Cryptocurrency

Worldly and otherworldly topics
User avatar
Mashy
Posts: 56
Joined: Mon Mar 01, 2021 8:33 am
Contact:

Re: Cryptocurrency

Post by Mashy » Tue Mar 02, 2021 8:09 am

Thank you for taking the time to write that up @Utisz ! I normally skip text walls but you ELI5AHAWFWACASAYAFSR (Explained Like I'm 5 And Had A Weird Fascination With Alternative Currencies At Such A Young Age For Some Reason)

I still don't full understand but hope to learn more every day. I have some questions, I hope they make sense.

Can you talk more about what happens when bitcoin reaches the 21M max supply? Miners won't be paid BTC to mine anymore? I see some altcoins have already reached their max supply? There is a crypto website I frequent that has a warning "Warning! This coin has no max supply!" about certain altcoins, why should that raise concern? I hear dogecoin has changing max supplies which makes it inflationary? How is that possible?

I read about 51% attacks and from what I understand it won't realistically happen to BTC - not impossible, just super super improbable. I've heard it happening to smaller coins with fewer nodes on the network.


Executing smart contracts on the ethereum network is probably the coolest idea I have ever encountered in my lifetime so I am pretty excited to see how it'll play out in the coming years. But. I minted my first Non-Fungible Token yesterday (artwork on Rarible) and the ether gas fees were super expensive, like $40 AUD. I did a double take when I saw my eth balance in the morning. I heard they're coming out with Eth 2.0 to address scalability issues but I have no idea how it works and I read contradictory reddit posts on it all the time.

It seems smart contracts are still years away from optimised mainstream adoption. Although technology does grow exponentially, so maybe by next year developers will have improved the user experience so it's not so frustrating.

User avatar
Utisz
Posts: 676
Joined: Fri Nov 22, 2019 4:35 am

Re: Cryptocurrency

Post by Utisz » Tue Mar 02, 2021 11:37 pm

Mashy wrote:
Tue Mar 02, 2021 8:09 am
Thank you for taking the time to write that up @Utisz ! I normally skip text walls but you ELI5AHAWFWACASAYAFSR (Explained Like I'm 5 And Had A Weird Fascination With Alternative Currencies At Such A Young Age For Some Reason)
That wall of text was also me trying not to write a wall of text. :cool:
I still don't full understand but hope to learn more every day. I have some questions, I hope they make sense.

Can you talk more about what happens when bitcoin reaches the 21M max supply? Miners won't be paid BTC to mine anymore? I see some altcoins have already reached their max supply? There is a crypto website I frequent that has a warning "Warning! This coin has no max supply!" about certain altcoins, why should that raise concern? I hear dogecoin has changing max supplies which makes it inflationary? How is that possible?
Correct yep! No more bitcoins for miners at that point. The payment will halve every 210,000 blocks (i.e., roughly every 2,100,000 minutes, or every 1,458 days) until it drops below 1 A Satoshi (0.00000001 BTC), at which point miners will not receive any more bitcoins, and no more bitcoins will be generated (the total maxes out at 21 million). This will happen in around 2140. Payments will become miniscule long before that of course. I found this graph here showing how the supply decreases with time:

Image

In terms of what happens then, something I did not talk about before are transaction fees, which are used to prioritise which transactions to process.

There's a fixed amount of space in each block to store transaction fees, where currently about 2,000 transactions can be recorded per block. And a transaction is not official until it's recorded in a block for a while (with other blocks added after it, which makes it more secure in that competing chains of blocks will not arise elsewhere). And there's a fixed rate at which blocks are mined on average, namely every 10 minutes. So bitcoin can currently only process about 2,000 transactions every ten minutes. Often there will be more demand for processing transactions than can be processed. So to get in the next block, practically speaking, you need to bid within the highest 2,000 transactions. This is another limitation of bitcoin specifically: that it does not currently scale well at all in terms of the number of transactions.

Those who are still mining in 2140 will then only receive transaction fees, which will be how ever much people are willing to bid to get into the next block. Google tells me the current fee for one transaction is about $20, and that there are about 2,000 transactions per block, so aside from the 6.25 bitcoins they currently get, miners make around another $40,000 on transaction fees for each block.

