[MURG] The main problem...

[Eugen Leitl]

On Sat, Feb 07, 2004 at 04:10:04AM -0500, Ed Minchau wrote:

> I would not be so quick to dismiss this.  There is no
> qualitative difference between a "human-level" AI and

The qualitative difference is that you need many, many orders of magnitude
more resources to achieve equivalent performance of an AI or an upload via
tissue modelling route. I've also pointed out that a mere recoding will
result in an instanteous performance increase of two to three orders of
magnitude. We clearly can't do that recoding. 

It all boils down at to how much resources you need to bootstrap AI, and how
difficult it is to fall up several theory levels in modelling purely by
machine learning. The news from computational sciences is that it's hard. We
don't know how, yet.

There's considerable uncertainty here, and that will decide whether our goose
is cooked, or not.

> an uploaded mind, only the route chosen.  Supposing
> that a human-level AI would not be sentient is to deny
> that it is a human-level AI at all (since humans are
> AFAIK sentient).
> 
> Part of the problem with this discussion is that we
> are trying to predict what would happen after the
> Singularity (heralded by the arrival of human-level

Singularity? You're soaking in it. I'm not trying to forecast what's going to
happen. I'm trying to point out that some, potentially avoidable, scenarios
are highly undesirable.

> AI); a time beyond which predictions become
> impossible.  It isn't that we don't know, we _cannot_
> know.

You're giving up before even trying. 

-- Eugen* Leitl <a href="http://leitl.org">leitl</a>
______________________________________________________________
ICBM: 48.07078, 11.61144            http://www.leitl.org
8B29F6BE: 099D 78BA 2FD3 B014 B08A  7779 75B0 2443 8B29 F6BE
http://moleculardevices.org         http://nanomachines.net
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: not available
Url : http://askja.bu.edu/pipermail/murg/attachments/20040207/0cc4f934/attachment.pgp