Hacker News Reader

Imagine you are going to hire something expensive, and the owner wants a rental bond. So you go to a 3rd party, they set up a smart contract which automatically returns the bond to you 2 weeks after the lease ends, but the 3rd party can redirect the money to the owner by manually intervening.

The example is basic, but now the 3rd party can't take the money, go broke, take risks using your bond, etc, etc, etc. There is clear-in-code records of what the powers of the parties are and when things happen. You can prove where the money ended up.

This isn't an earth-shattering example, and things can still go wrong. But the failure cases have shrunk vs using paper. It is an improvement on the status quo.

I've done your example in the real world, without smart contracts. The 3rd party is a notary. If I understood you right the notary here is replaced by a notary plus smart contracts.

A notary in my juristication (Germany) is a quasi-employee of the government, so has little risk to go broke etc. The particular bonds for these cases are usually non-transferrable.

OTOH smart contracts introduce new risks like bugs and costs like high gas prices - for what gain?

Edit: To clarify, the modes of failure are vastly worse for the smart contract scenario. If the notary goes broke, my money will not be part of the insolvency estate. If OTOH the smart contract has a bug, everone's money is gone.

> A notary in my juristication (Germany) is a quasi-employee of the government,

In a sense; precisely. At the moment, to fill that function, you need a quasi-government employee to have a trustworthy bondholder.

With a smart contract, literally I could be the notary and while that would be super weird, I can't actually do anything not specified in the contract.

Something is possible that wasn't possible before. That is why people can get excited about the idea.

yet you can do something not matching the real world outcome. and there's no way to roll that back. and people still need real-world trust in you... seems not soo smart to me.

I don't know about Germany, but in Belgium you pay a crazy amount of money for the notary service.

So it only makes sense to hire them for when dealing with large amounts of money. So what do you do with small amounts of money?

Plus, not only notaries but also escrow services might be replaced with smart contracts. Smart contracts might be way cheaper.

It will be very interesting to see this play out. But I wouldn't just discard smart contracts because we have some ancient way of doing things right now. And let's be honest, every time I have to deal with a notary, the work looks like it comes from Medieval times.

But the smart contract solution still needs a notary? It just can now be an unregulated third party. What problem has been solved here?

Speed and money

In Switzerland that's the default for deposits: you create a bank account in your name, but you're not allowed to touch this money until you've left the accommodation and the landlord has agreed that you left it in good shape. I've heard that it's also rather hard for the landlord to claim this money, they need a strong case.

all those cases are examples of how old institutions have managed to solve issues relevant to the smartcontract case. But most people don't have access to such institutions, and it turns out building political institutions capable of providing such tools and trustworthy is really hard. Even worse, some people don't have as much trust in longterm longevity of governments.

for all those people, trust-less smart contracts is a replacement of institutions they don't have, available as long as its tech works.

I still don't understand what happens when you actually break something expensive that you hire?

Doesn't some 3rd part have to verify the damage? Doesn't the 2nd part then have to still take a part of the bond?

What if the 2nd part just says you broke something, when you didn't, who solves that?

This is an exact example of what I'm talking about, where, in your example, I actually see 0 utility, really negative utility, in using a smart contract.

At the end of the day, this third party escrow service is still the only one that needs to evaluate the performance of services rendered - so where did your "trustless" distributed network go?

I get that the smart contract can ensure that there are only 2 outcomes (money sent or money returned), but it's not like you can remove the legal system here - there is nothing to stop the 3rd party from colluding with one of the other parties to make the decision their way. Heck, even is you still wanted to use crypto, why not just get rid of the smart contract altogether and send the money to the 3rd party's escrow account? You still have a full audit trail so you're guaranteed to know is the 3rd party sent the money somewhere else.

The Automated Market Maker (AMM) systems are smart contracts using something more equivalent to a ballast system with pools - "liquidity pools" (LP, ironically and perfectly these act like limited partner shares of a pooled investment) - allowing the current price to be more of a ratio, which is stored in the smart contract after every trade. The largest AMM's have user-spawned markets which contain two assets (this is in direct contrast and competition to the concept of begging exchanges to add your asset and pretending that market makers organically showed up, all while getting extorted by the exchanges and market makers, the market has chosen AMMs). When someone buys in an AMM market, they send AssetA adding to the pooled balance of AssetA, and receiving AssetB, reducing the pooled balance of AssetB. The ratio (simple division) of these two assets is stored at the end of that trade in the same smart contract (the spawned liquidity pool ballast), and can therefore be read by any other user and other smart contracts. Reading is a free operation.

