[tzumby]

What is surprising is how large the number of techies who don’t get this. It’s not about crypto currencies as much as it is about the game theory that allows it to run a distributed self replicating state machine that is Byzantine tolerant - this is the true innovation

[shieVae4I]

It's not a particularly interesting Byzantine mechanism. Like every other such mechanism (in the absence of trusted computing and similar tricks) it has a n = 3f+1 limit. See the selfish mining paper for why it's not n=2f+1.

It's just the Byzantine mechanism that happened to win. If anything, its innovation is that its n is based on hash power, not on number of users or nodes.

And the cost? Having to drag an immutable database around that can't be simplified; and, obviously, the opportunity cost of all the energy that's being used to support the consensus.

The game theory part is likely to be inexorably intertwined with the cryptocurrency part. The incentive to defect is blunted by people's investments, either in terms of hash power machines (for PoW) or in coin holdings (PoS). It would be very difficult to construct something that would be self-replicating/hype-incentivizing but not reward early comers the same way.

[xpe]

> It would be very difficult to construct something that would be self-replicating/hype-incentivizing but not reward early comers the same way.

Can you clarify what you mean by "the same way"? I wonder if you meant "reward the early-comers in _some_ way."

Yes, early miners of a cryptocurrency have an advantage, in the sense that there is less competition in computing resources. However, when you factor in risk (i.e. likely return on investment and opportunity cost), this "advantage" may not seem worth it. It depends on one's risk profile, awareness, opportunity, and skills.

Thinking along game theory lines, when designing a system that requires up-front work, it seems clear that early participants will look for risk-adjusted rewards downstream.

I don't know if I agree with the "very difficult" aspect of your comment. Many real-world systems exist that are not clearly explained by game theory; humans are more complex than their theories.

Just to give one example, family genealogists typically are more than happy to share their historical research and family trees with others without expectation of personal gain. They do it largely because they want their ancestors to be remembered in the context of history. Of course, there is also some incentive for the _ genealogist_ themself to be remembered and perhaps to be perceived as important. But my point stands -- family genealogists don't expect to be compensated at all, much less in a pyramid-scheme kind of way. They are happy to create something of value and share it. Of course, a big difference between these family trees and cryptocurrencies is that the former are non-rivalrous.

[shieVae4I]

I suppose I was implicitly assuming "in an economic manner". You're very right that humans do things that narrow, economic game theory can't account for (e.g. the RAND secretaries playing Nash's "So Long Sucker").

To put my argument in context: consider IRC. An economically motivated BFT system would give the IRCops and channel ops some kind of quantified share of power that they stand to lose if they defect. But doing that, I think, would not just be a difficult programming exercise; it would also undermine the ops' intrinsic motivation and lead to a worse environment.

So the likes of Bitcoin work well if the protocol is designed for (game-theoretically) selfish participants whose interests lie mainly in the number going up. But that's a very narrow niche. Your genealogy site would probably suffer if it had a "number of family names added" counter and the users' status was measured entirely by such a number.

[baby]

You’re talking about bitcoin which is not bft. And there’s no n=3f+1 limit, see algorand.

[shieVae4I]

Allow me to quote the Algorand paper, https://people.csail.mit.edu/nickolai/papers/gilad-algorand-...:

> Weighted users.To prevent Sybil attacks, Algorand assigns a weight to each user. BA⋆is designed to guarantee consensus as long as a weighted fraction (a constant greater than 2/3) of the users are honest. In Algorand, we weigh users based on the money in their account.

That 2/3 is your n=3f+1 right there. It's just that, as I said of PoS in general, the n is based on coin holdings.

This is not controversial, as the authors of Algorand are completely aware of the limitation:

> Most Byzantine consensus protocols require more than 2/3 of servers to be honest, and Algorand’s BA⋆ inherits this limitation (in the form of 2/3 of the money being held by honest users). BFT2F [35] shows that it is possible to achieve “fork∗-consensus” with just over half of the servers being honest, but fork∗-consensus would allow an adversary to double-spend on the two forked blockchains, which Algorand avoids.

[baby]

If you read the whole paper you will see that it scales to more than the 3f+1 participants by randomly selecting a committee (of 3f+1 validators) at each round.

[pornel]

I think part of the problem is that mining was a very clever solution to a logical puzzle that seemed impossible to solve. Like bloom filters, it's tempting to find a use for it, because the algorithm is so cool.

However, the theory and practice turned out to be very different. Neat algorithm can't be used as an excuse for the whole ecosystem that emerged around it: insatiable electricity usage of PoW, pump and dump, ransomware, money laundering, etc.

[tzumby]

Yes, this is the kind of discourse I would expect to see around blockchain technology from a group of smart nerds. I can't argue with that, the oldest blockchain kind of failed to prove itself valid in significant ways.

[qualudeheart]

Right. The hilarious volatility and scams are imo much less interesting than this.

[xpe]

Your comment is unclear.

> What is surprising is how large the number of techies who don’t get this.

What do you mean by "this"? Do you mean all of the points made by the author in the tweets?

Or do you mean "the following" instead of "this", referring to what you think is most innovative about Bitcoin?

[tzumby]

If you don’t believe that there is value in a distributed self replicating Byzantine tolerant state machine, that’s fair, everybody is entitled to their opinion.

[onlyrealcuzzo]

Considering that it hasn't produced any value of note in 13 years, I'm gonna go out a limb here and say I don't think it's worth 2% of the entire world's wealth - and certainly not 3% of the world's energy.

[xpe]

In this comment, I'm assuming you are talking about Bitcoin.

> I don't think it's worth 2% of the entire world's wealth

Taken literally, if 2% of the world's wealth is in Bitcoin, so isn't that its value, by definition?

