part 0: my penance

about two or three years ago i made a post like "cr*ptocurrencies are good actually"

i was very wrong! this is because i am a computer fool who was taken in by the swan song of cool technology. i was that "wow! cool robot!" meme. there are some things i will stand by. (to be fair, the robot itself is very cool)

btw, this post is unrelated to FTX falling (which is very funny! lol! anyways there have been constant computercoin disasters that i don't care about, because that's not the cool part of the cool robot.)

part 1: why cryptocurrencies seemed so cool

i learned about cryptocurrencies way early on, like in 2011, because i had zero friends and was very online. tragically, i was a child, and did not have $10 to play around with. if i did, i would be a cryptocoin millionaire and also i'd probably have physically transitioned and also would have gotten a lot of medical procedures done.

but as i said, i'll stand by some of that "cool robot" stuff! there are all these creative proof-of-X concepts, like proof-of-storage, which could do something neat! you have eth*reum and others which provide a type of virtual machine on the ledger itself! and you see so so many of your favorite and obscure datastructures and cryptographic primitives: merkle trees, zk-snarks, etc. (And yeah, proof-of-stake / eth 2.0 is here, but that was never really in question)

and there's legitimately so much cool work, it's just such a cool sounding platform to build on!

there's a few problems:

1. u need $$$$ to actually deploy a dapp you "own" in a meaningful way. (as in, close to "start your own ISP" money)

2. 99% of your peers will be insane snakeoil salesmen

3. (also all of the other tiny problems, like solidity language being bad. like, imagine actually using it? lmao)

Anyway, there are a few things a cryptocurrency (or similar construction) could solve! (E.g. It'd be straightforward to require payment for receipt of email, like a stamp. This would reduce spam and would be a natural usecase for netbucks.) And I wouldn't be surprised to see parts of their architectures picked apart for real-world uses. (E.g. A non-decentralized ledger with web-of-trust or CAs in place of a consensus mechanism just for recording transactions.)

The only thing that cryptocurrencies help with are private payments, but this takes extensive domain knowledge in (1) opsec (2) money laundering and (3) using cryptocurrencies.

Anyways, the point of this is that cryptocurrencies are bad even at their best / most technically justifiable use cases!

part 2: why cryptocurrencies are bad at their best

Exhibit A is the Eth*reum Name Service. It's a pretty straightforward DNS-equivalent and has solid technical foundations. I can find no fault with it except for the people. The problem is simple: Anyone can mint an ENS name and auction it off. So, you have wikipedia.eth and a myriad other domains bought by the earliest bidder to be resold.

Any big name that might have some interest in decentralized apps (like wikis or games) are suddenly being held hostage by snakeoil nerds. The platform has little-to-no profit incentive, and when you put a three-to-five figure barricade up, who's gonna give a shit?

The funny thing here is this: Despite strong tech foundations, ENS is useless because the only people using it are scam-artists.

part 3: how NFTs fail to meet their meager standard

Exhibit B is NFTs, which are hilariously worse!

The reason they are worse is that (1) like other aspects of cryptocurrencies, they have some contrived use cases, but (2) unlike other aspects, these use cases still make sense even if your peers are selling snakeoil. For example,

• Conferring ownership of rare items in a dead MMO. (E.g. If Runescape died today, minting "partyhat NFTs" for each partyhat owner would actually make sense.)
• As a replacement or augmentation for notary services. (This is one tangible benefit! Being able to self-notarize. But that's not sexy/profitable enough.)
• As a replacement for existing key-signing / web-of-trust infrastructure. (It's like if we took PGP, made it 10x harder to access, 100x sillier, and 1000x more expensive. This is not an improvement!)
• As a framework for sharing items between games? I guess? (Is JSON not good enough?)
• Used in part with legal contracts for conferring ownership of IP.

This last one is the most interesting. Used in the best possible way, an NFT should provide an easy-to-verify timestamped hash claiming ownership of the data which yields that hash. An artist could, feasibly, upload their artwork to their own site/service, while also minting an NFT to claim ownership. The art is licensed (or rights transferred entirely) to whoever owns the NFT.

The problems with the majority of NFTs get increasingly funny:

• This is legally risky, for being unproven. That's not itself a flaw!
• But this legal framework is not used on popular NFT services! You don't own the property rights to the NFTs you buy any more than I own the right to Pikachu for owning a Pikachu card.
• Worse, these NFT services only store URLs not hashes. This kills any of the purported benefits an NFT could bring.
• Even worse worse are the rest of the problems: The (former) environmental costs, the (existing) costs of minting, and all the outright stealing/plagiarized works.

Lol! But, wait: I promised to tell you that NFTs are funny because they can be useful, despite being used for scams in practice.

part 4: looking at the legit use case for NFTs is even funnier

But what about all the other usecases? Well, no big MMOs have died and reimbursed players with NFTs yet, nobody is seriously using it for cryptographic infrastructure (LMAO COULD YOU IMAGINE), no "real games" are using it (yes, im gatekeeping candypop bubblesock saga), and nobody is using it to confer IP.

The *one* use case that is used in practice is self-notarization. No qualms or jokes about it, if you want to self-notarize a document, you can use cryptocurrencies, and if you've used online notary services, you might have done this already.

The funny thing is? You don't need NFTs to notarize! You can notarize in any transaction.

Even funnier? You don't need eth*reum to notarize! You can notarize in any transaction (on basically any major cr*ptocoin). You don't need the "smart contract" capability of eth*reum.

The only benefit that "the bl*ckchain" brings is that, at any given time, there are a few thousand (or tens of thousands) of servers out there that can verify your self-notarized document. It's likely some kind backups of "the ledger" will exist for a few decades, even after all cr*ptocoins officially die in 2029.

part 5: git will still be around after all computercoins die in 2029

As I said above, there are some neat, creative, and new ideas that came from all the sad motherfuckers who wasted years of their life building up their favorite cyberdollars. I expect all of these to be stolen and repeated, including the ability to self-notarize.

See, the reason the self-notary works is because (1) merkle tree (2) decentralized and (3) duplicated on a lot of machines.

DNS, Git, and certificate authorities each get us 90% of the way there. Git's the best candidate, because all it'd require is replacing the SHA-1 commits with something better, like SHA-256(SHA-256(x)). (That's not a buttcoin reference! Doublehashing is useful for mitigating length extension attacks.) That, and collaborate with second-parties to maintain duplication of the git history.

So, what's the TLDR / ELI5?

Cryptocurrencies over over over over promise and under under under deliver. Every ounce of passionate technical work is surrounded by a ton of investhit snakemen. Where they does deliver, it's primarily in "extremely contrived use-cases that aren't used in practice", often better-served by existing infrastructure.

One exception is making private payments. Computercoins can add a sliver of extra privacy! But you need to Already Know How To Do This in order to not footgun yourself here. Given this is the bare minimum use case, you'd think this shouldn't be prohibitively difficult!!

The other exception is "self notarization", which allows you to prove that you had a document with contents X and a given time. This is the literal one and only use case! And that's hilarious! To have this cyberpunk-libertarian-computercoin-netbeanz-utopia reduced to a notary service is HILARIOUS!

As it stands, cr*ptocurrencies actually provide a really convenient infrastructure for that. (But that infrastructure could exist without cr*ptocurrencies, and we're 90% of the way there with Git.)