Why Can't Computing at the Heart of Bitcoin Be More Useful?
spectrum.ieee.org
spectrum.ieee.org
1. The input need to be unpredictable. Block N is the seed for block N+1.
2. It needs to be difficult to calculate, yet trivially easy to verify.
3. Difficulty needs to be adjustable.
The only idea I've ever seen that met all three of these criteria was PrimeCoin's usage of calculating certain Cunningham twin primes (which is still of unclear/tiny value).
If you don't see any value in that, it's gonna be hard to justify that computation.
Whereas things like Bitcoin and Ether are fungible, meaning there's no difference between my 1BTC and your 1BTC.
This is simply unjustifiable, when thousands of people are actively running the planet to the ground to make a quick buck in an online casino.
Are you talking about the stock market? Or the companies that have already more or less destroyed the planet to extract every last bit of profit?
I'd love to know what your top alternative is
If you put 1000W of electrical energy into a Bitcoin miner, you get 1000W of heat out the far end. It's the same efficiency as a resistive element space heater, just has done some work (useful or not, it's still work) on the way through.
Same thing for a computer running BOINC or Folding@Home. Power goes in, heat comes out, but you've done something useful in the gap.
I typically heat my office (off grid solar shed) in the winter on waste compute - I've got a few computers in here that run Folding@Home/BOINC and I just run them if I've got surplus power. I have propane backup for the really dark grey days ("generator days" if they extend past about a day - it's cheaper to run a gallon of gas through an old generator than to radically expand my battery bank), but most of my winter heat comes from running F@H/BOINC. To the point that on a clear, sunny winter day, I have my window open and fans sucking cold air in or I'll roast in here. I can hold about 1.8kW of load no problem on a clear day.
Now that I have solar on the house, I've considered adding an old compute rig or two in there for winter heat. It's less efficient than our heat pump on warm days, but we also get some cold, grey, foggy days in which the heat pump likes to ice up badly. Normally, if the heat pump isn't keeping up with things, the thermostat will call for the backup coils (which are just big resistors) to aid, and they're the same efficiency at turning electricity into heat as a couple computers. Old Xeons aren't amazingly power efficient, but you can also get them for free or nearly so and throw compute at your preferred projects.
One can optimize for a wide variety of things - energy spent, compute performed, Bitcoin mined, minimum cost, etc. It works better, and is a lot more efficient, if you look at everything as a system and integrate it. I don't run BOINC tasks on my homeserver when we're in air conditioning season, but I light up a couple cores of them in the winter.
If you put 1000W into a Bitcoin miner, computer, resistive space heater, etc, you get 1000W thermal out.
If you use that 1000W to turn the compressor in a heat pump, you get far more heat out (3-4x is a reasonable average in a lot of areas), because you're not simply generating heat from the energy - you're using the energy to move heat. A heat pump is an air conditioner in reverse - you cool the outside air and heat the inside air.
A standard CPU dissipates heat as a resistive heater would - you pump 100W into the CPU, you get 100W of heat out. I'm not sure what you're using to claim that a CPU "is literally a heat pump" here - it's not, by any standard definition of a heat pump.
And heat pumps are vastly more efficient than resistive heating (or using the heat from a microchip).
In many places, for every 1W of power a heat pump consumes, it can move 4W or more of heat into the space.
However, you're correct about coils icing up. That impacts some regions more than others. I'm in a fairly dry high desert climate, and our coils don't ice up unless we have fog (at which point, yes, they ice badly). There are defrost cycles that reverse the unit and melt the ice, but it is a problem in climates where it tends very humid in the winter.
However, there are a lot of places where it works fine, and one can always use a dual fuel setup, where a heat pump is used down to whatever temperature it starts having problems, then switch over to something else (gas fired furnaces are the usual backing option) in the extreme cold. You still get the energy savings of the heat pump while it works well, but can keep a house warm down to quite chilly temperatures.
