THere's no rush to mine seti data. it doesn't matter if we detect alien signals now or in 20 years. so just store it and then mine it all in a single day with quantum processors in 20 years.
SetiCoins are also a good idea. But How to make this practical? the characteristic needed is a calculation that is hard to do but easy to check like factoring or hashing. You can't make detection of a signal be the mining event because that's never going to happen. You can't just make it some bolus of processing in which two parties agree as they might collude. So what about the seti signal is hard to do but easy to check?
Yes lets take the most rock solid pillar of internet communication and extend it so it only works right in Chrome Broswers.
What a Microsoft-1990s move. It's the reason everyone hated microsoft for a decade. Embrace and extend.
But they never had the gall to go this big. Why not embrace and extend TCIP too google? They already are doing DNS so it wouldn't be that hard. Facebook's VPN might give it a whirl too.
Does a vpn provide much security over https? Perhaps a vpn also shields the dns lookup from scrutiny. But in this case Facebook is pilfering your vpn lookups too, so thatâ(TM)s moot as a distinction.
Whatâ(TM)s the benefit of vpn in the age of https everywhere
If you have an authentication server why do you need or even want block chain. Furthermore if you want to distribute the authentication to many servers how do you control the authentication list if there's no proof of work. and if there's proof of work, then it gets expensive because that's why its called work
the not "mushy" sound is interesting because one common reason for a device to sound mushy is because the device is non-linear so while it could play any one frequency in a "flat" way, it can't play them all at the same time. This is almost dictated by the physics. And this is why, in part, why speakers that divide up the base and trebble to different speakers do better.
A related phenomena is the direction of sound. a given speaker may angularly radiate different frequency bands differently. Again giving both mush and imbalance in a room.
the fact the the apple speaker is reported to have clarity, directional sound, and self-equalization based on the room suggests they are actively compensating for the non-linearities.
here's the key sentence in the consumer reports review that pretty much invalidates the review: Consumer Reports evaluates sound quality for speakers, smart or otherwise, in a dedicated listening room in which our experienced testers compare each model with high-quality reference speakers. Each test unit that allows for user controls is tuned for optimum sound quality—we want the speakers to sound their best.
What matters is how it sounds in a typical room with typcal variations. hard reflective surfaces or absorbing couches.
The most striking six comments I have read in almsot every review of this device are consistent: 1. It has stereo-like directional sound effects. 2. It sounds good in any room. Somehow it is adpating 3. There is no "sweet" spot but sounds good in most places in the room. This one is very intriguing. 4. It's definition is very good. not mushy sounding like typical devices. 5. remarkable base that doesn't turn to buzzing when cranked to max volume 6. the ability to command siri is not affected by how loud the sound is, and works at great ranges.
It "sounds" like this is a sonic engineering masterpiece.
By the way, bose also does sonic engineering to make inexpensive parts sound great. I would think that computer people would actually appreciate this since it's effectively what microchips did-- replace expensive bespoke electronics with universal cheap electronics via free-to-replicate software engineering.
I keep finding some as malware like Spigot or Conduit has wormed itself into my broswers. The way that crap works is it's capturing and editing all your transmissions and receptions to analyse content, inject ads, and to track correlations across sites. So even if you think you are clean you may not be and even if you are using your personal computing devices they very well can correlate you back to your work devices just by the websites you visit.
I really really hate these adware companies. They are like computer herpes that can't be gotten rid of.
Facebook probably will have some way to see you are using a covert app, and probably be able to see other people who are using the app at the same time. Likewise google, may verywell be the DNS that originates this. And finally how can you be sure that Signal or Skype doesn't share it's transaction analysis with its "partners and customers" as the EULA you didn't read might say. And perhaps the person you are talking to is also taking notes in google docs, etc..
De-anominization isn't that hard. More to the point, you can't really know ahead of time if is. And finally, your early contacts with a reporter probably don't happen via covert channels. Few people plan ahead like Snowden.
It comes with a warranty. If you want you can pay $40 more for the extended warranty. Or if you are like many people your Visa card already doubles the free warrantee. And of third party sellers will sell you an extended warranty less than apple too.
It's an appliance. I don't see people whining that the repair cost of the dishwasher is close to the cost of a new one.
