f7ebc20c97

Hacker News commenters have hated (and been wrong about) cryptocurrency ever since the first Bitcoin article appeared here. It's very strange. I chalk it up to some sort of Dunning-Kruger contrarian effect + demographics skewed toward cynical authoritarian statists (not sure why, since tech used to be full of libertarians).

It's all just a repeat of the negative comments over and over... https://news.ycombinator.com/item?id=599852

I wonder when efficiency gains will be slow enough so that a PoW heater will pay for itself before it's obsolete.

Isn't 99% optimal good enough?

Please, let's not get bogged down in a discussion about heat pumps.

They're both ~100% efficient at converting electricity into heat.

Not necessarily. The mining algorithm could hash the block once using SHA256, then use that and a fixed-size nonce as a fixed-size parameter for the ASIC-optimized hash function.

Proof-of-work cryptocurrencies moved toward complex hash functions that are "ASIC resistant" in order to attempt to keep mining power as distributed as possible (for CPUs), because whatever highly-funded manufacturer designs the most-optimized and fastest ASICs gains concentrated mining power. ASIC hash/W efficiency doubles every couple of years, so using them as heaters becomes cost-ineffective as difficulty adjusts.

But what if we turned this all on its head and intentionally designed a hash function that is "ASIC optimized" so that the perfect ASIC miner design could be found quickly (perhaps at its very conception), and miners became a generic commodity item, "grey goo miners", and used them as space heaters in cold climates, or maybe even wherever electric heating was needed in industry? That way mining would be as distributed as the need for electric heating, and its energy would be put to good use. It might also put a ceiling on the total electricity consumed because it would be hard to compete with any miner whose electric bills are "free" (relatively).

Blocking copy-paste, blocking screenshots, and blocking file downloads...for only $1.3 billion dollars!

Are you sure we're not in another tech bubble?

You think TPTB want to handicap the ability to launder money?

You're not the first guy who's been motivated to search for a cure for his own disease. Go straight to the source. Contact the leading researchers and ask about doing a PhD with them or something (I'm not sure how academia works, I never could figure it out). Surely they'd be happy to have motivated researchers. You'll also be on the cutting edge of advanced experimental treatments and you can experiment on yourself someday (if you want).

Ah, I see. Thanks for the info. Government intervention (trading something for nothing) is behind what I thought was a free market, as usual.

Free banking has a downside (bank runs) so the government inevitably narcotized it, like they do everything. Established political power will do anything for stability.

The "Great Moderation" is the triumph of forced central banking. They've tamed the business cycle. But you can't get something for nothing. Fixed inflation rates have stabilized the economy by protecting decaying firms and suppressing startups. It's even done the same for demographics. From stability came structural stasis, came a decline. With fixed inflation, the equilibrium interest rate is now negative. CBDCs and "degrowth" are coming soon to enable that and protect the establishment at all costs.

The Austrians are always right. We need free banking, we need a business cycle, but it's bitter medicine.

What's difference between IPFS and Oxen??

Does a hypereducated liberal democracy even work in the long run, or will a morally relativist people aloof from reality argue over lifeboats and bilge pumps forever until the ship sinks and they all drown?

Sorry for being cynical.

On the trajectory we're on, we're going to need people like you (OP) to figure out what we'll build on the ashes of our current civilization. I see you like diversity libertarianism / Archipelago. A more robustly decentralized version of the current US constitution, which was actually really good until the corporate coup.

You're in luck! We're in the midst of a transition to a brand-new architecture. You're going to love social credit.

Expert social engineers (sociologists, psychologists, economists, environmental scientists, et. al.) will be chosen to define "well being" using the latest scientific research and data. Money won't exist. You'll have a social credit score from 0 to 1000 depending on your overall value to society, the economy, and the environment. Higher scores will qualify you for better food, housing, and transportation. Lower scores will subject you to various restrictions or even imprisonment. Purchases will deduct different amounts from your score depending on their true overall societal cost; prices will include all economic externalities as determined by the experts.

Since the score asymptotically caps out at 1000, all economic inequality problems will be solved. In addition, there won't be the bureaucratic corporate management overhead of the current institutional patchwork because an expert-tuned AI management system will give you instructions on what work you need to do to help your score (including diet and exercise), using your mobile device or neuralink.

Luckily, there won't even have to be a violent revolution to reach this utopia, because just about every world leader at the UN, WEF, IMF, BIS, major political parties, large corporations, and so on are ALL on board with this.

Guns are also "easy" to manufacture, but are successfully restricted.

The clear outcome of this is the current ptb enslaving the world without needing so much as a police force.

Wow, is it even possible to form a science fiction setting without clichés?

Everything will be banned once someone does it the bad way, because there aren't enough resources to determine each case on an individual basis.

Nothing is free. The energy input of this system is the time-value of the coins in payments waiting to be buried by succeeding blocks. It's kind of like proof-of-stake, but the stakers are the transactors themselves.

Basically, how big of a payment you can do depends on how much volume there is. It's a positive feedback loop, though.