Of course, if it gets really cold, a ground source unit becomes worth looking at (heat exchange with the ground, either via a deep well or a bunch of coils under the yard).
This also doesn’t address the wasted materials and labor used to make the chips.
/s
That's not necessarily always the case. There's a threshold where the delta between the outside ambient temperature and desired indoor temperature is too large, it becomes more efficient to use something like resistive heating rather than a heat-pump
(In other places and to a larger extent pop culture, "geothermal" refers to tapping into actual heat sources underground.)
No. The value of a PoW coin is based on the wasted electricity.
If mining does something useful (like heating) for a large fraction of the miners the mining difficulty will go up to adapt to such change, requiring additional mining.
At the end of the amount of wasted electricity will remain the same.
I have a degree in EE.
The wasted electricity is in additional heat that you don't want in your home and you have to vent.
Suppose millions of people could replace resistance heaters with mining rigs that cost exactly the same and consume exactly the same amount of energy every day.
Such people would be running the miners ever if the financial gain is very close to zero, because they were already using the same amount of electricity for heating.
That means they would push the mining difficulty up. But, if the demand for coin remains the same, other people will run mining rigs that releases increasing amount of heat in the environment.
You just pushed the amount of wasted heat from one place to another.
Regarding heating things with electricity, a heat pump is way more efficient than resistance/mining heating because is moving existing heat from one place to another.
The approximation for electrical use in heating is using heaters that can output roughly 10 watts per square foot of your house. 12-13 if your house is poorly insulated.
At full power, that means that a 1,600 sqft house requires 16 kW of electricity. At a cost of 13¢ per kWH, that's about $2.05 for that 16kW of electricity. Of course, no residential house will ever realistically have 16kW available to it, nor would heating ever require a 100% duty cycle. But this is for comparative purposes.
The equivalent amount of gas power would require about 54,000 BTUs from a gas furnace (about 3.4 BTUs per watt). One cubic foot of natural gas provides 1,032 BTUs which means that generating 54,000 BTUs of heat requires about 52 cubic feet of gas (we'll round that up to 65 to account for inefficiencies of a furnace; top end furnaces are in the 80-80% efficiency range). 100 cubic feet of natural gas costs about $0.48, so the cost for natural gas would be around 30¢
Again, neither will (nor could, nor should) run constantly, but it's a sobering comparison.
EDIT: Comparisons were hours vs. seconds; fixed.
They simply periodically pump a bit of heat back out to keep the exchanger free of ice.
You'll periodically need a backup source of heat, but most of the time you'll be fine with the heat pump alone.
Define "cold". Mitsubishi units have 100% of capacity at 23F/-5C and 75% capacity at -13F/-25C:
* https://www.mitsubishicomfort.com/benefits/hyper-heating
They're used in Alaska:
* https://www.nrel.gov/news/features/2021/even-in-frigid-tempe...
Fujitsu says -15F/-25C as well:
* https://www.fujitsugeneral.com/us/residential/technology/xlt...
At the 'break down point' an air source heat pump (HP) is no worse than a baseboard heater: in both cases the COP is basically 1.0. So you might as well go with the HP as it'll be more efficient for probably the majority of the year.
You can get gas furnaces that are 'dual fuel': they'll run the HP refrigerant through the exchange and circulate the conditioned air, and once it's "too cold" switch to gas.
If you're going to buy a only-cooling AC unit, then spending a little more on a heat-cool HP is probably a smart idea. There's a payback period, and it depends on the price of your electricity and fossil fuel (NG, propane).
It’s an elegant exploitation of the design reasons why the work must involve something like random search (for an easy to validate target).
On squinting, it’s kinda like a smart contract for a compute job, where the compute job itself constitutes proof of work to validate the settlement of the contract!
It could be the fact that it's not trivially easy to verify the work completed, unlike a hash.
