[ExI] Strong AI Hypothesis: logically flawed

Dan danust2012 at gmail.com
Wed Oct 1 21:45:20 UTC 2014


> On Tuesday, September 30, 2014 5:57 PM, Stathis Papaioannou <stathisp at gmail.com> wrote:
> The other copies are not the you, now, in this world, but
> your past self is not the you, now in this world either.
> The copy of you that wakes up from cryonic sleep in a
> minority of worlds bears the same relationship to you,
> now as you, now bear to your past self.

This relies on a certain view of persistence, which in itself is not uncontreversial. I'm sure you've heard of endurance, exdurance, and perdurance (or whatever other theory might be cooked up here) with regard to persistence. It's unclear to me which is the correct view to take here. And, again, I don't postulating one view is correct simply to make the solution you want is the way to go here. (The same goes for personal identity beyond temporal issues. The idea that self is an illusion is not settled and not without serious (in my mind) problems.) 

> If you're blind or disfigured that is a real deficit and makes a real
> difference. If you die every night in your sleep and a different
> person wakes up in your bed with your memories every morning that is indistinguishable from ordinary life. In fact, it is arguable that that is what actually happens.

It's true that the deficit is one you know about via memory: you recall that previous you didn't have said deficit and now you have it. (I was going to quip: But what if you really were already blind or disfigured and it's just your memory (and all other testimony) of not being so that's wrong here?) But my point is knowing you die every night is something that the new you would get used to -- or maybe not, but what could you do about it?

But this doesn't clear up the matter of whether this is actually is the case or whether it clears up anything about possible worlds or a multiverse. That you couldn't tell any difference from the inside or out in the case of dying every night doesn't mean that this applies to supposed yous in other possible worlds or other universes. You're still trading on the idea that those other yous are somehow you rather than instances of you in another universe or possible world. And internally or externally, in many of these cases, they'd be distinguishable. The you that is here now exchanging emails with me is distinguishable from the you that's in another possible world who's never exchanged emails with me, for instance.

You could argue there are possible worlds or universes so close that you couldn't tell. Maybe, but it still doesn't answer whether they're the same you. Also, there are ones that are not so close where they are distinguishable, outside and in. Again, the problem is whether and how the yous (said in strong Cockney accent;) in different possible worlds or different universes are part of an overall you -- presuming either of these exist. That's not clear here. All of this seems like postulating a just so story on possible worlds or a multiverse to get the optimistic answer you want. (And we're not even getting into all the other possible worlds or universes where you life is far worse or nonexistent. Why just focus on the ones where you exist in a better state?)

>> Also, these are epistemic issues that don't really clear up what
>> is the case. You might not know (or now know, considering that
>> the problem might be tackled in the future) how to resolve these
>> issues, but lacking a resolution doesn't erase the problem. Nor
>> does merely adopting a resolution that seems
>> uber-optimistic: no one really dies or needs to worry.
>> 
>> This also doesn't really resolve the issue of whether strong AI
>> is possible. I doubt it, but one conceive of it being the case
>> that they are necessarily ruled out (in our world, or, if
>> you please, in all possible worlds). Thus, fantasizing it
>> might be different elsewhere doesn't guarantee just how it's
>> different -- regardless of our ability to know.
> 
> No, it doesn't have a direct bearing on the possibility of
> strong AI.

At least in the actual world, we're in agreement on that. :)

Regards,

Dan
 My latest Kindle book, "Born With Teeth," can be previewed at:
http://www.amazon.com/gp/product/B00N72FBA2
 (It doesn't deal with strong AI or possible worlds, AFAIK.:)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20141001/5eabf7c1/attachment.html>


More information about the extropy-chat mailing list