[EM] Election-Methods Digest, Vol 210, Issue 62

[Kristofer Munsterhjelm]

On 29.01.2022 19:43, culitif at tuta.io wrote:
> Hey Daniel,
> 
> Regarding actual critiques of the Bitcoin protocol, besides the
> proof-of-work protocol you pointed out, most of the concerns have to do
> with scalability. Solutions to this problem usually rely on a tradeoff
> between security and increased speed (like increasing the block sizes)
> and are often not long term solutions. Bitcoin has already had nearly
> 100 hard forks mostly related to issues like this. There's more
> long-term solutions like the Lightning Network happening (which allows
> transactions outside of the blockchain), but those have major issues
> with adoption. There's also many concerns about the centralization of
> it. Centralization is a major problem amongst most blockchain protocols
> and to date there's no foolproof solution most people agree on. The
> proof-of-work protocol is susceptible to centralization of the miners
> (as seen when a single coal mine in China flooded and the hash rate
> plunged 35%) and also rely on a small number of nodes. In general, most
> crypto enthusiasts agree that proof-of-work is fundamentally flawed and
> few newer protocols rely on PoW. 

As the perennial wet blanket guy here, it probably comes as little
surprise that I'm pretty opposed to Bitcoin too. I think there are
serious, possibly even fundamental limits to the cryptocurrency concept
itself.

The core innovation is based on the idea on using some sort of costly
competition to keep people honest. This competition is either one of
resources (proof of work, time-space, disk, etc), or one of previous
investment (proof of stake).

If it's the former, then the mechanism will waste materials and/or
energy. This is a great loss as that energy could be used for other more
productive things; and if whatever it uses has economies of scale, then
there will be an implicit centralization anyway, as a few big operators
will be better at running miners than a large number of smaller ones.

If it's the latter, then the centralization is even more direct, as
those who already have power gain more power.

Beyond this, blockchains are essentially signed append-only data
structures, except the digital signature is replaced with a proof of
some sort. This structure can't be pruned. This means that for any
serious transaction rate, the log will grow very quickly and soon become
difficult to maintain. Either you have a low transaction rate (and the
system is not a good currency) or you have a high transaction rate (and
disk space becomes a problem).

But ultimately, I think my issue with cryptocurrencies is political. I
won't go into specifics about that as I don't want the more ardent
enthusiasts to stumble across this post, but I could explain in private.
But it's something that even fixing all of the above would not fix.

-km