Yes, if the application and its users demand it. Reversibility is possible in peer-to-peer and decentralized payment systems. But not every app needs reversibility and the centralized intermediaries that come with that.
Craigslist explicitly tells people selling on it to avoid accepting online payments (like, PayPal) and only work with cash because the online payments are reversible and sellers constantly get scammed. You have to remember that transactions involve two parties, and it isn't somehow always the case that only the ones on one side matter.
The power balance simply isn't always "seller who is selling a large quantity of easily-replaceable items that might be a dud can easily amortize the cost of fraud across all of their payments by high margins"... sometimes it really is "seller who is selling a small quantity of expensive or precious items tends to get scammed by anonymous buyers who abuse chargebacks ".
> ...sometimes it really is "seller who is selling a small quantity of expensive or precious items tends to get scammed by anonymous buyers who abuse chargebacks ".
This. He's right you know.
The chargeback system is rife with extreme abuse, and the banks will always give in to friendly fraud without checking the contents of the dispute. Both the customer and the merchant lose to chargebacks and friendly fraud.
Venmo[1] does not have reversibility. Cash does not have reversibility.
If users really wanted credit card functionality at the expense of its costs, you could build that as a layer on top, without forcing everybody in the system to use it.
And consider, most disputes is not just about the quality of a product. That is handled with regular refunds. Credit card disputes are a lot of times because of fraud or card skimming, which is very often because the system is not very secure to begin with. Secure cryptography can remove a lot of this type of fraud.
Open source, public and permissionless infrastructure for the web. Very different than Venmo which is owned by PayPal and limited to USA, and not programmable, no interop.
about the OP’s thread: centralized services like PayPal, Venmo and Stripe can close your account. Nobody can close your private key, even if you are on a L2 for fast payments you can use the escape hatch to withdraw without possibility of censorship.
The irony of this post criticizing DeFi and blockchain in the wake of FTX is that those decentralized applications are working just fine in this downturn. Aave and Uniswap aren’t failing, they are thriving.
Isn’t that the point? The protocols are built to resist changes by single entities and continue working as expected, handling user deposits non custodially, regardless of market activity. HN can declare a dozen crypto deaths with each new CEX and Uniswap will just keep filling orders for whoever is sending value through it.
I don't know, you're declaring victory because applications didn't stop working during the downturn, but I don't know that the problem was that applications stopped working during the downturn but rather that various service providers ran into financial trouble.
The goals of DeFi and web3 is to create structures that do not rely on centralized service providers, and are able to resist control of single bad actors and provide certain clear and transparent security guarantees. In that they have so far succeeded.
I wouldn’t call it victory yet. We still have years of figuring out which blockchain, DEX and DeFi models work and which will fail, most of this new tech is only a couple years into development. But the long term 10+ year vision seems clearer.
DeFi really is just a series of distributed (not necessarily decentralised) applications that enable parties to enter into agreements involving virtual tokens without the need for a contractual agreement between them. Not relying on contractual agreements places enormous constraints on what DeFi can do. For example, DeFi can't handle counterparty risk at all. Therefore DeFi doesn't remove the need for conventional financial institutions that rely on contractual agreements, such as so-called "centralised exchanges" and "lending platforms". It's disingenuous to say DeFi is fine because it didn't fail unlike those centralised exchanges, because DeFi doesn't (and can't) provide the same services that centralised exchanges provide (namely, custodial services and trading that isn't limited to virtual tokens).
DeFi never claims to provide the exact same fiat and off-chain services as CeFi and CEX. It has specific goals, like replacing custody with non-custody, or replacing a centralized exchange with an automated market maker that no single party can control.
This has different risks and trade-offs, obviously, but users who opted for Uniswap instead of FTX as their crypto exchange are probably pretty happy with their decision.
This is like telling someone who got a terrible haircut that they should have shaved their head instead. If they got a haircut it's because they wanted a haircut not a shaved head.
There are many users who just want to hold crypto and perform basic lending and exchanges, and would be willing to spend fractionally more in gas fees to achieve this with higher security guarantees.
Everybody at Barber CEX got a surprise head shaving the other week, meanwhile Barber DEX is still cutting people's hair normally even though it is more expensive and harder to find.
That's the protocol working exactly as it's designed. Everything was transparently visible. Anyone could see where the funds were at any given point, what the liquidation price was, and what the total liabilities were.
If you ride a fast motorcycle and fall off because you were speeding, it's not really the fault of the motorcycle, is it?
Yes, it is the fault of the motorcycle. You can "fall off" a motorcycle because of the design of a motorcycle, and it's something that doesn't happen with a car. Cars have risks when speeding as well, but they're different, and lesser, because of the protections afford by the car itself.