I'm not trying to nitpick. I'm trying to push back on what seems to be a glossing over of the fundamentals.

I wonder if what you mean is that you don't think it is _wise_ for people to put so much of their assets in the form of Bitcoin.

Now, if we want to have the discussion of "how has Bitcoin affected the economy?" it is not enough to only look at a snapshot of how wealth is allocated at one instant of time. We should be looking at a differential view; i.e. before and after.

In particular, an economist would ask questions such as...

1. How has Bitcoin changed transaction costs? Who bears these costs? Over what time frame?

2. To what degree does the structure of Bitcoin provide advantages relative to other assets?

3. How does the development of Bitcoin factor into a longer-term view of experimentation and innovation?

4. To what degree has wealth been reallocated due to Bitcoin? To what degree did it disrupt existing wealth patterns?

And so on...

[onlyrealcuzzo]

People paid a lot of money for Enron stock before it exploded. Did that mean it was worth $70B the day before it exploded and suddenly it was worth nothing when people found out it was actually all a fraud?

Enron wasn't actually producing any value. Neither are self-replicating Byzantine fault tolerant systems.

Anything they can do can be done hundreds of orders of magnitude more efficiently. And nothing they are doing /requires/ the inefficiencies.

I mean, obviously people at Enron were doing /something/ in the same way Bitcoin is. And obviously investors were paying a lot of money for that at one point and then stopped paying for it when it became clear that /something/ was mostly fraud.

[xpe]

There are some pitfalls in comparing a particular stock (e.g. Enron) against a class of cash-like assets (e.g. Bitcoin).

If you want to make the comparison, perhaps you could lay out your logic?

So far, I haven't found a clear, direct, fundamental argument in your comments. I see some loosely related examples and some rhetorical questions. I am interested in your justification and terminology.

[xpe]

Market capitalization varies over time, sometimes quite rapidly, because it (partly) depends on human perception.

Am I detecting some conceptual discomfort with how market valuations work? You wouldn't be the first.

[xpe]

What is your definition of value / worth, as used above?

I can see that it is not market capitalization.

[mosheroperandi]

Post-communist Albania offers a interesting parable: https://www.imf.org/external/pubs/ft/fandd/2000/03/jarvis.ht...

Various pyramid-like enterprises sprung up in the new, underregulated markets. At their peak, the schemes were worth upwards of half Albania's GDP, were participated in by 1/3 to 1/2 of the population, and enjoyed support from the government.

For a time, all investors enjoyed good interest payments. But all the popular adoption and shared belief in value couldn't fix their underlying insolvency and the ensuing collapse once new deposits dried up.

Whatever quantifiable value, if any, Bitcoin produces for being a BFT ledger is not nearly enough to offset the constant, massive value drag require to pay for mining costs. That x% of the world's wealth is currently tied up in it is not proof against its fundamental insolvency.

[xpe]

> not nearly enough

Can you estimate (1) Bitcoin's value, as you see it and (2) the mining costs?

Various actors (of varying legality) benefit from Bitcoin to varying degrees.

[xpe]

From what I can tell, much of this discussion seems to be characterized by people miscommunicating.

> Considering that _it_ hasn't produced any value of note in 13 years...

What do you mean by "it"? Do you mean (a) Bitcoin in particular; (b) some/many/most/all cryptocurrencies; (c) a distributed self replicating Byzantine tolerant state machine? Or something else?

The comment by tzumby meant (c). Generally, the HN thread was about (b), best I can tell.

[baby]

You’re barking at the wrong tree. Bitcoin is not bft

[xpe]

I don't think anyone is denying that there is value in any subset of the following:

* distributed systems * self-replicating systems * Byzantine fault tolerant systems * state machines

The tweets by Pinboard are largely talking about social and societal issues. Both technical and societal issues are important. They are intertwined.

[baby]

Pinboard ignores the tech because he doesn’t understand it, yet he talks about societal impact without understanding it as well.

[xpe]

This is a blanket, unsupported statement. It isn't persuasive.

> ignores the tech ...

Not true. He doesn't ignore the technology of Bitcoin.

> because he doesn’t understand it

Please show evidence of this. You are not "in his/her/their head", nor is it likely that you've had a conversation where you can assess. It is more likely that Pinboard disagrees with you largely because of value differences.

[baby]

“Of bitcoin” yet targets all cryptocurrencies

[tzumby]

this = blockchain + proof of work / proof of stake , in my comment. The crypto currencies are a requirement to fulfill the Byzantine tolerance. They need to have value in order for the incentives to make sense

[xpe]

> The crypto currencies are a requirement to fulfill the Byzantine tolerance.

I find this writing to be unclear and/or circular.

Are you saying the following? _If_ the primary goal of a system is Byzantine fault tolerance, using a cryptocurrency is a requirement?

[tzumby]

Sorry, you have a point. I re-read my comment and it's not very clear. We need two things to achieve the Byzantine fault tolerance:

- a way to ensure that users of the system can't cheat by creating a vast number of nodes and inflate their "voting power" (think of a poll where you can get behind a proxy to vote multiple times because the only security is weather or not the IP voted): this is the hash-cash, proof of work computation that ensures you have real hardware behind your "vote"

- an incentive to mine and secure the network: the cryptocurrency. The amount of cryptocurrency you mined is just a translation of the energy and work you did to secure the network the same way as a banknote is a conversion of X number of hours you worked for someone.

The question, like I stated in another comment, is whether or not this distributed ledger is useful. Let's say we haven't discovered the killer use case for this yet (putting my skeptic hat). Is it fair to dismiss it just because of all the scams related to its current use ? Do we do the same for email just because there's so many phishing attacks that happen on the protocol ?

[Graffur]

I've been asking for similar clarification on other comments. I think this whole thread has become useless.