Apple has a history of dropping old well established standards about 2 to 3 years before people can see they were right. It's actually a bit uncanny how good they have been done guessing correctly.
a few example: Parallel port RS-232 port dropping support of all printers except postscript (which is what let the original mac/lisa clear the way for real WYSISWYG for the first time on third party printers. a BFD) floppy drives ADB.... everyone thought these moves were crazy at the time. so much legacy support was dropped. the they were always right in the end for the following reason:
by requiring that any device plugged in met some higher level of service expectation they could write software that took advantage of that requirement sooner than their competition who had to have legacy support. I give you the WYSIWYG revolution as exhibit A.
Someone who runs a website actively promoting illegal activities needs to be squeaky clean to get away with it. If you are making money off it, it's going to look bad. I don't know the Danish laws but in the US there's only so much you can hide behind on the first amendment which protects free speech not commerical speech.
Joce640k
I'm the grand parent poster here and you raise an interesting suggestion that I have considered thoughtfully and think maybe is not true. While the Double Spending scenario is correct, the "taking other people's bitcoins" argument may not be correct. I am pretty sure there's only two things you can do under the double-spend attack. One is to recover your own coins and the other is to vandalize other people's transactions. But what you cannot do it re-direct anycoin you did not own at some point to yourself. you can't take other people's coins (if they did not originally come from you).
here's why. When you spend a coin two things have to be true. 1. the blcok chain shows the coin is assigned to your public crytpto key 2. you can sign the transfer with the transfer you want to intitate with your private key.
a double spender doesn't know your private key. so they cannot take your coin.
what they can do is erase that transaction (so you never spent it) or they could erase the transaction where that coin came into your wallet in the first place (so you lose it). But that erasure won't in general transfer the coin to the doublespender but just to some previous holder of the coin.
The double spender can "steal" but it can only do so by resetting it's wallet back to an earlier state. So it's possible the erased transactions might take coins you got from the double spender. but the double spender can't actually make that historical wallet size bigger.
Caveat: I suppose one strategy would be for the doublespender to briefly buy every coin in existence (borrow some money to do it then sell the coins to pay back the borrowed money). then they could continually reset the chain back to that time when they owned it all.
Note that I did not address your argument that all the potential double spenders would keep each other at bay. this isn't logical. If they are held at bay, no one is making a profit by double spending. Yet they are also overinvested in mining gear and so can't make a profit on fees either. That situation can't persist. People won't just invest in excess mining gear just to ward off some other potential double spender at no profit to themselves.
consider the rational miners for the sake of argument and we can mop up the edge cases later.
You are right that there's no rational reason to chase 51% of the bitcoin hash rate just for the value of the coins/fees. But that actually is why it is profitable for the double spender to do just that. To see this realize that the mining coins or fees are going to be far less in value than the amount of value being transacted in the mining event. vastly so when it becomes fee based-- otherwise you would be paying more in fees to do a transaction than the transaction (and since the transaction includes the fee that becomes a tautology.) agreed?
From that Agreed logic we see that it would always be more profitable to double spend than to mine. e.g. if the bitcoin transaction included your purchase of a home for $1M and the fees on the entire bitcoin epoch including all ancilliary transactions was say $100,000 then double spending beats any fees by a longshot. (here doublespending means you get your $1M back but still own the house).
So... the number of miners is deternined by that $100,000 fee. getting 51% of that pool might not be worth the capital investment. So no rational reason to chase excessive mining capacity. But for the doublespender there's a huge payoff for getting 51% of the capacity regardless of the fees. It's not just $1M one time, but it's $1M on every transaction... forever.
I hope that clarifies the argument.
You can now see that logic dictates that the cost of achieving 51% needs be higher than the payoff (which will be as many of these $1M transactions as you can get away with). There is nothing in the bitcoin system that adjusts the cost of POW to the magnitude of the transaction. It does adjust to the fees which might indirectly reflect the transaction value but this is not a direct coupling as it will be clamped by competition from other fee based systems (like for example, writing a cashier's check or using a visa card).
If you take my made up numbers here at face value (not correct but makes the point) then if there's a 10 to 1 more profit in double spending, any irrational actors competing for that "1" will still not matter.
As I noted originally, I'd love it if you could prove me wrong, but I don't think the arguments you gave do this.
How might this come about? well I see four scenarios as nearly certain to come along at some point There are probably many others I haven't considered.