The arbitrarily parameterized instances part is really important. In Bitcoin this is done by feeding prior blocks as input into the hash and then getting the hash to match the adjustable difficulty value, so you can generate new problems automatically based on the history.
The Stack Exchange answer that someone linked to here mentions this issue by saying that, if you just had a generic and deterministic useful problem, people could precompute solutions to relevant instances of it, which would amount to premining on the blockchain.
The "doing work to verify transactions" aspect comes in when you can make the problem instances actually encode a specific version of history. It's not obvious how to do that where each and every instance would be of intrinsic interest or value to someone.
Maybe there's some kind of physical simulation where this could be the case (like, each individual hash represents a particular randomly chosen chemical from a very, very large family of chemicals whose properties are being explored by simulation in some way, or something), but I'm still not sure how you could then make a successful result be easier for other people to verify without repeating the same amount of work. (I'm not familiar with a technique for proving, in an easily-checkable way, that the result of a complex physical simulation was such-and-such.)
https://gridcoin.us/ https://gridcoin.us/guides/whitelist.htm
https://gridcoin.us/guides/whitelist.htm
I haven't totally gamed it out but I think the committee could break Gridcoin by means such as refusing to approve any projects, or approving only "captive" BOINC projects that agree to distribute rewards the way the committee wants. Alternatively, BOINC itself could get shut down, or all of the currently-approved projects could have their credentials taken over by someone else and start issuing rewards in a sketchy way.
As I understand it, Gridcoin basically has an identified group of oracles who together appoint and unappoint other oracles who are empowered to issue rewards. While this is clever, it's not very decentralized or permissionless.
(Maybe I should pick on BOINC itself even more than the committee as the "authority" here, since I think BOINC's infrastructure is directly used by the whitelisted projects to assign tasks and announce rewards for completing them. So BOINC is really in a position to tamper with every part of the reward process.)
Hasn't this comparison been made over and over, and every time it comes out in favor of traditional banking by a mile?
https://www.bloomberg.com/opinion/articles/2021-01-26/is-bit...
Other cryptocurrencies that use less energy can be used to transfer money to Venezuela so Bitcoin still loses against other cryptocurrencies for fringe use cases.
You also have to do the same thing for Bitcoin if you want the comparison to remain fair. Bitcoin doesn't magically replace the entire financial industry, so you don't want to accidentally include the costs of the things it doesn't replace (such as the bank manager who's deciding whether or not to authorize a mortgage).
BTC uses more annual energy than the Netherlands. What did it accomplish when compared to the economic output of the Netherlands?
There is more to Proof-Of-Work than just hashing [1].
[1] https://cryptorials.io/beyond-hashcash-proof-work-theres-min...
People need to compute things asynchronously. And they are probably willing to pay. Being able to distribute those out as part of the the mining/validation process would make the energy less “wasted.”
I’m not sure why the grid computing schemes of the 00s didn’t pan out, I suspect aws was just simpler and cheaper than trying to assemble networks of nodes using Globus or whatever.
Hashing for say, Bitcoin, works because you're finding a hash that has certain properties (e.g., having a certain number of leading zeroes). Anyone can take the input (to the hash function) you've found and trivially test that it does indeed produce a hash with those properties. Finding the hash is hard, but testing the hash is easy.
Doing SETI/Folding work doesn't work like that. Consider Folding: every unit of work produces useful output. Every input produces equally valuable output. Additionally, the cost of producing the output is high, but the cost of verifying that output is also high.
With hashing-based PoW, everyone in the network can verify that a hash with the correct properties was produced. There's no way to say "yes, this one Folding@home worker found something unique". The work needs to be checked, which means a majority of nodes in the network need to verify the output, which means >50% of the network needs to compute the same thing again (you can't rely on less than a majority, or anyone could cheat).
https://www.forbes.com/sites/michaeldelcastillo/2020/08/06/v...
(of which the top 3 already account for more BTC than Elon) also unwise?