Of course the value of the transaction fees will dominate the bitcoin payout long before 2140, even if bitcoin values reach really absurd levels.
I see some altcoins have already reached their max supply? There is a crypto website I frequent that has a warning "Warning! This coin has no max supply!" about certain altcoins, why should that raise concern? I hear dogecoin has changing max supplies which makes it inflationary? How is that possible?
Scarcity (i.e., limited supply) is an important resource for any money. If we found a cheap way to turn sand into gold, the price of gold would fall off a cliff, for example: gold would no longer be a good basis for money. Even if people thought that it was likely that a cheap way to turn sand into gold would be found, there would be zero confidence in money based on gold. It's a similar issue for cryptocurrency: if folks think there might be some trick where someone could cheaply and quickly create a massive amount of that cryptocurrency, then the cryptocurrency is doomed to fail.

Scarcity is also a concern for fiat currency (backed by a government, but not by a commodity of inherent value).

Take this note for 100 quintillion (thousand thousand thousand billion) pengő, the Hungarian currency of 1946:

Image

Two years prior, the highest note needed was 1000 pengő. Prices at one point were doubling every 15 hours due to hyperinflation.

Part of the problem was that the Hungarian government started printing money to try to offset economic problems. And there was no theoretical cap on how much they might continue to print, so basically there was no cap on the scarcity of the money. The pengő was seen as more and more worthless, which spirals out of control as people lose faith in the currency, get their money out of the banks, buy physical things with it raising prices even more as things sell out, making it more worthless, leading to hyperinflation in order to have some sort of money you could try to buy a banana with.

So if there's no theoretical cap on an amount of money, there's going to be this question of scarcity. I don't think that there necessarily needs to be a max supply like Bitcoin, but I guess there should at least be a max supply over a period of time. Like if BTC continued to pay 1 BTC per block forever when it reached that level, I think it would not be a fundamental issue because everyone would still know how much supply there would be at a given point in time, and that there would never be enough to ever devalue what they hold. If, on the other hand, there's no max supply over a fixed period of time, some serious questions have to be asked regarding how scarcity is guaranteed.
I read about 51% attacks and from what I understand it won't realistically happen to BTC - not impossible, just super super improbable. I've heard it happening to smaller coins with fewer nodes on the network.
It would take China (the government) to do it now. They have 60% of the pool. If Xi Jinping clicks his fingers, it could get done for bitcoin. But it would likely take someone like Xi Jinping to mount such an attack at this stage, yes. How likely that is, however, I don't know. It would wreck the currency, and it's hard to say what they would get out of it.
Executing smart contracts on the ethereum network is probably the coolest idea I have ever encountered in my lifetime so I am pretty excited to see how it'll play out in the coming years. But. I minted my first Non-Fungible Token yesterday (artwork on Rarible) and the ether gas fees were super expensive, like $40 AUD. I did a double take when I saw my eth balance in the morning. I heard they're coming out with Eth 2.0 to address scalability issues but I have no idea how it works and I read contradictory reddit posts on it all the time.

It seems smart contracts are still years away from optimised mainstream adoption. Although technology does grow exponentially, so maybe by next year developers will have improved the user experience so it's not so frustrating.
Smart contracts are a pretty interesting idea, yep. They remind me of the title of an Adam Curtis documentary: "All Watched over by Machine of Loving Grace". They are complicated and expensive though.

The first issue is that for smart contracts to become truly anarchistic in terms of no central control or middle-men, the people involved will have to understand, validate and maybe even write the code of the smart contract that they are agreeing to for themselves, which is not easy, even for programmers, and especially for complex contracts. Otherwise someone will have to be paid for that service (a type of nerd lawyer I guess), and can you trust them? The second issue is that Bitcoin and Ethereum allow Turing complete languages, basically any computer programme, which is a bit worrying as some programmes may not work on some inputs (they may run for ever and never output any result), and there is no way to know, in the general case, a priori, which programmes have this problem, on what inputs they will fail, or what should happen in this result (maybe the contract is declared null and void and everyone gets their original investment back somehow). The third issue is indeed the cost.

