Hacker News Reader

Comment by wyager | Hacker News Reader

It's bewildering that so many people fail to get the "minimizing trust" aspect of Bitcoin; I do not trust anyone to successfully generate and hide (in space) super-secret private keys that might be worth trillions of dollars. Not to mention all the other obvious problems with this silly idea (What happens when the satellites stop working after 15 years because they got too far from the sun, or radiation fried their silicon?)

View on HN

I think people are so accustomed to having to trust each other that the notion of something being "trustless" is really foreign.

If you tried to live a trustless life, you couldn't get surgery, ride on a plane, get food from a grocery store, etc. - all of those things rely on implicit trust in other human beings. The whole of human society was built on trust!

Trustless people are in for a hard life. Trustless protocols are in for an easy one -- harder to game and verifiable as authentic.

Trustless protocols are still trusting the hardware, firmware, operating systems, compilers, software, maintainers…

All of that is at least auditable. You know what's not auditable or algorithmic? Literally everything else financial.

Oh great, then how about you implicitly trust me with your bank login info and email it to me?

> you couldn't get surgery, ride on a plane, get food from a grocery store,

I look up my doctors' ratings before I use them. I don't fly sketchy third-world airlines with bad safety records. I don't buy food from grocery stores in places with bad food safety. "Trusted" and "trustless" are shorthand for the degree of system-external due diligence you need to do before having a sufficient degree of confidence.

But you're trusting the ratings and the safety records.

This is silly. When looking at the ratings for your doctor, do you look up the rating of the raters?

The point isn't that you have to trust everything. The point is that trust underpins everything. At some level, you are trusting something.

> do you look up the rating of the raters?

Obviously. Most rating sites are scams. There might not be a formal meta-rather, but I can use eg reddit.

In that, you are still trusting that the system you have devised had not been successfully gamed or hacked by an adversary. For that matter, you may just be trusting that you don't have an adversary.

The rabbit hole runs stupid deep.

You've just effectively demonstrated how trust functions in society, not its absence.

But you do trust mining groups not to get to 51% capacity, which has happened before. You do trust that ECDSA/secp256k1 and double SHA-256 don't have unknown shortcuts, even though that hasn't been proven.¹ You do trust the Bitcoin Core developers, and the developers of your wallet program, not to add a backdoor, and you do trust your own ability to keep your private keys confidential. In minimising trust, you're also centralising a single chain of it; the real world is resilient against misplaced trust (so long as many entities aren't secretly untrustworthy at once, which we call a conspiracy), but Bitcoin isn't. It's a small number of single points of failure.

I'm glad you said “reduced trust”, rather than “trustless”, because otherwise this would've been a several-page rant.

¹: It's unlikely either of them will have significant shortcuts discovered any time soon, but that's not the point.

Reply on news.ycombinator.com