From: Anders Sandberg (asa@nada.kth.se)
Date: Tue Nov 06 2001 - 07:48:15 MST
On Mon, Nov 05, 2001 at 03:03:11PM -0500, Smigrodzki, Rafal wrote:
>
> ### I would tend to differentiate between actions conducive to survival but
> not dependent on having a sense of self (as in a nematode running away from
> a gradient of a chemical poison) and actions intended to enhance survival
> (as in the eradication of the smallpox virus). The former are not evidence
> of a wish to live, merely the evidence of instinctual survival mechanisms,
> with the animal not aware that the action prolongs survival - the only
> sensation that the nematode feels is a changed activity of poison sensors,
> not a feeling of having escaped death. In the latter example, coordinated
> action is undertaken, with a clear goal of prolonging survival, with a wish
> to live, rather than just evade pain.
But what about the cat trying to escape being locked in the same room as
an angry dog? It will try different escapes and escape strategies. The
same goes for the rat in the corresponding situation, although with
perhaps less behavioral flexibility.
> Does this mean that you are allowed to shoot sociopaths whenever you feel
> like it? It does advance the life-wish of innocent humans.
>
> #### If there was a highly reliable method of diagnosing sociopathy (for the
> sake of this discussion let's define it as a stable lack of concern for the
> survival wishes of other people, despite being fully aware of them) then it
> would be perfectly reasonable to pre-emptively control such persons,
> especially by preventing them from ever being in a position of force without
> direct supervision.
OK, this is a reasonable position I can agree with. An increased risk
is handled by adding a commensurate safeguard. But I don't see *why*
you soften your initial position here?
> >However, the commision of murder(=killing of an innocent
> >self-aware entity against its wishes) or even a verifiable willingness to
> >commit murder, is reason enough to lose the right to live
>
> If you were an anti-abortionist this would be a great argument for killing
> abortion doctors and pro-abortion people, wouldn't it? The biggest danger of
> your argument is that it can be interpreted so broadly.
>
> #### Anti-abortionnists do not define murder as I do - they have
> epistemological difficulties leading them to erroneously conclude that
> embryos have human rights. So, their ethics might be unimpeachable (they
> want to kill those who initiate violence against defenseless children) but
> their knowledge of the world is poor, therefore they make wrong decisions. I
> don't think that my basic argument can be abused any more than other
> principles. After all, "striking back" can also be very broadly interpreted.
The problem here is that it might not just be epistemological
differences, but simply a different definition of human.
I once met a girl who greeted me with "Hi, murderer". She had heard of
my interest in uploading, and considered it murder since at the end of
the process you end up with a disassembled (dead) human and a non-human
simulacrum. A consistent position I disagree with, but understand. But
if she subscribed to your view of when it is allowed to initiate force,
she would not just be right in killing me when I tried to upload
someone, but also in killing me *now* because I would gladly upload a
willing test subject if it was possible - and hence show my disregard of
(deluded) human life from her position.
> ----
> This goes far beyond merely being allowed to strike back at somebody
> initiating force against you, since it also allows you to make pre-emptive
> strikes against people judged (by whom?) as *potentially* able to initiate
> force.
>
> There is a terrible danger in making rights contingent on mental states that
> cannot be checked, rather than based on actual behavior.
>
> ### You are entirely correct - however, the problem you are pointing to is
> not an ethical issue per se - it is much more of a problem with having a
> limited knowledge of mental states. Presently we have only very unreliable
> tools for predicting somebody's likelihood for sociopathic behavior but in
> the future you might be able to predict with a great degree of certainty
> that the person sitting next to you in the cafeteria would immediately kill
> you if you were defenseless, had no social structure to protect you and he
> could take your breakfast. The least you would need to do would be to
> prevent ever being at the mercy of such a person (never let him join the
> police, run for office).
Exactly. And it might be reasonable to try to get the person treated or
supervised. But if you attack him without him attacking you first, he
would have a moral right of defend himself regardless of how twisted he
was.
The problem is that some information is subjective to a degree that no
amount of brain scan can give it to us. If we have different
philosophical views of what it means to be a human, then we might want
to kill each other as being clear menaces to human life but a judge
could not do anything other than rule on what definition of humanity the
legal system accepts - none of us can ever be shown wrong.
> Beside the obvious
> slippery slope of making other, less central rights (such as freedom and
> property) also contingent on having the right thoughts and beliefs *as
> judged
> by others*, revoking the right to live now depends on a judgement of mental
> capability which is quite subjective, easily affected by prejudices and
> self-interest and in general hard to control.
>
> ### These are important arguments, especially the self-interest one, but
> again, they are valid only if our methods of predicting human behavior are
> subjective - if you can run a virtual reality test on a person (with the
> person being unaware of the simulation), and he repeatedly shows objectively
> a disregard for the life of others (will almost always kill an elderly
> couple in an expensive car, stranded on a highway at night, just to get a
> ride in the car) then the answers are much less subjective. His beliefs are
> not so important as is his willingness to kill innocent persons, which
> revokes his own right to live. If you kill in VR, you might kill in RR, so
> it's OK to kill you for real.
Aha, so all the players of Quake and Doom should be executed? OK, they
knew it was a simulation. But what about the killers who actually think
*this* world is a simulation? They are wrong, but ethically it is not
obvious they are different from the Quake players.
Detecting potential killers this way is a great idea - but I think
killing people for a virtual action (even if they thought it was real)
is not just excessive but immoral. No crime has been comitted: I view
crimes as bad things actually done to another being with rights, not
just a one-sided action - it is an interaction between two ethical
subjects. When you punish (severely) non-crimes, the whole scale of
justice tilts.
The greatest risk is that your right to life is suddenly entirely in the
hands of the government (or whoever runs the testing). What happens when
a mistake is done, or some malicious person(s) forge data? You can never
appeal if you are dead.
A deeper problem is that when you start to take away the most valuable thing
there is - life - then there is less resistance to take away other
valuable things. If people can't make informed political decisions, why
allow them to vote? People who can't manage their economy, shouldn't
more competent people take their money and use it wisely on their
behalf? If you think so, maybe you should move to us in Sweden ;-)
> A society where the
> right to life is not universal and can be revoked not just for actions but
> because your *thinking* violates certain standards sounds like an extremely
> dangerous situation.
>
> #### By the way, correct me if I am wrong, you agree that the right to live
> might be revoked for actions, right?
Only for initiation of force, and even then only to a minimal extent. I
do not, for example, believe that executions are ethically acceptable.
-- ----------------------------------------------------------------------- Anders Sandberg Towards Ascension! asa@nada.kth.se http://www.nada.kth.se/~asa/ GCS/M/S/O d++ -p+ c++++ !l u+ e++ m++ s+/+ n--- h+/* f+ g+ w++ t+ r+ !y
This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 08:11:50 MST