I hadn't read anything about Ethereum 2.0, but I've heard of proof-of-stake (basically proof of vested interest in the cryptocurrency continuing to function correctly). It would reduce the wasted computation compared to bitcoin mining, for example. But it seems less robust, maybe. For example, to me, proof-of-stake in a cryptocurrency does not imply a lack-of-an-even-bigger-stake in that cryptocurrency failing or being derailed in some way.

Sharding is also a known idea (splitting the verification into parts and solving each part on one machine, rather than having all machines verify everything). But it would by definition be "less secure" than something like Bitcoin.

That said, Bitcoin has proven "secure enough", so it might be time to explore solutions that are less secure (and still "secure enough") but more scalable.

User avatar
Mashy
Posts: 56
Joined: Mon Mar 01, 2021 8:33 am
Contact:

Re: Cryptocurrency

Post by Mashy » Wed Mar 03, 2021 5:38 am

Thanks for another interesting post :thumbsup: Here are some more questions that a cursory google search couldn't answer for me:
Utisz wrote:
Tue Mar 02, 2021 11:37 pm
I hadn't read anything about Ethereum 2.0, but I've heard of proof-of-stake (basically proof of vested interest in the cryptocurrency continuing to function correctly). It would reduce the wasted computation compared to bitcoin mining, for example. But it seems less robust, maybe. For example, to me, proof-of-stake in a cryptocurrency does not imply a lack-of-an-even-bigger-stake in that cryptocurrency failing or being derailed in some way.

Sharding is also a known idea (splitting the verification into parts and solving each part on one machine, rather than having all machines verify everything). But it would by definition be "less secure" than something like Bitcoin.

That said, Bitcoin has proven "secure enough", so it might be time to explore solutions that are less secure (and still "secure enough") but more scalable.
Re: proof of stake. What's the computational difference between a new block being "mined" and a new block being "validated"? Is PoS more green because the process of validation itself uses less energy or because not every node on the network has to participate, or both?
From what I understand in PoS you can only validate if you contractually secure a certain stake (e.g. 32 eth), and then a node is randomly selected to "validate" a new block? What stops it from falsifying information? Sounds like the opposite of distributed ledger technology?

User avatar
Utisz
Posts: 676
Joined: Fri Nov 22, 2019 4:35 am

Re: Cryptocurrency

Post by Utisz » Wed Mar 03, 2021 6:33 am

Mashy wrote:
Wed Mar 03, 2021 5:38 am
Thanks for another interesting post :thumbsup: Here are some more questions that a cursory google search couldn't answer for me:

Re: proof of stake. What's the computational difference between a new block being "mined" and a new block being "validated"? Is PoS more green because the process of validation itself uses less energy or because not every node on the network has to participate, or both?
So proof-of-stake effectively replaces what they call proof-of-work, as used by Bitcoin and other cryptocurrencies.

Proof-of-work means that some node on the network solved some task that is known to be difficult, which in this case is finding a key in over a quadrillion possible keys that fits the block encoding all the transactions to be recorded. It is hard to find the key because nobody has any clue which key works, you just have to try all of them, and there are so many. However, if someone gives you what they claim to be the key, it is fast to verify; you can see if the key unlocks the block in fractions of a second. So it is hard for someone to mine a block, but once it is mined, it is easy for everyone to verify that the block is valid.

(More technically, the key is a number. There is a fixed mathematical function, called a hash, defined as part of Bitcoin, that can take a block and produce a number. There is no known way to predict what number the hash will give for a block. The only way is to compute the hash and see what it gives. A key "unlocks" the block if, when inserted into the block, the hash over the modified block returns a value below a target number. To make mining easier (which happens when the last blocks took longer than ten minutes, trying to speed up mining) the target is raised, meaning more keys/hashes are accepted; to make mining more difficult, the target is lowered. As there is no known way to predict which key/number will unlock the block, I understand that miners start with 1, 2, 3, ..., and keep going, typically up to several quintillion, until they find one.)

Let's say in bitcoin it was easy to mine blocks. This would make it so that lots of people mine blocks at the same time. And others would be scrambling to mine the next block in the sequence. And so it would lead to competing branches of the blockchain, i.e., utter confusion, more blocks than can be verified, attempts to hack the process by mining long chains all at once with false transactions, etc.