This user-spawned and self-replication smart contract factory, alongside the price storage, has been key to the DeFi boom over the past year. No oracle required.

People (organizations) are using oracles to try to keep markets with similar assets in sync, and for smarter routing for the best price. The arbitragers take care of that too, which is simply Oracle + Fund movement. Flash loan users in the repo market are arbitraging as well, which is simply Oracle + Borrow + Fund movement + Repay in a single transaction included in a single block.

It is a growth industry when you stop asking why and start asking how. A lot of oracles are ignoring some market pairs, and the oracles themselves could have their own inefficiency.

>real-world contracts require some information about the real world

For a blockchain to know about the real world, it does indeed need so-called "oracles".

These are typically trusted external entities (e.g. an actuarial firm) who "inject" signed information into the blockchain that can then be consumed as input by smart contracts "living" in the blockchain.

For example, an Oracle could inject, say every day, signed with their published key pair, the average price of Bitcoin into the Ethereum blockchain, and you could image a family of smart contract paying out if and only if the oracle-published Bitcoin price meets certain conditions.

This solution is - of course - very far from ideal as the information is only as good as the trust you place in these external institutions.

On the other hand, this is not very different than the trust you place in well established institutions such as banks, rating firms, escrows, etc...

What benefit this brings is:

    1. Transparency (anyone can check the chain to see what contracts are in place)
    2. Automation
    3. Complex hierarchical structures where the "building blocks" or "leaf cells" is provided by the oracles.
    4. A single failure of the Oracle to inject "truth" into the chain would ruin their rep. for ever.

To compare this to a real world contract, one party that has - say - written a "promise to pay" if certain conditions occur could chose to reneg on that promise, which would lead to lengthy and costly legal proceedings.

This latter scenario can't occur on a blockchain: the cheater would have to bribe the Oracle, a far more complicated affair.

[EDIT]: and as pointed out by another comment, there are federated oracles: a smart contract can use an aggregate of multiple sources that provide the same info and form a consensus based on those.

[EDIT2]: another interesting point is that as more and more financial activity "moves" to the blockchain, the need for Oracles diminishes as the information required by a smart contract becomes native. You can thus view oracles as a temporary onramp of legacy financial info into the blockchain.

Federated oracles is the answer you are seeking. Essentially build systems to quantify reputation and provide rewards for those that publish data of interest in a manner that can be driven via consensus or through interfacing with an agreed upon central authority prior to a contract. You don’t “trust” the oracles you make them compete and avoid malicious actions.

You are focused on off-chain interactions for smart contracts. The infrastructure is not there yet I’m afraid, but there is steady progress. I believe Taiwan is working on IoT infastructure for publishing high resolution air quality data by having individuals set up sensing nodes that gather the data and publish it. Not sure if there is any DLT involvement, but no reason there couldn’t be.

On chain contracts have lots of use cases in the meantime.

> On chain contracts have lots of use cases in the meantime.

Thanks for responding. Can you suggest a starting point for learning more about what these use cases are, because I'm still struggling to understand what those would be. When I try searching for things like "uses for smart contracts" I'm pointed to things like this page, https://corporatefinanceinstitute.com/resources/knowledge/de..., which honestly is a bunch of pie-in-the-sky bullshit that is not actually technically feasible at present.

I agree. I've been holding BTC for a couple of years as I feel that international accessibility coupled with scarcity makes for a great store of value. But DeFi I still fail to wrap my head around.

My main concern being: DeFi could get harsh competition from traditional financial actors by them getting up to date on accessibility. Thus appealing to a new, more decentralized userbase.

Why do people think cyberspace is not the real world? I think smart-contracts as envisioned by early cypherpunks where meant to run in a world where Governments are obsolete, they won't work well in our current society.

This is the kind of "cyber-babble" I'm talking about. Smart contracts can only respond to things that happen on the chain, but the vast majority of events that humans produce contracts to govern in the first place happen in what is colloquially termed "the real world". Yes, I think the blockchain is "real", but I think most people understand what I'm talking about when saying I want a contract to be able to reference actual performance of a service rendered in 3D space.

> I think smart-contracts as envisioned by early cypherpunks where meant to run in a world where Governments are obsolete

"Governments" have existed in some form since the beginning of the species (families and tribal structures, etc.) and pretending that some piece of technology may make them "obsolete" belies a very simple understanding of human nature.

I'm really glad you asked this question, because I've wondered it as well.