BTC is inefficient (OP) to
why did Elon buy (you) to
understanding the difference between rich and ethical (last comment) to
rich investors being wise or not (you again)
To make it short: Just because Elon or rich investors buy doesn't mean they are unwise (for themselves) just that it's bad for the world (ecologically), so basically both can be true. Just the difference between thinking about themselves and thinking about other people as well...
Bitcoin uses less electricity than Christmas lights and provides a completely electric, transparent, uncensorable currency.
You could power Christmas lights for several Christmas if not a lifetime with the electricity used in a BTC transaction. Let that sink in.
People way overestimate how much energy Bitcoin uses and way underestimate how much simple things like christmas lights, video games, hockey rinks, etc. use.
Mining rate and therefore energy costs are based on the blockchain mining parameters, not the number of users. If there were a ton more users, you'd likely see more transactions on the Lightning network, with Lightning nodes settling out on the main Bitcoin blockchain, but it wouldn't impact energy use substantially.
About the only thing that impacts energy use at this point is the price. One can reasonably estimate that, steady state, the value of the energy used is roughly equal to the total block rewards.
Bitcoin doesn't work that way. Energy usage doesn't scale anywhere near linear with the number of users.
Think of it this way: the Bitcoin network is buying security, not transactions, from miners.
If gold was invented today, what would it look like, and how would it function?
I'm not saying it works particularly well yet but the idea is sound.
But there are literally thousands of those!
As for the banking system, Bitcoin is part of it now. Banks offer cryptocurrency-backed accounts, and if it becomes big, you will probably get all the stuff that banking does like loans, escrow services and fast, reversible transactions. There will be regulations and I guess a fractional reserve system will follow.
And how is it uncensorable? All it takes for a state to censor it is to declare it illegal with harsh sentences for those who get caught and embargoes on countries that don't play along.
As for Christmas lights, I think you are way off. Care to share the numbers.
The value of decentralized system far outweighs the cost of maintaining it especially when it comes to money. Could bitcoin be more efficient? Absolutely, but to call it big mistake of modern technology is just silly.
It takes time to build/deploy/grow these exchanges (since they are essentially a 2 sided marketplace).
Edit: There's no reason to downvote OP's post, folks. It's a good question that some people may not know the answer to.
Did you buy the top or something?
It gets hyped, it falters, people stop talking for a while, and the cycle repeats itself.
If not, then maybe just step back and observe the chain grow block by block, day by day, unfettered, for the rest of your life.
This reads more like a shallow shill-piece for the people in the article. It's odd to see it at IEEE.
None of the presented angles would work for something like Bitcoin. Without another native currency, how do you assign value or determine usefulness of, say, models? No matter how you slice it, if you're going to make it into a viable PoW-replacement you will need some kind of fundamental theoretical breakthrough that none of these have. TEEs are definitely not it. The status quo is otherwise that it is an intractable proposal (which other commentors have covered already)
I have a feeling that the author doesn't fully understand what they're writing about and are acting as a useful idiot so some guys can get free PR.
"Please don't post shallow dismissals, especially of other people's work. A good critical comment teaches us something."
e.g. one can be amazed by nuclear physics yet dislike nuclear weapons. Or appreciate chemistry by hate meth cooks.
- The usage of electricity in Bitcoin is a feature not a bug. It's designed to use the most democratic and ubiquitous commodity avalible. Any nation/individual can evaluate their electrical opportunity cost and at any point make the decision to dedicate resources to towards re-balancing the network without permission. This is not possible with proof of stake.
- We produce about 160,000 TWh of power, ~50,000 TWh is wasted due to inefficiencies. Bitcoin using 120 TWh (0.25% of wasted power)
- A lot of Power projects get scraped due to inconsistencies in demand. Random Ex: Texas power outages during summer peak AC usage. It's infeasible to spin up and spin down extra generators to meet spikes in demand. #Bitcoin fixes this by providing excess supply a discounted demand during normal hours that can be routed away during peak needs else where. Ditto for renewables.
