy use?
Isn't it true, that as time goes on, mining gets harder (= more energy intensive), so if you want to earn bitcoins you need to invest increasingly growing amount of energy? Isn't it also true that this process of mining needs to continue to ensure the integrity and trustworthiness of the blockchain?
No. Mining adjusts so that it always takes 10 minutes to find a block on average. This can happen on a single computer, or on a billion. If the price of bitcoin keeps on increasing, there's a larger inventive to mine, which will cause the difficulty to increase. The marginal profit of mining will always tend towards zero, and the total spent on trying to mine will tend towards subsidy plus fees. The subsidy halfs every so often.
Mining only appears always to increase as time goes on because Bitcoin is increasing in popularity, but the protocol itself doesn't mandate that at all. Like ikeboy said, difficulty is adjusted to keep the average time to find a new block at about 10 minutes. The network is perfectly content to let the rate stay still if no new computational power is added to the network.
I would like to see some real numbers. Finance is 8% of the entire GDP of the US. The transaction volume will eventually go up unrelated to mining power. It seems like a waste to people who don't understand systems of incentives, but it is a small and insignificant price to pay for more ideal money.
How much electricity is really being wasted? Economics works even if some people's reaction is an emotional one instead of data driven.
> It seems like a waste to people who don't understand systems of incentives,
The implication that bitcoin critics are simply incapable of understanding bitcoin's "system of incentives" is laughably condescending as well as totally inaccurate.
> but it is a small and insignificant price to pay for more ideal money.
I don't think "decentralized" necessarily translates to "more ideal". Certainly, it's not ideal for most people given all the other drawbacks that come with bitcoin (difficult to spend, difficult to secure, subject to price fluctuations, subject to privacy concerns, vulnerable to critical mistakes like sending money to the wrong address or losing a wallet in a hard-drive crash). Bitcoin is only "more ideal" if you want to perform transactions outside the legacy system; a real use-case, but not a very popular one.
Decentralized isn't why it has better properties of ideal money. Interestingly the winklevoss twins have made by far the best presentation of the value of money.
One paper estimates that the power consumption of Bitcoin mining is approximately equal to the power consumption of Ireland. My own estimate is that it is the same as the power consumption of Cambodia. Either way, Bitcoin's power usage is basically insane, considering that the transaction volume could literally be handled by a Raspberry Pi running a SQL database.
> It seems like a waste to people who don't understand systems of incentives,
Well, it may be that I don't understand incentives here, but it seems to me that Bitcoin is constructed to incentivize wasting exponentially growing amounts of power.
Yes, legacy financial system is a waste, too, but we should be striving to stabilize and reduce that waste instead of constructing an engine that could happpily raise it to an unprecedented amount.
The missing factor is that it isn't exponential. Right now block rewards are the mining incentive. Eventually transaction volume will go up much higher at no cost to the miners, which will flip the transaction volume to power used ratio.
Legacy financial systems are not just wasteful, they are horrendously wasteful and incredibly inept. It may be the case the competing crypto currencies take over or coexist with bitcoin, but they will have to overcome its substantial network effect and development lead (although it is open source of course).
You could say that about any number of human activities, though. "Geez, all these game conventions use so much power, can you imagine how much coal is being used to power all these lights and booths right now?" If anything, having the monetary system tied to the power efficiency of computing devices should greatly incentivize increasing their efficiency further.
Edit: Downvotes aren't supposed to be used for disagreements, but whatever.
I'd wager a large amount of virtual currency that these are not specifically aimed at virtual currencies, but virtual implementations of real-world currencies.
I can see the advantage for the user (25% of the BTC mined), and I can see the advantage for 21 (75% of the BTC mined) and Intel (sell some chips). But what's the incentive for the electronics manufacturer in embedding these BitSplit chips? If they're not getting a cut and they're not being paid, why would they integrate a chip even if it's free? The downside risk of embedding a chip that may become obsolete, break down, or negatively impact energy efficiency surely outweighs the marketing benefit of saying '21 inside'.
Why is 25% a benefit? If these chips profitably mined bitcoins, it would make sense to run them yourself rather than give them away. 25% isn't going to cover the electricity used.
21 will surely dole out some of their 75% cut to hardware integration partners as a distribution fee. The ripoff here is for anyone ignorant enough to run one of these devices.
The way I read it, the 25% is not going to the user directly. It just simply goes into an account the manufacturer controls and is earmarked for the user. The user will only be able to spend it as the manufacturer allows. This is more along the lines of Google Opinion Rewards rather than being part of a mining group.
I have to wonder how they're going to market these things to people without breaking any consumer laws. It seems like you'd almost have to trick people into using them for any extended period of time (like after they get their first electric bill), which will make them a prime target for lawsuits, Attorneys General of every state, the FTC, Congress, etc.
Intel will advertise these chips, and claim they have many in stock. They will take preorders in Bitcoin. The lead times will balloon out to several months, under the guise of "burn in testing". People will be angry at not getting their boards, and attempt to cancel their pre-orders. As they pre-ordered in BTC, however, they will not receive the same value back as they ordered, due to fluctuations in the price. A handful of people will receive their devices late, and by then the difficulty curve will have ramped up, making these devices not cost effective.
The idea is to stick these chips into household applications (e.g. USB chargers), give the appliances out for free, stick the electricity bill on you, and give 75% of the Bitcoin revenue back to the parent company.
Yep, a free USB charger that you amortize over its entire use, at many multiples of what it should cost. It's great for people who love playing a guaranteed losing shell game between buying electronics and electricity.
I'm becoming more and more convinced that it's impossible to profitably mine bitcoins without screwing someone else in the process. The trick is to screw them as long as possible without them knowing.