The blocks that form the blockchain of Bitcoin are then reliable because they are slow and hard to compute. This makes it hard for people to "fake" blocks, let alone chains of blocks, because of the work involved. And without these fake blocks, we avoid fake transactions. The difficult of mining a block means that, despite the electricity used by Argentina being invested into mining blocks, one is mined only every ten minutes. This gives ample time for everyone to verify the block before the next one is mined and added. The only way (it is thought) that you can cheat in this system is to control the majority of computing power, the so-called 51% attack.

This system is called proof-of-work. It develops trust in that it provably requires massive resources (50% of mining capacity) to even try to commit fraud. If you try to commit fraud with say 20% of the computing power, the other 80% of the network can simply ignore your block, mine a valid one to replace it, and produce a longer (valid) chain without the fraud in the history. (The longest chain is accepted.) While proof-of-work is, ahem, proven to work, it is super wasteful. All these miners are computing numbers that mean absolutely nothing, it's just some arbitrary difficult task needed to satisfy proof-of-work and make it so that massive resources are needed to try to commit fraud.

Another idea is then proof-of-stake. The idea is to trust those that have the most stake in the cryptocurrency to formalise the transactions ... to do the mining. The mining in this case doesn't need to be artificially difficult. If the miner tries to commit fraud, they are undermining the currency, and risking their stake in it. Also, the rest of the network can override their actions so long as they have more stake combined. As I understand it, if you have say 20% of the stake (e.g., in a simplified model, hold 20% of the coins), then you will be asked to mine and verify 20% of the blocks. A 51% attack is possible, but only if someone now controls 51+% of the stake/coins. Otherwise, if you try to commit fraud with less stake, again the other stakeholders will be able to mine more blocks than you can and create a longer chain without your fraudulent transactions. (There are more advanced proof-of-stake models that also take into account things like how long you've held the coins to avoid newcomers gaining massive stakes to try to overthrow the currency, but I guess these are details.)

In both models, all nodes in the network should still be validating and keeping an eye on what's going on, which is a separate issue. What changes is just how blocks are mined, and who gets to mine them. In proof-of-work, it's more or less random who finds the right key, but over time, you should mine a percentage equivalent to how much of the computing power (mining capacity) you control. In proof-of-stake, you should mine a percentage equivalent to your stake in the cryptocurrency.
Mashy wrote:
Wed Mar 03, 2021 5:38 am
From what I understand in PoS you can only validate if you contractually secure a certain stake (e.g. 32 eth), and then a node is randomly selected to "validate" a new block? What stops it from falsifying information? Sounds like the opposite of distributed ledger technology?
Yep, it's still a distributed ledger. It's indeed random, but not evenly random, rather random weighted based on stake. So if you have X% of the stake, you have an X% random chance to mine the next block. Unless you have 50+% of the stake, if you try something funny, others will eventually be able to mine more blocks and overwrite your funny business in the distributed ledger (creating a longer valid ledger).

User avatar
Ferrus
Posts: 270
Joined: Sun Feb 28, 2021 11:10 pm
Location: Barcelona

Re: Cryptocurrency

Post by Ferrus » Thu Mar 04, 2021 12:08 pm

Utisz wrote:
Mon Mar 01, 2021 10:19 pm
Yep, the regulatory issues are still grey, and could pose an existential threat if policy makers do not have a vested interest in seeing bitcoin or cryptocurrency flourish. This is also an issue if they have a vested interest against bitcoin or cryptocurrency, an obvious one being that the government can actually control the dollar, so there's a built-in bias for government/regulators in favour of a fiat currency that they can control by definition over a cryptocurrency that they (by default) cannot control at all. If/when cryptocurrency begins to threaten fiat currencies as a real alternative, and thus government start to lose control over the currency used and over their own monetary policy, that will be an interesting time to witness.
Although also, if governments decide to absorb it, they can.

