The attributes that _I_ would like to see in a cryptocurrency

in #cryptocurrency7 years ago (edited)
  • Inflationary
  • Hierarchical
  • Scalable
  • (Ideally, I'd love it if all of the "work" (computation) that is done to generate the currency actually did something useful)
  • Might be nice if the 'work function' or whatever wasn't able to be run on GPU's, only CPU's.

(And of course, the features that we have in just about every cryptocurrency that's out there - decentralized, electronic, anonymous-ish, etc. I won't write about that stuff because we know it already)

A little bit about why each of these things -

Inflationary

Currency should be similar to nature. In nature, things that are 'old' get eaten up by things that are 'new'. "New bucks" should be worth more than 'old bucks'. Deflationary currencies like Bitcoin are the opposite; and encourage people to hold on to their money, effectively rewarding the "old" and punishing the new.

Secondly, there's a practical reason for this - "bad money drives out good" (Gresham's Law). People will spend the inflating currency, and hold on to the deflating one. That makes my new currency become one that is actively spent, all the time, and 'drives out' the "good" money (which would be Bitcoin, et al).

Hierarchical

Bitcoin and every other currency that I've seen all tends to be universal - if I buy a cup of coffee in Ocean Beach, San Diego, that transaction will show up in a ledger running in Adelaide, Australia. And it doesn't need to, it makes no sense to occupy the bandwidth. And it also means that there's a very strict chain-of-custody you can track for currency, and some privacy here might be nice. The rest of the internet works this way - when I grab an HTTP web page from a server in Mountain View, servers in the UK never see the request or response - because they're not involved, and don't need to be.

I believe Cryptocurrency should be the exact same. I'm imagining blockchains of blockchains, and them only talking to each other because a transaction passed across some hierarchical border. I'm imagining, maybe, a chain per continent, or something like that, and then maybe if it got popular enough, subchains per region, and maybe even sub-sub-chains. Of course, technically, I have no idea how we could manage to do that, but that'd definitely be one of the problems to try and figure out. I'm taking inspiration from a lot of the fundamentals of the Internet, here - routing protocols, layering, that kind of stuff.

Scalable

The whole "the block size needs to get bigger" thing sucks. Better to build in some of those lacks of constraint into the protocol. Maybe block-size is the trigger that creates a sub-blockchain, or something? E.g. if your "California" blockchain starts to get so busy that the entire blocks that contain transactions are starting to get full, that's when you spin up your "San Francisco," "Los Angeles, " and "San Diego" blockchains, maybe? Or maybe they somehow dynamically split off and form up some how?

But, regardless of the solution, the problem, simply stated, remains the same - "if something gets big you don't have to suddenly force everyone to upgrade their software."

Useful Work

It breaks my heart that all the Bitcoin mining that's done out there is all using pointless computations that serve no purpose other than wasting power. I would love it if it actually did something beneficial - like SETI at home, or some kind of cancer thing. I think I read about one of the newfangled currencies out there doing something like this. I would totally steal that concept.

CPU only?

This is really just because I like the egalitarian nature of this. I like the idea that the "little guys" can mine just as hard as the "big guys" and might somehow manage to get a payoff. Highly parallelizable work functions end up making people get custom ASIC's and GPU-rigs and I don't like that as much. But that's really just a "nice to have."

(edit)

Notes

Might be interesting to have it be IPv6 only - no need to waste IPv4 space with various nodes and whatnot, and it's a brand-new thing, you can skip all kinds of routing and port-mapping misery and whatnot by just starting out having it be IPv6 only. And who knows, maybe the built-in addressing and networking schemes in v6 might inform or modify how the addressing would work with this currency, too?

Sort:  

Interesting thoughts. Thanks for sharing.

  • Inflationary

New bitcoin is created all the time and many consider it inflationary for that reason. When you say new should be worth more than old, you're breaking fungibility which is a key property of money. What about those who save? What about those who see their purchasing power go away at no fault of their own? Inflation favors those who can easily get more money (i.e. those who already have a lot of it) over those who only have a little but are saving what little they have to try and get ahead. It harms those who need every scrap of purchasing power possible. I can see an argument for expanding the money supply based on population or some other natural mechanism, but I'm also okay with some deflation. The tech industry, as an example, has products getting better, faster, and cheaper and that's a good thing. With good price discovery and unmanipulated money, deflation could be a good thing, as long as it's not hyper and prices can respond accordingly. I also think the good money driving out bad money thing will become irrelevant soon. The Exodus wallet, as an example, has shapeshift built in. See Andreas Antonopoulos' talk on money as a form of expression like language. It won't matter how someone chooses to express themselves because we'll be able to seamlessly go from any currency to any other. Centralized monopoly control of currency creation makes this sound fantastical but it's already starting to happen.

