Hacker News Reader

Comment by hiq | Hacker News Reader

> Price and volatility aside, if we simply could have used a smart contract where money is automatically deposited upon project completion, we would have ensured payment.

A smart contract doesn't allow you to do that though, so I'm not sure how much you thought this through, or if you're only writing about your initial impression and dismissing what you learned after (I do think you realize this, as most people do after a few weeks of experience with Ethereum).

Either the smart contract would have automatically transferred money after a period of time, in which case it's basically like paying in advance. Or the smart contract would have received some external information, either from the consulting company, the client, a 3rd-party or a combination thereof, to decide whether the project has been completed or not. At that point, you're back to the problem you had without Ethereum.

View on HN

Thanks for asking us to expand! We can always write a novel, but not everyone wants to read about oracles and unsolved problems in distributed compute.

You are correct, for this smart contract to work you need defined rules around how the money is released should either party falsely claim success or failure, so you need to rely on automation or some third party to validate fulfillment of the contract. So you're back to an oracle problem, how do I know I can trust the validator?

The nice advancement here is that anyone can be that third party! Your smart contract can explicitly identify a third party for multisig, or you can rely on a validator pool of mechanical turks instead of a licensed escrow agent, and accordingly the cost of escrow can decrease.

The nice part is that this is customizable, and built into the very rails that move money. It's a much lower barrier to entry than trying to build something like that for ACH, and that's what excites us about cryptocurrency.

It seems to me like there is already something better than "smart contracts" now with the current legal and financial system. The current system can handle all sorts of nuance, edge cases, and disagreement that smart contracts can't. What's the Oracle for "Did you really go bankrupt?" or "Does the interaction between Alabama and California law necessitate payment in these specific conditions?"

Ethereum is one of the more credible cryptocurrencies and yet, it is slow, throughput is woefully inadequate for scale, expensive to transact, and smart contracts don't really work. When I say "Don't really work" I mean that you can't use them for the things you'd want to use them for - i.e. you can't write an Ethereum smart contract for "If I deliver X hours of contracting work, meeting these specifications, then you will pay me Y amount by Z date automatically." If I'm wrong and you can do stuff like that, please point me to the solidity.

I'm glad you included the line welcoming comments from those who feel this market shouldn't exist, otherwise I wouldn't have written this comment. I hate to be negative regarding a launch - but this feels like adding a product to the Ponzi Scheme Industry. Maybe it's like writing software to help manage slot machines, except your slot machines aren't regulated and you tell people they're investing rather than gambling.

Thanks for the thoughtful response, and you're absolutely right, the current legal system is designed to handle all these. Right now, Oracles are also way more complex to build, and they can't handle the nuances of meeting X specification in a way that's simple and easy to do.

To be clear, that doesn't mean that a simple solution will never be created. Obviously we think it will, but that doesn't mean we're right.

But to address a slightly different part of your framing, our situation actually got more complicated when attempting to go the legal route. We suddenly had additional lawyer, collections, and time costs associated with the process...and we still never managed to collect. That's certainly an area where the current legal system fails (for a lot of people!).

Cryptocurrencies might not be the perfect alternative, but they are an alternative that, in our opinion, is exciting to explore.

But it’s not just a matter of “no done yet” is it? Adding an oracle removes any “benefits” of smart contracts and puts you right back where you were with just regular contracts (but this time done badly and with bugs by incompetent bullshitters)

The fact that you justify your product with a use case which does not work and can not work is very telling. But I’m sure as a strategy it’s working very well for you.

I think it will help if I share a little more of the story.

We are not using that use case of smart contracts as justification for our product, but the excitement for the future of smart contracts and the innovations in the cryptocurrency space encouraged the founders to get involved. They started buying cryptocurrency and found a pain point in tracking and rebalancing all of their holdings using Excel which could take upwards of 30 hours a week. Making an easier solution is the justification for the start of this product

Hope it works out!

Think cars near the time they were first commercialized, and cars now. Blockchain and smart contract tech is several times more complex than cars, much probably. Of course you will get frustrated if you expect smart contracts to substitute the entirety of the legal system because it's not going to happen, I can assure you. If you look at cars and dismiss them because "they cannot fly" that's on you. There's a lot smart contract tech is doing, can do, and will do. Not only through tech but also through culture. There's plenty of interesting use already, but I'll mention a different example, more on the culture side of things:

People in Brazil are using a new technology to transfer money between bank accounts faster and in an easier manner, it's a hit and growing very fast. It's not blockchain or ethereum, but a. the team involved cites those techs as their main influence, some of them having prior work/interest in those areas aswell and b. they say they have plans for implementing integration with public cryptocurrencies, smartcontracts, and several of their technologies such as smart contracts. Why, you could ask, wasn't this released to the public before, if it's not groundbreaking tech or even a tech innovation at all and provides so much benefit? "Culture".

I got your point, this is tangential, but I would argue cars are much complex than DLT because they interface with the real world in a mission critical manner. The only reason they don't fly is the safety margin.

It's refreshing to hear well-thought-out responses to the "isn't it just escrow" question, which I admit I've asked in a skeptical way re: cryptocurrencies in the past! Decoupling oracle-ness from a need to be a legally trusted holder of money is indeed something that broadens/democratizes the pool of viable oracles.

