From: Peter C. McCluskey (pcm@rahul.net)
Date: Thu Dec 17 1998 - 16:57:20 MST
mark@unicorn.com (mark@unicorn.com) writes:
>Peter C. McCluskey [pcm@rahul.net] wrote:
>>You
>>probably assume that you can evade governments by using encrypted
>>communications, but I doubt that you will be able to reliably keep
>>the government from observing your communications after the person
>>you sent it to decrypts it (unless you are very carefull about whom you
>>communicate with), and identifying you by analyzing your writing style
>>if needed.
>
>What does it matter if they can only link me to 'foobar@nym.net'; are
That would work if you never communicate in ways that the government
can trace. Most people either want to continue using meatspace
communications or doubt their ability to consistently avoid mistakes
which will allow the government to trace some of their messages back
to them.
>they going to arrest me by email? And evading writing style analysis is
>easy, just feed your message into a program which converts it into a
>standardized style, or modifies it to match someone else's.
It's easy to attempt, but it's hard to verify that such software works.
The more you are willing to accept some alterations of your social personality,
the more believable this approach sounds, but then it becomes harder for
most of us to believe it's worth the cost.
Take for example you sell some software that you write under an anonymous
contract, and don't trust the buyer.
Will the program you mention just change some variable names and
whitespace? ("Merely" interfering with the company's ability to maintain
it by obscuring your design, unless maybe the software is close to passing
the Turing test.) It isn't obvious that that will be enough to remove all
patterns. And if it tries to do more, I have doubts about it's ability to
reliably avoid introducing bugs.
>>It it's universal, you can find all the information you need, but you
>>have no obvious way to alter the information without someone seeing you
>>perform the alteration.
>
>And how will they prove that the video showing me modifying the data is
>real and not another fake? If anyone can fake video then *nothing* can be
>trusted.
With several independantly controlled videos showing you modifying the data,
the only thing we need to trust is that there isn't a big conspiracy to
frame you, i.e. it creates no new risks of injustice as long as people
accurately estimate the reliability of the evidence.
Also, it's a lot easier to keep the truth consistent than it is to create
a consistent set of fakes.
>>With the majority of the people substantially consenting to the takeover.
>
>What the hell do you expect them to do about it? You can't vote for a
Vote Libertarian?
>non-crook when there are only crooks to vote for, and you can't start
>shooting them until a large fraction of the population have accepted
>that shooting them is the only solution.
Considering the weapons they have, it's hard to imagine that shooting
them is a solution.
>>By the standards you are using here, almost all evidence is a joke. Are
>>you arguing against using evidence, or did you have some other point in
>>mind?
>
>Video evidence is a joke. Anyone who trusts it is an idiot. No non-sentient
>witness can be trusted.
Anyone who has religious-style faith in it is an idiot. So is anyone
who religiously rejects it.
mark@unicorn.com (mark@unicorn.com) writes:
>den Otter [neosapient@geocities.com] wrote:
>>ROTFL. Right, of course a sentient witness *can* be trusted.
>
>Uh-huh... and when exactly did I say that they could? No, I didn't say
>that, did I, actually? I said that non-sentient witnesses couldn't be
>trusted. Yet people like you would throw out a hundred witness statements
>if you had a faked video which showed a fake crime.
Setting up this straw man hardly justifies your attempt to argue that
video evidence is undesirable.
>In the last year or so, I've had my photo taken by one red-light
>camera while driving through a green light, and two speed-cameras while
>driving at 30mph in a 30mph limit. Living in an authoritarian surveillance
>state I have a very different opinion on the infallibility of automated
>systems; luckily either those cameras were out of film or some sentient
>worked out that they were fucked up, otherwise I'd have lost my driving
>license without commiting any crime.
Thses sound like good examples of why multiple cameras are better than
one.
-- ------------------------------------------------------------------------ Peter McCluskey | Critmail (http://crit.org/critmail.html): http://www.rahul.net/pcm | Accept nothing less to archive your mailing list
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 14:50:03 MST