  • Hierarchical

Interesting. What you're describing sounds a bit like the lightning network, segwit, and side chain solutions people are coming up with. Tying it to physical location seems like a huge mistake to me. If there's a hierarchy, there's a level of centralization, a point of weakness which governments and banks can exploit. What if a province in China decides it no longer wants to allow transactions to flow there? All your money is effectively worthless if you happen to live there. The blockchain protocol works and solves the Byzantine General's problem precisely because it is not hierarchical. Decentralization is important for security, resilience, and censorship resistance. I know some cryptos use master nodes or witness nodes (like Steemit, BitShares, and soon EOS), but I think it can be done in a dynamic way. Meaning if a group of nodes become threatened or start being bad actors, they can be replaced (why I love DPOS, delegated proof of stake).

  • Scalable

EOS seems to be way ahead of everyone here. It does things in parallel instead in serial. Check out Dan's posts on this stuff. It's amazing. The block size debate is only a problem in bitcoin because the protocol has no built in governance model, and its pay to play instead of bandwidth based on stake (like graphene systems). Steem, as an example, raised the block size recently like it was big deal because the witnesses are motivated to serve the users or get voted out. There was actually a bug which got exposed when we (the witnesses) did so, and we quickly changed it back. Hardware improvements will continue to make this a non issue for most. Also, with DPOS, you don't need miners at all for security and performance. Much more efficient.

  • Useful Work

Gridcoin does this and it's pretty cool. Still though, I prefer DPOS which doesn't do any "work" at all. It's all via scheduled witnesses (see Dan's blog on the missing DPOS white paper). It's cooperation over competition. Massively more efficient which gives Steem 3 second block times instead of bitcoins random times of ten minutes or more.

  • CPU Only

Another reason to avoid proof of work. ASIC development technology will continue to advance (nanotechnology anyone?) and algorithms that were supposed to be ASIC or GPU resistant aren't. Not only that, but again, those with money will just buy up clusters of super computer CPUs / GPUs anyway so the vision of egalitarian "1 CPU, 1 vote" doesn't really work in practice.

I do think you've got some interesting ideas here, so I'm sorry if my review seems all negative. I'm on my phone, so please excuse the typos. I usually like to include links to stuff in talking about, but you'll have to google. :)

I know that more and more bitcoin is being mined every day, but the amount reduces over time. I consider that deflationary. Hell, even if it didn't reduce over time, I would still consider it deflationary because it's not a geometric progression.

Yes, inflation does hurt savers. And logically, you can probably guess that it will benefit debtors.

But it also punishes stagnation, and that's the argument I'm trying to make. Money sitting in a mattress will decay over time. It should. Money invested, however, will not. So the impetus is on people to actively invest their money in the economy, instead of just sitting on it.

With bitcoin, the long-term trend so far is that the best thing is to just sit on it, it'll probably gain value. That sounds more like buying an expensive bottle of wine to leave in your cellar, or an expensive painting to hold on to - it doesn't sound like currency.

And, of course, I'm not arguing for crazy high inflation rates, I'm talking closer to 5%, maybe less.

The FED's target is, what, 3%? Maybe that works because of population growth, but it still feels wrong to me. I get that money should have velocity for it to be useful as money, but to me that's partly a matter of market cap here. Cryptocurrency is still pretty small compared to the global finance market. Once (if) it gets to the target price many envision, the velocity should go up as the price stabilizes and no longer providers huge incentive to HODL. I don't think we need to inflate the value away prematurely in order to get there. In some ways, that might even delay the process.

I'm coming at this from an economics perspective that is concerned with boom and bust cycles. IMO, encouraging investment via money manipulation beyond what the market itself demands causes these problems.

I think Bitcoin (any many other cryptocurrencies) will have problems as currency if they continue to charge a fee for usage. That's what I like so much about STEEM and SBD. No fees involved, just bandwidth based on stake.