It's like you're subtly trying to highlight how terrible crypto is, but not realizing it.
Let's expand the decentralized web without building a crypto wall around it.
One failure of FTX and BlockFi is that users had no way to ensure that the centralized custodian was not running off with their on-chain deposits by directing them into unsound deposits. Vitalik is suggesting a cryptographic mechanism here that would provide better transparency as to on-chain activity of a CEX.
Day traders want to trade, no matter how much you try to tell them their trades are fictional or "have no use cases."
Partly UX, partly marketing, partly L1 fees and speed. Partly that a lot of people don’t know what Uniswap is. It gets a passing mention in the news if lucky, or more likely no mention, even though it’s the second largest crypto exchange on some days. Most crypto CEX investors are either unaware or too lazy to care.
A lot of crypto people do use Uniswap. The tone of Vitalik’s post is: what if we took some of the non custodial, on-chain, cryptographic proof things that work well in a DEX, and inject them into more CEXes so that even lazy users end up with better security guarantees.
Backed tokens are still a trust liability with the issuing party. They have a place and are interesting, but it would make sense to limit the scope of exposure.
Take Tether for example. Every time BTC starts to dip, USDT starts to de-peg. They are not at all uncoupled. Tether doesn't have the market cap to cash out all of the BTC, and never will. The amount of apparent value in the crypto market heavily outweighs any possibility of cashing it all out.
And that doesn't even begin to touch the questionable liquid assets held by stable coins. Tether claim to be holding 82% of "extremely liquid" assets [1], but I'm unsure it's proven or tested. From the report [2]:
> The valuation of the assets of the Group is based on normal trading conditions and
does not reflect unexpected and extraordinary market conditions, or the case of key
custodians or counterparties experiencing substantial illiquidity, which may result
in delayed realisable values. No provision for expected credit losses was identified
by management at the reporting date.
Substantial liquidity could be caused by, say, global inflation or recession conditions. But that surely won't happen...
Yes, that's why there's a spectrum of stable coins with varying levels of centralized control, governance, risk, etc. Stable-ish coins like RAI and LUSD are backed by ETH only, but they do not have a hard peg, they allow for some small wiggle room (typically <10%) that allows the protocols to catch up with dramatic supply/demand imbalances when they occur.
This math is not for end users of an exchange, it’s for developers and researchers building new exchanges. The UX does not need to feel that different than any regular app.
A lot of people like trading, just look at Robinhood. It would be good if an app like Robinhood existed that ensured cryptographic guardrails to prevent the platform owners from lying about their solvency.
If you look at crypto's daily price action, the majority of it is attributable to trading and speculation. I think this is pretty common knowledge.
If you feel stocks and equities is gambling, then you might also feel that this sector of crypto is gambling.
But there are other sectors of crypto that don't register on this price graph, maybe because price is not the only metric of their success, or because their volume is lower. Smart contracts, non custodial wallets, ENS, trustless payments, DEXes like Uniswap, are all interesting and valid use cases of crypto and blockchain tech.
Vitalik isn’t running any DEXes, he is not in a position where he can steal or move user funds locked into a DeFi contract. He could suggest a change that might do something malicious at protocol level, but the rest of the developer community would reject it.
CEO of a bank can't do that too. The board can though. And of course customers can reject that decision and switch bank. But in reality that won't happen both in the bank case and in the tokenbro case. We have already saw how Vitalik stole lawful tokens from the receiver of The DAO program (code is law after all), and everyone has supported him. Exactly the same as banks can do, only without outlandish claims.
A double spend attack is a type of fraud for sure, but obviously one of the remedies for a double spend attack is to fork the main chain and change the consensus to nerf the attack.
Ultimately layer 0 of a blockchain is the community that uses it, and if they decide to fork en masse, they will do so. It's an essential property of the system itself. Blockchains would not be antifragile if they could not fork.
Then the problem becomes the same, who is empowered with calling a fork? If they are just by users who are using it, how is it different than having an election etc.? Except here the agenda comes from a shadow organization within a "decentralized" system. I would much rather my votes happen in public with everyone's consent.
I don't see how double spending and forks are the same thing. Double spending is when the same coin is spent twice on the same chain. When you fork, you are creating a whole new future history (sorry, can't come up with a better word). If I have a ledger for my business, and someone created a copy and added different transactions to it, that would not be double spending, right?
Nothing really got stolen in the DAO case the way I see it. ETC still exists, its just that nobody wants to use that chain. Code is still law, but the users decide which code to run. I can see the beauty in it.