1. Due to distorted economics like subsidized electrical power or cheap cooling, some countries accumulate a dominant share of the mining. Then for some reason the government of that country kills the market (e.g. remove power subsidies or siezes the machines) for mining. Suddenly there's void of miners that someone with a big cluster (amazon?) could step into.
2. due to geopolitical tensions (a war?) one country with a small number is cut off from the rest of the world's miners. Again now some small player witha lot of CPUs can dominate the ledger for that country. We see internet cut offs in Iran and Syria and other places now and then. And there's growing numbers of firewalls around countries. IMagine if China and the US came to blows in the south China sea or North korea executed a cyber blitz kreig on south koreas. Isolations of the internet would occur.
3. The reward system for bitcoin changes with time with fees replacing mining. What if the fees don't attract enough miners? that is say, as fees rise the number of transactions will fall. there is some equliubrium there. That equilibrium sets the POW difficulty and the number of miners who can make a profit. Do we know that equilibrium is sufficient to protect the transaction value of a double spend? I don' think there's anything built into the model to assure this.
4. What if some new crytpocurrency suddenly becomes more profitable for miners and they point all their ASIC's and GPUs at that leaving bitcoin a ghost town with just a few miners. Here's one such scenario. Someone, maybe etherium can, will figure out how to make the wasted effort of the POW actually valuable in itself. e.g. make the POW be the travelling salesman problem of delivery planning that UNited Parcel Service has to solve every day. Now you get paid for the POW by the currency holders and also get paid for the work you did for UPS. this will beat out coins where the POW is worthless as a calculation.
There are a lot of complaints one can lob at cryptocurrencies but I want to address just one that at the moment I see as ultimately fatal to all proof of work systems. Since this fatality hasn't actually materialized yet I have to wonder if I'm wrong about it being unavoidable in the end. But I don't yet see how I am wrong so here goes:
The ENTIRE magic and near Genius solution that bitcoin and others perform is the avoid the "double spend" problem when there is no central authority to manage a secret signing key. Solving the double spend with distributed signing is the magical part.
The DOuble spend problem is that in an a normal distributed ledger that anyone can write to a bad actor could spend a coin, see it entered in the ledger as beloging to the seller, then after getting the benefit of the sale from the seller, re-write a newledger in which that spend never happened. The bad actor can then re-spend the same coin.
Block chain by itslef doesn't solve this. it's just a ledger format. But when you add the proof of work part then you have an escalating difficulty barrier. The seller waits to see the transaction is confirmed. If they are paranoid, they could even wait for several more epochs of chain extension. At that point if the bad actor wanted to re-write history they would have to create a new block chain that was longer than the currently accepted one. And that would be hard because of the multiple epochs of proof of work.
this is exactly why the fear in bitcoin is that if one person accumulated enough mining power they could execute a double spend. But since this is addative: it takes 3x more mining power to unwind 3 layers of the block history (and N-1 x more mining power than the world, to unwind N layers), it's hard.
Or rather it's hard, but only if the world has a lot of miners. If miners lose interest two things happen. First it becomes easier for the bad actor to accumulate enough CPU power to overwhelm the rest of the world's miners. Additionally, since Bitcoin in particular scales the difficulty to the transaction rate it also requires less and less CPU power to do this. (as miners lose interest and so the POW difficulty goes down)
Thus Cryptocurrencies only protect all that capitalized outstanding wealth only as long as their's an active pool of miners. If that goes away then the protection of the blockchain is gone and the double spend re-emerges. at that point it gets crazy.
SO why might this not have happened yet. I think maybe it's because the cost of mining a coin is so high that the cost of mining 2,3, or 4 coins to unwind 1,2, or 3 layers might not have been worth the gained value of the doublespend. But that' now. As the coin becomes increasingly capitalized then a lot of wealth will be transacted in each epoch (and the lightning network is amplifying this now). Thus the temptation for a fouble spend will eventually exceed the cost.
At that point there is a total heat death of the currency as no one can trust it.
Once the miners stop buying as you put it, things will normalize to zero
Slashdot Mirror
but then SETI will say they can't afford to buy RAM chips.
THere's no rush to mine seti data. it doesn't matter if we detect alien signals now or in 20 years. so just store it and then mine it all in a single day with quantum processors in 20 years.
