[Cryptography] Anonymous rendezvous (was Business opportunities in crypto)

Jerry Leichter leichter at lrw.com
Sun May 2 13:06:49 EDT 2021


>> Introducing a bunch of fancy words doesn't change anything.
>> No, it still makes no sense. If you don't know who you're
>> talking to,
> 
> 
> But you *do* know who you are talking to.  What you don't
> know is where he is, what he looks like, or how to send thugs
> around to his place for speaking crimethoughts or to seize
> his assets.
> 
> You do know you are talking to someone with a reputation for
> speaking interesting and important truths on a strongly
> pseudonymous public forum, or talking to a business with a
> reputation for doing what it is paid to do, even though the
> taxman and the regulators may have difficulty finding it.
Did you read the original message you're replying to?

"[The participants] (a) are "strongly distributed" - "strongly" because we want to assume no connections between participants except those explicitly desired; (b) no trusted third parties"

What exactly is "reputation" if not (a) an implicit communication through intermediaries to deliver information about the party involved; (b) trusted third parties:  You are trusting their evaluations.  Zero for two.

And, frankly, I find it remarkable that anyone these days believes that anonymous - or pseudonymous - reputations in a large-scale open community solve *anything.*  Go look at product reviews at any on-line site.  Or at follower counts on social media sites.  How many of them are being faked by people paid to comment - positively or negatively?  Reputations worked as long they weren't faced with an active, intelligent adversary who set out to counter them.  Kind of like the typical cryptosystems produced by amateurs:  Really strong against the particular attacks the designers thought about; trivially attackable by a serious opponent.

Attacks on reputation systems are the mail spam of today.  And, sure, I know you'll propose some clever mechanism that stops them in their tracks.  Just as we have years and years of clever mechanisms that stop spam in its tracks.  And you know what?  None of them work.

Mail works great in controlled environments - controlled either because it's a relatively small group involved and the connection path lengths between people are short, or where there is some authority that controls who participates.  As it grew beyond that to the open Internet, we got spam.  The same goes for reputation systems.  They're great at toy scale; they might work with tight controls.  The latter is what some sites are trying - e.g., only allowing reviews from people who actually bought the product.  But that introduces other issues:  If it's the site itself that controls comments, they have an incentive to block negative ones.

                                                        -- Jerry



More information about the cryptography mailing list