When the gold standard existed most governments didn't actually have enough gold to cover all the bank notes in existence. It is true that it prevented creating too many bank notes as it would make it vulnerable or brittle to a run if it was known. Even so, some governments like France in the 19th century would "sterilize" their gold flows, see: https://eh.net/encyclopedia/gold-standard/ by using interest rates to manipulate the amount of monetary base in terms of assets and deposits. Very similar to what China has done the last 20 years in neutralising the inflationary impact of foreign currencies by deliberately reinvesting them in US treasuries by state fiat and thus keeping the price of the dollar up*. I see a lot of people have a rather simplistic conception of the money supply, it is true that overprinting money will eventually lead to hyperinflation, but in a normal healthy economy the amount of money is far more determined by the type and intensity of banking operations and lending than the exact quantity that a government prints. This is why QE didn't cause inflation, as because of various regulatory and economic factors banks dramatically cut back lending and thus the money supply, which is a product of fractional banking. The gold standard doesn't stop fractional reserve banking, which is another misconception, as you can lend on top of loans based on a gold back currency as long as people are ignorant of the actual amount of gold, or think it will cover the velocity of transactions up to that point, and without a control of fractional reserve banking the gold standard is just as vulnerable to sharp spikes of deflation in economic crises where suddenly banks increase their reserve requirements from fear of risk, as was seen in the Great Depression, which in itself reduces the willingness of people to invest because the money they have now will be worth less in the future. The main benefit of the Gold Standard was that it regularised exchange rates between currencies - all were based on gold, so there was arbitrage to be had if they were out of sync with the value of the currency in the international market, factoring in transportation costs, which were known as the gold points - based on a price-flow species mechanism described a long time ago by David Hume (but made a bit more complicated by sterilization by certain countries, see before). But this is also very difficult to sustain politically in the modern world. In the 19th century when this was the dominant form, there were no unemployment rates, and it was not considered the responsibility of the government to supply an environment for employment, merely that property would be protected and perhaps grudging doses of charity. People seem to think using the Gold Standard (or Bitcoin) would be some kind of victory against globalisation and the destruction of national governments - as Forex trading has been used to decimate economies in the past - but it would actually hand even more power to financial interests that they had in the 19th century, governments would no longer be able to devaluate their currency to improve competitiveness. Where or not this is a good idea or not - I am generally inclined to think in the long term it is not, but the recent Covid crisis has also shown the advantages of keeping a certain level of domestic industry shielded from international speculation - it seems very difficult policy to enact in an age increasingly dominated by populist movements - even though many of the populist movements are in favour of a Gold Standard. There is an irony there that this Gold Standard they want so much would be an even harsher version of the Euro, a transnational currency controlled by a central bank outside of the home country is another way to take that power away, without even the control of a supranational government, but entirely dependent on the economic competitiveness of the country to avoid financial interests staging an attack on gold reserves.

Something similar could develop with Bitcoins, if they were to become a universal currency, then governments would have to keep reserves of these currencies to prevent speculation against their own national currencies that would give bitcoin investors the capacity to punish economically underperforming countries, or those with excessive fractional reserve bond debt (back by bitcoin). You could get Bitcoin derivatives that base their value on a pool of Bitcoins and are then supplied according to a monetary multiplier factor. This will happen naturally with time and maturity of any market as there is a financial incentive for those who have hoarded bitcoins to do something like this, if they can generate enough belief in bitcoin derivative currencies. The only way to stop this is a government level ban on deposits less than 100% backed by currency.

* I remember people getting agitated about the amount of debt China owes, but actually it was as bad for them as a collapse in the dollar would destroy their economy and import-mediated growth, which is exactly why Xi has been trying to cut down on reliance on foreign exports, reinvestment in T bills towards a more national economy including heavy investment in homegrown tech, the success of which we will see with time.
Ex falso, quodlibet

User avatar
Mashy
Posts: 56
Joined: Mon Mar 01, 2021 8:33 am
Contact:

Re: Cryptocurrency

Post by Mashy » Mon Mar 08, 2021 8:38 am

Then if I understand correctly - in a Proof-of-Stake model, is a 51% attack more likely because the pool of validators is smaller?

And hypothetically if the validator was rich as fux s/he could easily take control of 51%+ of the stake, and override every other node in the network?

I just staked some Polygon (MATIC) on Binance, just to try it out. I am guessing I am contributing to Binance's stake and they are then paying some interest to all of us Binancians who contribute to the stake. It really is a matter of rich getting richer.

