Re: Hello directly from Jimbo at Wikipedia

[Marc Abel <m-abel@xxxxxxxxxxxxxxx>]

Hi Jimbo,

My concern about pseudonyms is that they can compromised after-the-fact.

Suppose Li Li is over in China writing blogs, and she selects Muzimei as
her pseudonym and implements it with any of several digital signature
schemes.  Then she writes a bunch of blog entries, and possibly a few
edits to her Wikipedia page.  Wikipedia trusts her because of her
reputation, and presumably no one knows her real name is Li Li.

Except there's an eyewitness to one of her blogs, or the authorities
raid her Linux box and her private key is compromised.  Li Li has a
serious problem, because all of her previous activity is signed by her
Muzimei pseudonym, so after 36 hours of interrogation she confesses, is
found guilty of separatism in an 18 minute trial, and executed 20 days
later.

I still lean in favor of approaches which protect Wikipedia on the basis
of actual content submitted, instead of information about the
submitter.  This is why Paul Syverson and I are advancing these types of
proposals.

Once again, one way to check content (until machines can "think" better)
is to have not-necessarily-anonymous parties with interests in specific
subjects ok what would otherwise be anonymous posts prior to the posts
showing up.  This system would be only as efficient as these "approving"
personnel for anonymous posts, but these people can be mutually selected
in the sense that an anonymous editor can nominate anyone, and Wikipedia
may refuse anyone.

It's a little like any meeting held under Robert's Rules of Order.  To
discuss anything, someone must present a motion and someone else needs
to second it.  If the content's not pertinent enough for even a second,
there really no need for the group to consider it.  In the mechanism I
am suggestion (as an intermediate-term approach to try), the mover may
be anonymous if the seconder is willing not to be.

Marc



On Wed, 2005-09-28 at 08:36, Jimmy Wales wrote:
> Marc Abel wrote:
> > This has law enforcement implications; if you can prove that Alice took
> > the cookie from the cookie jar, perhaps using eyewitnesses, you can now
> > show that Alice did many other unreputable things.
> 
> Can you explain this in more detail?
> 
> --Jimbo