Hacker News Reader

Comment by 110011 | Hacker News Reader

It's scary to see so much faith in bitcoin given it's horrendously wasteful proof-of-work scheme to avoid double spending.

There are better alternatives that minimize latency nearly a hundred-fold and avoid proof-of-work altogether. Example: Algorand (https://people.csail.mit.edu/nickolai/papers/gilad-algorand-...) Even if this new cryptocurrency is "just" a proof of concept there are many others like it that are much less wasteful than bitocin, and this should make people think twice before piling their faith on an important but ultimately flawed first attempt at a usable cryptocurrency.

View on HN

Thanks for the link, I hadn't heard about Algorand. I don't claim to be really knowledgeable but I found an article on Coinbase that raises the question of whether Algorand needs (or should have?) incentives, so it looks like there's some debate as to whether Algorand is really viable?

[1] https://www.coindesk.com/no-incentive-algorand-blockchain-sp...

I keep being told that Bitcoin will eventually adopt something less wasteful, but given how rocky recent evolutions of the protocol have been, I don't see why I should believe that, versus an altcoin supplanting it.

People seem to like Bitcoin as a gold alternative precisely because it doesn't evolve. But I don't see how that jives with the fact that its energy consumption is completely unsustainable.

Wasteful by what standard? Right now each day it consumes ~$1m worth of electricity in order to protect ~$1.5 billion of new payments. Economically, this "electricity safety fee" works out to less than a tenth of a percent.

Wasteful as in we can do better. Did you see the link in my comment? One of the authors of that paper is actually a Turing award winner.

Ultimately we've reached a point where there are more efficient alternatives to having hundreds of people computing billions of hashes to reach consensus. To put this in perspective imagine a group of people sorting by randomly permuting and checking if it works, while stubbornly refusing to adopt near linear algorithms.

Really interesting paper! +1 on trying to solve this problem, I was just trying to put the size of the problem in perspective (the house isn't really burning down).

I suspect people underestimate the game theoretic consequences of stake systems like Algorand as compared to PoW. (For example, the dynamics of contentious forks may look quite different.) I don't know which is better! But those types of consequences are likely much larger (for better or worse) than the electricity savings.

Well said.

0.1% per day? Are you sure? That's a 3.5% fee per year! That's astonishingly high for any sort of custodian's fee.

oh it's 0.1% of new payments per day.

if you wanna compare it to all stored value instead (about $100bn), then the "electricity safety fee" comes out to ~0.001% per day.

(these comparisons are just meant to help put the electricity costs roughly in scale against the benefits they help provide)

Ah yes, thanks, I misread. So it's 0.35% per year so still quite a lot.

[My initial calculation was wrong. It would have been 35% per year, which would have been more obviously incorrect!]

Reply on news.ycombinator.com