SetiCoins are also a good idea. But How to make this practical? the characteristic needed is a calculation that is hard to do but easy to check like factoring or hashing. You can't make detection of a signal be the mining event because that's never going to happen. You can't just make it some bolus of processing in which two parties agree as they might collude. So what about the seti signal is hard to do but easy to check?
Yes lets take the most rock solid pillar of internet communication and extend it so it only works right in Chrome Broswers.
What a Microsoft-1990s move. It's the reason everyone hated microsoft for a decade. Embrace and extend.
But they never had the gall to go this big. Why not embrace and extend TCIP too google? They already are doing DNS so it wouldn't be that hard. Facebook's VPN might give it a whirl too.
Does a vpn provide much security over https?
Perhaps a vpn also shields the dns lookup from scrutiny. But in this case Facebook is pilfering your vpn lookups too, so thatâ(TM)s moot as a distinction.
Whatâ(TM)s the benefit of vpn in the age of https everywhere
If you have an authentication server why do you need or even want block chain. Furthermore if you want to distribute the authentication to many servers how do you control the authentication list if there's no proof of work. and if there's proof of work, then it gets expensive because that's why its called work
I feel old, that I actually know what you are joking about.
the not "mushy" sound is interesting because one common reason for a device to sound mushy is because the device is non-linear so while it could play any one frequency in a "flat" way, it can't play them all at the same time.
This is almost dictated by the physics. And this is why, in part, why speakers that divide up the base and trebble to different speakers do better.
A related phenomena is the direction of sound. a given speaker may angularly radiate different frequency bands differently. Again giving both mush and imbalance in a room.
the fact the the apple speaker is reported to have clarity, directional sound, and self-equalization based on the room suggests they are actively compensating for the non-linearities.
if so $350 is shocking cheap!
here's the key sentence in the consumer reports review that pretty much invalidates the review:
Consumer Reports evaluates sound quality for speakers, smart or otherwise, in a dedicated listening room in which our experienced testers compare each model with high-quality reference speakers. Each test unit that allows for user controls is tuned for optimum sound quality—we want the speakers to sound their best.
What matters is how it sounds in a typical room with typcal variations. hard reflective surfaces or absorbing couches.
The most striking six comments I have read in almsot every review of this device are consistent:
1. It has stereo-like directional sound effects.
2. It sounds good in any room. Somehow it is adpating
3. There is no "sweet" spot but sounds good in most places in the room. This one is very intriguing.
4. It's definition is very good. not mushy sounding like typical devices.
5. remarkable base that doesn't turn to buzzing when cranked to max volume
6. the ability to command siri is not affected by how loud the sound is, and works at great ranges.
It "sounds" like this is a sonic engineering masterpiece.
By the way, bose also does sonic engineering to make inexpensive parts sound great. I would think that computer people would actually appreciate this since it's effectively what microchips did-- replace expensive bespoke electronics with universal cheap electronics via free-to-replicate software engineering.
eyhay, etslay oolfay ethay inguistlay
Sally it uses both whitespace and braces blocks, and has glottal sounds for sigils like $, @ and %
if you are P2P with no intermediary to tumble the connections then every router, gateway, backbone, and ISP knows exactly who is talking to who.
I keep finding some as malware like Spigot or Conduit has wormed itself into my broswers. The way that crap works is it's capturing and editing all your transmissions and receptions to analyse content, inject ads, and to track correlations across sites. So even if you think you are clean you may not be and even if you are using your personal computing devices they very well can correlate you back to your work devices just by the websites you visit.
I really really hate these adware companies. They are like computer herpes that can't be gotten rid of.
Facebook probably will have some way to see you are using a covert app, and probably be able to see other people who are using the app at the same time. Likewise google, may verywell be the DNS that originates this. And finally how can you be sure that Signal or Skype doesn't share it's transaction analysis with its "partners and customers" as the EULA you didn't read might say. And perhaps the person you are talking to is also taking notes in google docs, etc..
De-anominization isn't that hard. More to the point, you can't really know ahead of time if is. And finally, your early contacts with a reporter probably don't happen via covert channels. Few people plan ahead like Snowden.
It comes with a warranty. If you want you can pay $40 more for the extended warranty. Or if you are like many people your Visa card already doubles the free warrantee. And of third party sellers will sell you an extended warranty less than apple too.
It's an appliance. I don't see people whining that the repair cost of the dishwasher is close to the cost of a new one.
Nothing to see here but apple bashing.
