RE: Copying nonsense (was Re: [sl4] Uploading (was : goals of AI))

From: Bradley Thomas (
Date: Thu Dec 10 2009 - 14:31:39 MST

>There is no exact definition.

I think there needs to be, otherwise how do we know what we're talking about
when we refer to experiences as being "subjectively connected"?

Brad Thomas
Twitter @bradleymthomas, @instansa

-----Original Message-----
From: [] On Behalf Of Stathis
Sent: Sunday, December 06, 2009 10:29 PM
Subject: Re: Copying nonsense (was Re: [sl4] Uploading (was : goals of AI))

2009/12/7 Bradley Thomas <>:
>>If two experiences S1 and S2 are generated by the one computation but
> no content (because the program deletes data from the memory of the
> subject and inserts unrelated data in its place) then there is no
> subjective connection between them. On the other hand, if the two
> experiences are generated by unrelated computations but just happen to
> share content, so that S2 remembers being S1, then they are
> subjectively connected.
> This is interesting. What kind of content overlap does it take for two
> data sets to be "subjectively connected"? Is there a clear definition?

There is no exact definition. It is possible to lose large chunks of your
memory and personality as a result of brain injury and still remain you,
i.e. feel that you are subjectively connected to previous versions of
yourself. But if the injury is bad enough that you remember nothing of the
past and people who know you don't recognise you apart from your physical
appearance, then you may as well be dead. Things would be more complex if
memories, personalities and the sense of identity itself could be altered
easily and discontinuously. On the other hand gradual alteration might be
OK, even if the end result is radically different, since that is what
happens as a person grows up.

Stathis Papaioannou

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:01:05 MDT