This blog post[1] from Vitalik explains a little:
> Unlike wealthy countries like the United States, where financial transactions are easy to make and 8% inflation is considered extreme, in Argentina and many other countries around the world, links to global financial systems are more limited and extreme inflation is a reality every day. Cryptocurrency often steps in as a lifeline
Some of the users are on battle tested decentralized solutions like Ethereum and its rollups. Others transact through centralized exchanges because fees are cheaper. And others use more centralized blockchain networks such as Tron to avoid high fees. On one hand, using central exchanges doesn't match up to the decentralized promise of the blockchain. On the other hand, it's cool to see people using crypto without caring that it's crypto. They just want to have access to more stable currencies for payments and transactions!
I've seen this myself, from talking to LatAm businesses and from a friend in Argentina.
If you'd have told anyone in 2009 that an open source decentralized currency would eventually become a trillion dollar market, or that it would be recognized as legal tender in some countries, no one would have believed you. Yet here we are with people complaining about how it hasn't yet obsoleted the US dollar.
However if im making bets for my own projects, with my own money, I’m not seeing the incentives for the sort of massive change to actually happen. Love it or hate it, you can’t just burn down the world and start over, so something has to be aligned with the gate keepers to make this work. I don’t think it is. Without adoption it’s ceases to be valuable to those outside the magic circle, which means it’s not worth adopting.
I had been following BTC since ~2009 (when a CS PhD friend of mine at uni introduced me to it, as his research project was on distributed ledgers) - but from the very start it was made clear to me that projects like Bitcoin would never be suitable for microtransactions due to the reasons that became clear to everyone since then: at the values of microtransactions (i.e. under $3 USD) the cost of committing that transaction within reasonable time-frame for a microtransaction (say, less than a minute for a mobile-game IAP purchase) is simply prohibitive and makes the legacy incumbent card networks (Visa, Mastercard, etc) seem like nimble, customer-pleasing startups.
The reason is that "ML" (as a proxy for algorithmic processing of a variety of data) is not particularly adversarial to pre-existing technologies. While this is not entirely true (there is always a hidden or explicit tension between automation and expert assessments) by-and-large this tension can be managed. The new tech gets bolted on the old.
In contrast crypto sought to overthrow all pre-existing monetary and financial system technology with proposals that are naive, half-baked and ignore (or reinvent in unacknowledged manner) vital aspects.
The practical implication is that crypto cannot carve a legitimate niche and keep iterating.
The main outcome of a lot of wasted energy (in all senses) is to point out that indeed, digitization opens up the way to evolve the financial system.
So much we knew, but now it has been drilled into the heads of large swaths of politicos, regulators, bankers etc that are completely tech illiterate.
I think you’re thinking of LLMs which only became possible with the shear amount of conversational data we have today. They obviously share a linage with ML, but it’s a form of training that was impossible several years prior. You can’t say that IBM DeepBlue was an early LLM and measure the timelines like that.
If you know of any watershed moment for crypto that needs to happen, I would love to hear it! I just don’t see anything happening on the horizon.
It’s possible in 60 years we might see similar leaps from distributed ledgers, zero knowledge proofs, verifiable computation, fully homomorphic encryption, and other tech being spearheaded by the crypto sector.
ZK-proofs are more advanced than just cryptographic signatures. The important point is being able to prove something without revealing the proof itself. Classic example is proving that you have a solution to a sudoku board without revealing the solution.
Most ZK proofs actually rely on proving something with some probability. That means I don't prove to you with 100% certainty that I know the sudoku solution, but rather the chance of me lying about knowing a solution is < 0.00000001%. Traditionally, zk proofs required many iterations to justify the probability of the proof. Blockchain use cases advanced research in "succinct" proofs. If you search for ZKSNARK, ZKPLONK, and ZKSTARK you will find some examples.
To compare to a cryptographic signature, I can use classic cryptography to prove that I know a value by sharing a signed hash of that value. However, you can only verify my proof when I reveal the pre-image. Doing the proof entirely on encrypted data is homomorphic encryption, and modern zk proofs make use of homomorphic encryption to prove things about arbitrary computation.
Distributed ledger technology, though yes that is clearly a blockchain technology that wouldn’t exist otherwise. Fair point.
No -- at least not in the sense that RSA uses. Zero knowledge proofs verify computation not data. The zero knowledge property means you can verify a computation ran without revealing its input. Some researchers call it "moon math."
A lot more than the very marginal value ML did for its first decades of existence. DeepBlue came many decades after the field came into being.
And I’m very aware of LLM and the different types of ML. It still stands that it was very marginal for decades. But it would be easy to write some dismissive comment in the 80s that the only thing it’s done is play chess for millions of dollars - exactly like crypto haters do today despite it being a completely nascent field.
Crypto has been around for nearly a decade and a half and has been really useful for malware and scammers and not much else. It's fundamentally broken as a currency because most coins are designed to be deflationary so they behave like securities. It's fundamentally broken for micro transactions because transactions are ridiculously expensive and slow by design.
If that simile of ML is like crypto is true, you stand to be quite rich! So best of luck! Best comments have net 0 right, so hopefully you’ll think of me and this thread when you’re a trillionaire eh? ;)
if crypto was about the everyday person, it would have been made useful