Apple has a history of dropping old well established standards about 2 to 3 years before people can see they were right. It's actually a bit uncanny how good they have been done guessing correctly.
a few example: ....
Parallel port
RS-232 port
dropping support of all printers except postscript (which is what let the original mac/lisa clear the way for real WYSISWYG for the first time on third party printers. a BFD)
floppy drives
ADB
everyone thought these moves were crazy at the time. so much legacy support was dropped. the they were always right in the end for the following reason:
by requiring that any device plugged in met some higher level of service expectation they could write software that took advantage of that requirement sooner than their competition who had to have legacy support. I give you the WYSIWYG revolution as exhibit A.
I hear that to expand the readership they will have the horse/western aspect fold in bronies.
Really, I love firefly but I love it as it is.
Someone who runs a website actively promoting illegal activities needs to be squeaky clean to get away with it. If you are making money off it, it's going to look bad. I don't know the Danish laws but in the US there's only so much you can hide behind on the first amendment which protects free speech not commerical speech.
Are subs considered LCS?
Well no. You can prevent me from selling them. But you can't change who signed them. So no you can't assign the coins to yourself arbitrarily.
Joce640k
I'm the grand parent poster here and you raise an interesting suggestion that I have considered thoughtfully and think maybe is not true. While the Double Spending scenario is correct, the "taking other people's bitcoins" argument may not be correct. I am pretty sure there's only two things you can do under the double-spend attack. One is to recover your own coins and the other is to vandalize other people's transactions. But what you cannot do it re-direct anycoin you did not own at some point to yourself. you can't take other people's coins (if they did not originally come from you).
here's why. When you spend a coin two things have to be true. 1. the blcok chain shows the coin is assigned to your public crytpto key 2. you can sign the transfer with the transfer you want to intitate with your private key.
a double spender doesn't know your private key. so they cannot take your coin.
what they can do is erase that transaction (so you never spent it) or they could erase the transaction where that coin came into your wallet in the first place (so you lose it). But that erasure won't in general transfer the coin to the doublespender but just to some previous holder of the coin.
The double spender can "steal" but it can only do so by resetting it's wallet back to an earlier state. So it's possible the erased transactions might take coins you got from the double spender. but the double spender can't actually make that historical wallet size bigger.
Caveat: I suppose one strategy would be for the doublespender to briefly buy every coin in existence (borrow some money to do it then sell the coins to pay back the borrowed money). then they could continually reset the chain back to that time when they owned it all.
Note that I did not address your argument that all the potential double spenders would keep each other at bay. this isn't logical. If they are held at bay, no one is making a profit by double spending. Yet they are also overinvested in mining gear and so can't make a profit on fees either. That situation can't persist. People won't just invest in excess mining gear just to ward off some other potential double spender at no profit to themselves.
consider the rational miners for the sake of argument and we can mop up the edge cases later.
You are right that there's no rational reason to chase 51% of the bitcoin hash rate just for the value of the coins/fees. But that actually is why it is profitable for the double spender to do just that. To see this realize that the mining coins or fees are going to be far less in value than the amount of value being transacted in the mining event. vastly so when it becomes fee based-- otherwise you would be paying more in fees to do a transaction than the transaction (and since the transaction includes the fee that becomes a tautology.) agreed?
From that Agreed logic we see that it would always be more profitable to double spend than to mine. e.g. if the bitcoin transaction included your purchase of a home for $1M and the fees on the entire bitcoin epoch including all ancilliary transactions was say $100,000 then double spending beats any fees by a longshot. (here doublespending means you get your $1M back but still own the house).
So... the number of miners is deternined by that $100,000 fee. getting 51% of that pool might not be worth the capital investment. So no rational reason to chase excessive mining capacity. But for the doublespender there's a huge payoff for getting 51% of the capacity regardless of the fees. It's not just $1M one time, but it's $1M on every transaction... forever.
I hope that clarifies the argument.
You can now see that logic dictates that the cost of achieving 51% needs be higher than the payoff (which will be as many of these $1M transactions as you can get away with). There is nothing in the bitcoin system that adjusts the cost of POW to the magnitude of the transaction. It does adjust to the fees which might indirectly reflect the transaction value but this is not a direct coupling as it will be clamped by competition from other fee based systems (like for example, writing a cashier's check or using a visa card).