On that note, if an outstanding question is "how do we increase the supply (and thus decrease the price) of oracles beyond those who happen to know how to interact with smart contracts," I wonder if something like your Hedgehog tech, paired with standardized smart contract plugins, could make it possible for far more domain experts without crypto experience to enter and advertise themselves in the oracle market. Makes your system a one-stop easy-onboard volatility-reducing shop for the escrow problem you identified, for all participants in that marketplace!

Wow, you hit the nail on the head! To point to our long term vision, cryptocurrency has the nice property of being data, but the double-edged feature of being money.

In order to tell people how to get data from point A to point B you suddenly have to be a fiduciary financial adviser! We want to take that duty seriously and help people interact with the cryptocurrency ecosystem as a trusted adviser.

If we can help you customize and interact with smart contracts and plug them into existing regulatory frameworks, then we can unlock a lot of value for everyone.

I still struggle with smart contracts. I am glad they exist, but right now I think they will be some super rare thing, used in cases that previously required eacrow.

I was in the e-signature world, and nobody tended to ever care about authentication. Nobody really wanted to pay for it, almost ever.

They smart contracts have 2 of the largest features of escrow: handing control of your money to a third party and not being able to use the money now/bad payment "terms". It mitigates some counterparty risk, as it's harder for the oracle themselves to steal the money.

It's also probably expensive, as oracles have to meditate disputes and take on risk.

So interesting, but unsure what it would be used for.

There is also decentralized protocols for judging events like https://kleros.io/ (not affiliated)

While I love the idea of JaaS (Justice As A Service) I suspect that in most cases the cost will make it unworkable. In the situation OP faced, figuring out whether the contract has been correctly executed and whether the payment should have been made would have required specialized knowledge as well as time to look into the details of the case.

Do you have money transmitter licenses in all the states? Do you plan on getting them?

We do not currently have money transmitter licenses as we do not move any money for you (yet), we merely pass your trade requests to your exchange.

To your point, though, we are pursuing these licenses in every jurisdiction where required so we can start offering this service. We're also registering as an investment adviser so we can offer financial advice. This takes time and a lot of capital, but we believe the end result will be worth it.

I'm trying to understand, by passing the trade request you are like a hosted wallet that you control? Instead of the user broadcasting a transaction, the hosted wallet does that? I don't understand how that isn't money transmission, unless the hosted wallets are actually multisg, because as I undertand it multisig isn't considered custodial.

So currently we don't host any wallets. Clients open custodial accounts with pre-existing exchanges and then leave an API key for us to retrieve transaction and balance data. When they request a trade, we simply let the exchange know that the client has requested a trade, and we don't move any money on the client's behalf.

We are currently seeking money transmitter licenses so that we can offer a hosted wallet of our own, where we hold the exchange accounts on the client's behalf, and act as their counterparty for every trade so they don't have to maintain accounts at multiple venues. This version of the application is not yet available.

Thanks for the clarification. The api is used to send instructions to the money transmitter. I am imagining the product is more UX and aggregation of data for the end user. Congrats on your launch.

> A smart contract doesn't allow you to do that though

This is why smart contract is such a bad name. A lot of people have a complete misunderstanding of what a smart contract is. They think it can enforce things like a real-world contract, but "automatically". It's just a decentralized script and without real-world input through an oracle it is of course impossible for the contract to know whether a project is completed. There's nothing smart or automatic about that part.

You hit the nail on the head. Unless it’s totally virtual - on chain finance, gambling, etc. the contracts are only as good as the oracles. I guess that’s why DeFi works so well.

Fortunately decentralized oracles are an area of focus right now, but it’ll be a slow process since most expansion will require a special integration and potentially expert oracles.

In fact, Ian Grigg's Ricardian Contract construct was originally created with centralized enforcement in mind. To him, the purpose of these contracts was to create an immutable record of the intent of any transaction that could then be followed all the way back to its inception in the event of a dispute.

So part of the magic of the smart contract is illustrating intent from the beginning, and making that intent customizable! Oracles are still an unsolved problem, but we absolutely believe that reducing the barrier to entry for becoming or acting as an oracle will result in cost savings as the ecosystem matures.

One way to do this is that the agency places a deposit in the contract, client places deposit+payment in the contract. Then when the job is done both parties confirm and the agency receives deposit+payment and the client receives their deposit. This way both parties are incentivized to find a mutually satisfying outcome.

It's not ideal in terms of funds used, but it improves the trust situation a little bit.

The deposit you mention if you do this on Ethereum just corresponds to legal fees you have to pay in case of conflict in the traditional contractual approach.

This incentive already exists without Ethereum, and Ethereum doesn't improve the situation in that regard.

It seems like what they want here is an advance + escrow. If you're working on, say a website with xyz features, how is a smart contract going to evaluate whether you met or didn't meet the criteria?

Good question, we've answered a couple times in this thread already, so please forgive the copypasta:

> So you're back to an oracle problem, how do I know I can trust the validator?

With DLT it seems trust and predictability is virtually the same thing. "Everybody loses everything" is predictable but not efficient. DLT is a proposal for efficiency.

We need to automatically negotiate contracts or we shift all the cost of fixing it later in courts up front. There will have to be a lot of smart defaults for the contracts to be smart, and very deep integration of tech so that the algorithms can be fed with enough data to negotiate in a very high-dimensional space. That requires infrastructure, but the DeFi culture does not seem to be interested in funding infrastructure.

Reply on news.ycombinator.com