There will be a distribution of transaction values with a median less than the mean, as seen in credit cards, swift, and all other payment systems. There's a long tail of large value transactions that would have to wait longer. So if you buy a car you'd have to wait for the equivalent of 1000 more car-valued transactions to occur, but this is helped by the ton of smaller-valued transactions backing you up.

If the maximum someone is willing to wait for a transaction to be confirmed is 1 hour, then the ceiling of acceptable transaction value is 0.1% of the sum of total value transacted per hour, but this gradually rises as the system gets more popular. If you think about it, a similar phenomenon happens in any payment scheme, more or less. Even Bitcoin would make you wait a long time for a trillion dollar transaction because the cost of mining a double-spend would take a long time to reach a trillion dollars. In the early days you wouldn't trust it with even a $10,000 transaction because anyone could farm out a huge mining operation.

I had some funding, worked on it for a year solid, built working prototypes, learned a lot, prepared for a whitepaper, but then I realized that an attacker could still profitably double-spend if they made a ton of small payments at the same block to different people. This would show up as a suspiciously-large transaction volume though...

I think it means that (worst case) to accept an avg payment you'd have to wait for 1000 avg blocks instead of 1000 avg transactions...? If the fee was 1%, instead, though, you'd only have to wait 100 blocks, so it becomes reasonable? I'm kind of stuck and have nobody to talk to about it...

Edit: You've inspired me to again work on this! Thanks!

Let's consider this "worst-case" scenario with 100% of txs are made by the same attacker with an intent to double-spend. Before generalizing, let's fix the amounts to 1000 txs/block with a payment price of $100/tx with a $1 fee/tx.

Block 1: The attacker sends 1000 txs (the "payments") worth $100 each to 1000 different merchants. He pays $1000 in fees. He stands to have revenues of 1000 x $100 = $100,000 if he can double-spend the payments. His payments are currently buried by 1000 x $1 = $1000 in their own fee costs, for a net profit of $99,000 if the merchants accept his payments right now. The merchants are cynical and rationally choose to wait until this block is buried by 1000 x 100 x $1 fee/tx = $100,000 in fees before accepting the payments. (It doesn't even matter if they know he's trying to double-spend or not).

Block 2: The attacker sends 1000 loopback txs to himself to bury his payments in another 1000 x $1 = $1000 of fees. $98,000 to go...

Block 3, 4, 5, ... 100: The attacker sends 1000 loopback txs to himself per block for 98 blocks to bury his payments in another 1000 x $1 x 98 = $98,000 of fees for a total of $100,000 in costs. The merchants are now satisfied with the payments and accept them as honest transactions. The merchants send the attacker $100,000 in goods.

At this point, the attacker replaces block 1 with another block that is just full of loopback txs to himself rather than payments to the merchant. He then replaces blocks 2-100 with a new set filled with loopbacks. Then he makes block 101 with one loopback tx to himself so as to outbury the old blockchain with an extra $1.

The attacker's current revenues are $100,000 in goods, but he had to pay $100,001 in fees to do it.

We can generalize this to mean that to trust a transaction, worst-case, you have to wait until the the total payment value of the block itself is buried by its own value in transaction fees. Therefore, you'd have to enforce minimum % fees because, if you allow low transaction fees then an attacker could send a huge loopback transaction that makes everyone else in the block to have to wait a long, long time until enough fees bury that block. Maybe you could solve this using fixed block sizes that force transactions to have to bid for a spot. I guess you have to do that anyway, though...

It gets better as more transactions are added to the end of the chain (in any later blocks). If fees are fixed at 0.1%, the average transaction only has to be buried by 1000 more average transactions. Most transactions are smaller than average, so they don't have to wait that long.

This also means that transactions are buried faster the more heavily the blockchain is used. For example, you'd only have to wait one block if there were 1000 transactions in it.

The only problem is if the double-spender is trying to double-spend 100 transactions at the same time...

Wouldn't they only need to store the outputs that they themselves signed, in order to prevent themselves from signing the same one twice?

Yeah. The developers hardcode (eg) 100 validators into the client. Transactions need to be signed by 67 of them. Why not require only 51?

Maybe there could be a permissionless way to get all clients to deterministically agree on who the validators are...maybe using something called proof of work...

I've considered a completely different system called proof-of-fee that has no miners or stakers:

- Each transaction pays a fee (that is either redistributed or burned)

- The best blockchain is the one with the most fees paid

- Each transaction signs the previous block

- The blockchain length is limited to 1 block per 15 seconds

- Nodes put their own transactions into the current block or make a new block with it (depending on if the blockchain length is maxed out) and broadcast it. No miners or stakers, everyone's equal.

Therefore, in order to replace a buried transaction (to double-spend), you'd need to pay more fees than all the transactions in blocks that were burying your old one. This would be unprofitable for deeply buried transactions.

There would be a lot of reorganization going on around the blockchange time at the 15 second mark, but that's an engineering problem.

It's a permissioned system. A bank, basically.