If you take my made up numbers here at face value (not correct but makes the point) then if there's a 10 to 1 more profit in double spending, any irrational actors competing for that "1" will still not matter.
As I noted originally, I'd love it if you could prove me wrong, but I don't think the arguments you gave do this.
How might this come about?
well I see four scenarios as nearly certain to come along at some point There are probably many others I haven't considered.
1. Due to distorted economics like subsidized electrical power or cheap cooling, some countries accumulate a dominant share of the mining. Then for some reason the government of that country kills the market (e.g. remove power subsidies or siezes the machines) for mining. Suddenly there's void of miners that someone with a big cluster (amazon?) could step into.
2. due to geopolitical tensions (a war?) one country with a small number is cut off from the rest of the world's miners. Again now some small player witha lot of CPUs can dominate the ledger for that country. We see internet cut offs in Iran and Syria and other places now and then. And there's growing numbers of firewalls around countries. IMagine if China and the US came to blows in the south China sea or North korea executed a cyber blitz kreig on south koreas. Isolations of the internet would occur.
3. The reward system for bitcoin changes with time with fees replacing mining. What if the fees don't attract enough miners? that is say, as fees rise the number of transactions will fall. there is some equliubrium there. That equilibrium sets the POW difficulty and the number of miners who can make a profit. Do we know that equilibrium is sufficient to protect the transaction value of a double spend? I don' think there's anything built into the model to assure this.
4. What if some new crytpocurrency suddenly becomes more profitable for miners and they point all their ASIC's and GPUs at that leaving bitcoin a ghost town with just a few miners. Here's one such scenario. Someone, maybe etherium can, will figure out how to make the wasted effort of the POW actually valuable in itself. e.g. make the POW be the travelling salesman problem of delivery planning that UNited Parcel Service has to solve every day. Now you get paid for the POW by the currency holders and also get paid for the work you did for UPS. this will beat out coins where the POW is worthless as a calculation.
There are a lot of complaints one can lob at cryptocurrencies but I want to address just one that at the moment I see as ultimately fatal to all proof of work systems. Since this fatality hasn't actually materialized yet I have to wonder if I'm wrong about it being unavoidable in the end. But I don't yet see how I am wrong so here goes:
The ENTIRE magic and near Genius solution that bitcoin and others perform is the avoid the "double spend" problem when there is no central authority to manage a secret signing key. Solving the double spend with distributed signing is the magical part.
The DOuble spend problem is that in an a normal distributed ledger that anyone can write to a bad actor could spend a coin, see it entered in the ledger as beloging to the seller, then after getting the benefit of the sale from the seller, re-write a newledger in which that spend never happened. The bad actor can then re-spend the same coin.
Block chain by itslef doesn't solve this. it's just a ledger format. But when you add the proof of work part then you have an escalating difficulty barrier. The seller waits to see the transaction is confirmed. If they are paranoid, they could even wait for several more epochs of chain extension. At that point if the bad actor wanted to re-write history they would have to create a new block chain that was longer than the currently accepted one. And that would be hard because of the multiple epochs of proof of work.
this is exactly why the fear in bitcoin is that if one person accumulated enough mining power they could execute a double spend. But since this is addative: it takes 3x more mining power to unwind 3 layers of the block history (and N-1 x more mining power than the world, to unwind N layers), it's hard.
Or rather it's hard, but only if the world has a lot of miners. If miners lose interest two things happen. First it becomes easier for the bad actor to accumulate enough CPU power to overwhelm the rest of the world's miners. Additionally, since Bitcoin in particular scales the difficulty to the transaction rate it also requires less and less CPU power to do this. (as miners lose interest and so the POW difficulty goes down)
Thus Cryptocurrencies only protect all that capitalized outstanding wealth only as long as their's an active pool of miners. If that goes away then the protection of the blockchain is gone and the double spend re-emerges. at that point it gets crazy.
SO why might this not have happened yet. I think maybe it's because the cost of mining a coin is so high that the cost of mining 2,3, or 4 coins to unwind 1,2, or 3 layers might not have been worth the gained value of the doublespend. But that' now. As the coin becomes increasingly capitalized then a lot of wealth will be transacted in each epoch (and the lightning network is amplifying this now). Thus the temptation for a fouble spend will eventually exceed the cost.
At that point there is a total heat death of the currency as no one can trust it.
Once the miners stop buying as you put it, things will normalize to zero