- Bonus: SHA-256 did not get broken this week.
- You live in a place where the "cost" of mining = Electricity + Amortized cost of mining rig.
- Someone else lives in a place where "cost" of mining = (Electricity + Amortized Blackmarket Rig Cost) - (5% Inflation per month - 30% reduction in remittance fee -etc..)
> Random Ex: Texas power outages during summer peak AC usage. It's infeasible to spin up and spin down extra generators to meet spikes in demand. #Bitcoin fixes this by providing excess supply a discounted demand during normal hours that can be routed away during peak needs else where. Ditto for renewables.
Could be, but are the price signals really in place to make this out work in practice in many places? How many electric utilities have so far succeeded in charging people more at all during peak demand times? (maybe other than electric customers they've identified as industrial)
Edit: To be clear, I agree that there is such a thing as surplus power generation and usefully smoothing out demand, I just don't agree that we're in a scenario where cryptocurrency mining is predominantly used or predominantly incentivized to be used that way.
This is true and is a function of context and locale. That's the point. Opportunity cost.
- Actually Bitcoin mining is being adopted by renewable providers as very good solution to offset intermittence and wasted excess.
- Further something like Natural gas flaring is something that cannot be stored but can largely routed through mining setups. Which is being adopted at scale.
Which is why Bitcoin mining is only feasible in the areas that have the lowest electricity rate?
> We produce about 160,000 TWh of power, ~50,000 TWh is wasted due to inefficiencies. Bitcoin using 120 TWh (0.25% of wasted power)
How much of those inefficiencies are caused by Bitcoin? My understanding is that a substantial fraction of that inefficiency is things like "there's power loss in running power from a power plant 100s of miles away" or "converting from AC to DC power induces power loss." In other words, Bitcoin isn't converting electrical waste into useful work, it's just creating more electrical waste on top of the power it directly uses.
> A lot of Power projects get scraped due to inconsistencies in demand. Random Ex: Texas power outages during summer peak AC usage. It's infeasible to spin up and spin down extra generators to meet spikes in demand. #Bitcoin fixes this by providing excess supply a discounted demand during normal hours that can be routed away during peak needs else where. Ditto for renewables.
Are you aware that the electrical grid already manages demand in large part by telling the smelters to stop smelting for a few hours? There's no need to invent new demand to manage existing demand.
Nope, not where Electricity is cheapest, but where it's opportunity cost is lowest. In developed nations with a stable currency this might be high, in others that don't have the privilege of printing the world's reserve to bail out bad decisions, different story. And that's what the market is reflecting.
>My understanding is that a substantial fraction of that inefficiency is things like "there's power loss in running power from a power plant 100s of miles away" or "converting from AC to DC power induces power loss
Precisely, most Bitcoin mines are remotely located right next to the power plants, partially for this reason.
> There's no need to invent new demand to manage existing demand.
No you need demand to bring excess supply.
There is this intuition that guessing a bunch of random numbers doesn't contribute high-leverage information or "work" to society. That there are a bunch of "information processing" tasks that have more inherent value.
But consider SETI@home - is decoding massive amounts of space noise valuable if nothing ever comes of it? Is the computation wasted?
Consider Folding@home - there are now far superior algorithms like AlfaFold that take far less compute. Is the computation wasted?
Consider all the matrices being multiplied on gigantic neural networks. Some of the resulting models will be enormously valuable to society. Most of it will be dead-on-arrival experiments - simply wasted compute. Even worse, the leverage can go the other way - a ML model can cause far more harm than the energy footprint used to train it.
One of the nice properties of the BTC PoW function is that you can be sure that your hash finding algorithm is probably not that sub-optimal.
We need to stop thinking that tech inherently neutral and without moral or ethical implications when there are clear benefactors and externalities.