I outlined it over in another comment[1] so I'm not gonna copy it all over but the point isn't to eliminate all trust. The point of trustless architectures (of which blockchain and smart contracts are one) is that you are eliminating implicit trust.

You are taking all the implicit trust, lowering it into explicit trust assumptions, and formalising who is allowed to make what decisions when, what happens when they do, and how the other parties are permitted to respond.

You are moving all of those implicit assumptions about how a contract, interaction, or relationship work and formalising them into something explicit and upfront so that all participants can evaluate their risk tolerance and trust levels prior to agreeing to a given contract or interaction.

And of course you are also sprinkling in a heavy dose of automation to smooth out the complexities of these explicit, mechanised contracts such that the happy paths are buttery smooth and the unhappy paths are at the least bearable and correspond to the contract you signed on to at the beginning of your interaction.

TLDR: It's low trust automation + formalising implicit assumptions into explicit ones.

1. https://news.ycombinator.com/item?id=46181371#46192445

▲

Kbelicius a day ago | parent [-]

Clicked the link but ctrl+f doesn't find any posts by you.

> The point of trustless architectures (of which blockchain and smart contracts are one) is that you are eliminating implicit trust.

That is also the point of laws and contracts as we have them today. How does, explicitly, blockchain improve on that?

> You are moving all of those implicit assumptions about how a contract, interaction, or relationship work and formalising them into something explicit and upfront so that all participants can evaluate their risk tolerance and trust levels prior to agreeing to a given contract or interaction.

What implicit assumptions aren't removed by laws and contracts as we have them today that are removed by blockchain and smart contracts?

> And of course you are also sprinkling in a heavy dose of automation to smooth out the complexities of these explicit, mechanised contracts such that the happy paths are buttery smooth and the unhappy paths are at the least bearable and correspond to the contract you signed on to at the beginning of your interaction.

Without any examples of what is being automated, how and what it is that is made buttery smooth... you really aren't saying anything here. Can you expound on any of those claims?

TLDR: By what you said the only thing that blockchains and smart contracts bring is a new medium to write contracts on.

	▲	OneDeuxTriSeiGo 21 hours ago \| parent [-]
		Ah sorry. I tried to link it in the context. The exact reply is here: https://news.ycombinator.com/item?id=46192445 -------- > That is also the point of laws and contracts as we have them today. How does, explicitly, blockchain improve on that? It's essentially automated tooling. The happy path (i.e. buyer and seller are in agreement) "just works" but when there's a disagreement you can rely on the contract to walk through all of the conflict resolution paths with whatever level of complexity the contract builds in for consensus from multiple third parties, etc. i.e. It's tooling that replaces manual bureaucratic arbitration with state machines and consensus algorithms. For two party smart contracts this means there's no third party but there's an inherent risk of exploitation by one party or the other by the design of the contract. It's inherent to two party contracts relying on any physical exchange but if you trust the party the contract is weighted in favor of, it cuts out any opportunity for arbitration and the complexity that comes with that. Now the only trust assumption is the two parties trust in each other. For contracts with some arbitration process however things get more complicated. Who all is involved in arbitration. Who does the buyer trust. Who does the seller trust. What's the reputation of one of these arbiters? This reputation can be loosely represented as a set of markets for the arbiter with demand from sellers and demand from buyers. If those two markets are out of sync from each other that suggests an impartial arbiter and both parties can reason about that. > What implicit assumptions aren't removed by laws and contracts as we have them today that are removed by blockchain and smart contracts? Well. Part of it is that laws are an inherently fuzzy thing and how they are upheld is entirely dependent on a long running and constantly evolving chain of interpretations from past court decisions. And of course how they are upheld in a specific case comes down to how well lawyers are able to convince a judge or a collection of jurors who were more or less selected at random with anyone semi-literate about the law thrown out ahead of time. So it boils down to "who is best able to sway the opinions of this random collection of people who are as illiterate about the law as the lawyers could manage to get them". Which mostly just boils down to feelings. Of course contracts often go to arbitration instead of to court proper so it's a different case there but arbiters are single authorities that almost universally side with the bigger entity (i.e. whoever is paying them to handle arbitration). So unless you are two large orgs, arbitration is inherently biased. So an alternative is a largely automated system where multiple third parties who are selected ahead of time by the buyer and seller can be relied upon for arbitration and where their decision is for all intents and purpose final. The buyer and the seller have equal decision making power in the selection of these third parties and they can evaluate the reputations of these third parties prior to entering the contract. i.e. you are moving away from trust in a large system with a thousand moving parts all performed by infallible people swayed by emotions and an endless process of appeals OR a single arbiter almost always paid by the larger party who will always rule in their favor. Instead putting your trust into a strict set of automated rules with a formal analysis of outcomes backing it + some optional assortment of selected third parties + a consensus mechanism for those third parties. > TLDR: By what you said the only thing that blockchains and smart contracts bring is a new medium to write contracts on. Yes. It is exactly that. A new medium to write contracts on. Manual bureaucratic systems and thousands upon thousands of people working in a complex legal system are replaced by a machine. Humans are still in the loop of course but only for making specific decisions at specific times in the process. And at the time of agreeing to the contract the relevant parties can ideally rely on tooling to explicitly outline at what points each party is taking on a degree of risk, the likelihood of that risk, and the process for moving forward in those cases. An extremely reductive TLDR is that the goal is to take a system that relies on an army of lawyers and legal analysts and reduce it down into something digestible and navigable by a single lawyer (or even a well educated layperson) with all the existing complexity abstracted away by formal methods tooling.