Also I found a really good MIT video lecture series on Blockchain if anyone is interested. Very clear and concise!

User avatar
Utisz
Posts: 676
Joined: Fri Nov 22, 2019 4:35 am

Re: Cryptocurrency

Post by Utisz » Mon Mar 08, 2021 4:43 pm

Mashy wrote:
Mon Mar 08, 2021 8:38 am
Then if I understand correctly - in a Proof-of-Stake model, is a 51% attack more likely because the pool of validators is smaller?
Good question! :?:

I was kind of assuming that a 51% attack is more likely under proof-of-stake than under proof-of-work, but on second thought, I might have been comparing apples (an established cryptocurrency based on proof-of-work like BItcoin) and oranges (some new cryptocurrency based on proof-of-stake). Thinking about it more, I'm not sure that 51% attacks are "easier" under proof-of-stake than proof-of-work.

It doesn't necessarily relate to the pool/number of validators, but rather to the resources required to gain control of 51% stake/work, and what you can stand to lose/benefit from gaining control of 51%. (edit: But yep, with fewer people participating with fewer resources, it gets easier to get to 51%)

Under proof-of-stake, if you can buy out half the cryptocurrency then you can effectively gain control of the currency, depending precisely on the protocols in place (e.g., maybe there's a black list for naughty participants, but there's then the issue of who defines the black list, and under what circumstances).

Under proof-of-work, you can buy/rent the computation power to do a 51% attack based on proof-of-work. The cost then relates to the amount of time you run the 51% attack for (which depends on what your plan is, I guess).

Assuming the plan is to fuck over the cryptocurrency with a 51% attack (which would probably be the end result in any case as folks lose confidence), we can assume the cryptocurrency drops in value after the attack. So then the question is what you lose. Under proof-of-work, you lose all the costs of the computation power needed for the attack, plus the value of whatever cryptocurrency you were awarded during that computation. Under proof-of-stake, your stake loses value.

I would imagine in most cases that the loss under proof-of-stake would be more costly, and it would be harder to acquire 51% stake than 51% computation power. But it gets more complicated in cases where you can borrow or short cryptocurrency. So not sure. :huh:

I would say that 51% attacks seem unlikely in both cases, particularly after the cryptocurrency has matured for a bit.
And hypothetically if the validator was rich as fux s/he could easily take control of 51%+ of the stake, and override every other node in the network?
Yep! That's also true under proof-of-work though. They could buy/hire all the machines they need to get to 51% of the mining resources.

And if they were to sink a cryptocurrency, under proof-of-stake, they would lose that stake.

User avatar
jyng1
Posts: 260
Joined: Mon Mar 01, 2021 1:13 am

Re: Cryptocurrency

Post by jyng1 » Fri Mar 12, 2021 6:59 am

Anyone wanna buy a tulip $69.3 million jpeg? NFT's are leaving crypto for dead...

Image

User avatar
Mashy
Posts: 56
Joined: Mon Mar 01, 2021 8:33 am
Contact:

Re: Cryptocurrency

Post by Mashy » Mon Mar 15, 2021 6:52 pm

I remembered I have a story similar to the bitcoin pizza story,

I first bought bitcoin with my brother in 2013, I bought 1.7 BTC when it was only ~$300AUD. (He later traded that for ethereum without my knowledge and it was a very good move). A few years later my brother asked me if I wanted to sell some of my ether directly to him. I needed a sweater at the time so I thought sure easy way to withdraw into fiat without paying capital gains tax. It was like a $2000 sweater if we go by today's Eth price :heh: and urbanoutfitters sent me the wrong sweater too..


jyng1 wrote:
Fri Mar 12, 2021 6:59 am
NFT's are leaving crypto for dead...

Wait why? You buy NFTs with crypto

User avatar
jyng1
Posts: 260
Joined: Mon Mar 01, 2021 1:13 am

Re: Cryptocurrency

Post by jyng1 » Mon Mar 15, 2021 8:13 pm

Mashy wrote:
Mon Mar 15, 2021 6:52 pm
jyng1 wrote:
Fri Mar 12, 2021 6:59 am
NFT's are leaving crypto for dead...

Wait why? You buy NFTs with crypto
At least you can buy something with it ;)

Post Reply