From: Tennessee Leeuwenburg (tennessee@tennessee.id.au)
Date: Tue May 16 2006 - 16:42:55 MDT
m.l.vere@durham.ac.uk wrote:
>> If morality is purely relative, then by definition we cannot instill an AGI
>> with our own morality (otherwise it would, from the perspective of the AGI
>> at least, be objective -- a given)
>>
>
> I dont think this is so. Whilst IMO, the balance of evidence is overwhelmingly
> against the existence of an objective morality, loads of people believe their
> relative moralities to be objective. I am certain it is theoretically possible
> to give an AI any set of goals we like, and have the AI follow them as if they
> were an objective morality.
>
That's self-contradictory and confused.
(1) So what if people believe their relative morality to be objective?
That people are stupid, confused or mistaught is not actually helpful to
deciding the issue. Their moralities remain in fact relative, and what
they believe about their objectivity is irrelevant.
(2) If we can make an AI follow imposed moral goals as if they were
objective, then morality is not by definition relative. We will have
just disproved that with an example. Instead, morality may at least
sometimes be objective, and we may proceed to deciding whether we should
impose our morality on an AGI, and in what way.
>
>> As you can see, this problem boils down to two things:
>> (a) Debates about morality
>> (b) Understanding AGI
>>
>> I suggest that (a) is not SL4 or easily solvable, and that (b) is what we
>> should concern ourself with?
>>
>
> So, we work on how to build an AGI, and how to make it follow a 'morality'.
> Then we let someone else decide what that morality should be. Sounds clever.
>
>
Hmmm, irony. I think you should try using this less, and making more of
an effort to understand what I'm on about. The question about how to
make an AGI follow a morality is the problem which must be solved first.
I think it's clear that any debate about AGI morality needs to be
informed by the nature of that AGI and the nature of morality. There is
no shortage of philosophical discussion about the nature of morality as
such. In order to contribute anything new to the philosophical debate,
we must introduce new concepts, such as how an AGI morality may be
constituted, before we may be regarded as saying anything useful.
I don't think any of the debate so far has done so.
Cheers,
-T
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:56 MDT