>Aave is used for being able to borrow against your crypto assets, so if you need a loan you don't have to sell, this isn't a service any bank offers.
Every bank offers this service, unless you're just talking about getting a loan with crypto as collateral.
The block-chain can be thought of as cryptographically secure state. That is, everyone agrees on the state of the blockchain and no one can modify it save for modifications in accordance with the "rules". A smart contract can be thought of as a cryptographically secure program, i.e. it cannot be modified. So you have cryptographically secure inputs (the blockchain), and a cryptographically secure program (the smart contract), which means the output is also guaranteed to be cryptographically secure. That is quite a strong premise. You can give your program to anyone and they can execute it for you without you needing to worry about nefarious modifications. This level of trust allows for the creation of a global computing platform. Software authors need only to bring their program and can have it executed on-demand without needing to maintain their own infrastructure. That is quite an innovation.
Of course, in its current form, this "global computing platform" is actually pretty shitty. It is prohibitively expensive to store / operate on a lot of data. However, crypto-currencies are there first viable "apps" built on this platform. If you think about it a crypto-currency is basically a very primitive program that stores a series of transactions as two addresses and an amount. The total amount of data per transaction is tiny. And yet, just by operating on this tiny amount of data you can basically implement the entire financial system: currency, stocks, bonds, options, etc.
We're basically in the dot-com bubble for crypto. People are starting to realize that this new computing platform is cool and has a lot of potential, just as many realized the computers and the internet were cool in the late 90s. Yet, just like in the 90s, the tech sucks and isn't good enough to bring much of anything valuable to the market.
Not entirely coincidentally, aside from plain faulty implementations and operational security failings, these two problems represent the ways most successful attacks on smart contracts are accomplished.
I've gone through about three phases of opinions on crypto:
1. "Blockchain is an interesting primitive that you could probably build something neat (besides a cryptocurrency) out of! I can't think of anything good off the top of my head, but I look forward to seeing what valuable thing people come up with."
2. Smart Contracts are an interesting primitive that you could probably build something neat (besides rebuilding existing finance tools without regulation) out of! I can't think of anything good off the top of my head, but I look forward to seeing what valuable thing people come up with.
3. NFTs are an interesting primitive that you could probably build something neat (besides speculative digital assets) out of! I can't think of anything good off the top of my head, but I look forward to seeing what valuable thing people come up with.
In each case, the tech looks like a neat idea, but I continue to wait for the actually valuable use-case to emerge. At this point, I'm starting to just pattern-match all crypto to "unlikely to provide real value ever."
No
> A smart contract can be thought of as a cryptographically secure program, i.e. it cannot be modified. So you have cryptographically secure inputs (the blockchain), and a cryptographically secure program (the smart contract), which means the output is also guaranteed to be cryptographically secure.
The output is not guaranteed to be correct, what you intended, or even what you agreed to. Even the most trivial computer programme can have unexpected properties.
Why would I have to explain that here?
This brings me back to my main point. You clearly have not implemented a smart contact and have misconceptions about what it does. I suggest you implement one first. You can still think the idea is stupid afterwards. That’s fine. But right now, it seems like you are arguing about how you think the blockchain works, which doesn’t necessarily match how it actually works.
So this:
> Yes, the output is guaranteed to be correct.
is provably wrong and you either know it and are malicious or you don't know it and this discussion is useless.
I think what everyone is quibbling with is the assertion that this is the same as "correctness", which is (by and large) considered a different property that relates user expectations to actual runtime behavior.
In CS, correctness means with “behaves in a consistent manner with respect to a specification”. The smart contract is the specification, so yes the output of the smart contract is guaranteed to be correct with respect to the specification. Of course, there can always be be errors in the specification. If you have another definition of “correct”, let me know.
Canonically, within the industrial context, there are two processes for achieving compliance of implementation with specification ("correctness"). The first is testing, the second is formal verification. In either case, you need an artifact at a higher level of abstraction (the specification) and one at a lower level (the implementation).
In testing, someone who has access to the specification can evaluate the dynamic behavior of the implementation and determine whether it matches. In formal verification, you can do the same by static analysis and proof.
If you say "the code is the specification" then correctness becomes completely vacuous. The code is correct because it runs. Great. So what? If the implementation results in behavior that is completely unintended by the author (e.g. your smart contract contains a re-entrancy problem: a notorious source of bugs, or isn't robust to EVM stack exhaustion) and you still want to claim it's "correct" then I just don't know what to suggest - we're never going to agree, and I don't think your usage is anything like standard.
The contracts were verifiable and correct enough, it seems, but some of the infrastructure turned out to be faulty, leading to a platform that was not trustworthy.
If your code has bugs or errors or security vulnerabilities, those will all lead to incorrect behavior.
That's why smart contracts aren't useful.
Most contracts end in the happy case 99% of the time, it seems very worthwhile to automate that happy case and fall back to the traditional legal system in the failure case where either party isn't happy with the outcome.
It's just as true to say contacts aren't "necessary" 99% of the time, because people follow through on the agreement.
The name likely comes from options contracts, which are technically contracts but traded like equities. Most options are very simple tuple: (equity, strike price, call / put). The smart in smart contract can mean writing a smarter option contract.
I think they can be, I am not a lawyer.
But if our "legal contract" is "do what the code says" then the smart contract is a legal contract. Is it not?
If you just want to trigger actions based on something and then store it, you don't need the machinery of the blockchain.
Just like Tesla's AutoPilot, smart contracts were named that for a reason, to mislead (and make money along the way).
1. It is impossible to write code that always works exactly how you want, all of the time, forever.
2. Law exists. The legal system exists. The moment you have a problem with a smart contract, it's going to court.
You might as well save yourselves the time, expense, hassle and potential embarrassment by just going to a lawyer in the first place and having them draft a contract.
On loans: Give me a bank that can loan me $100M without even asking for an ID? Without a collateral, of course. Like Aave Flashloan provides on blockchain.
That seems like a really bad idea!
Ok... how do you repay flash loans when things go bad?
Please, do elaborate.