[ExI] all we are is just llms was: RE: e: GPT-4 on its inability to solve the symbol grounding problem

[Giovanni Santostasi]

Spike,
I don't think you can have real intelligence without consciousness, at
least the type of consciousness that matters and that we recognize as
humans. For a system to be really intelligent needs to know its own state
and self-reflect. I heard Wolfram explain something very interesting that
is coming out from its models of reality based on graph theory. In his
model of the universe, absolute time is the set of all possible events, but
this is a static and lifeless universe.

To make sense of the world (in particular physical reality) you need a mind
that organizes time as a sequence (while time without a mind is just an
infinite net of all possible events). If this deep insight is true then
reality happens when a mind knows the world, so consciousness and
intelligence are strictly related. This also solves the "explanatory gap"
that people claim is at the core of what consciousness is.

There is no explanatory gap as I mentioned many times because knowing and
experiencing are the same thing. What the mind experiences are a particular
individual time ordering that is equivalent to the laws of physics. Wolfram
literally derives the laws of physics from a net of possibilities and an
individual mind experiencing a particular "slice" of these nodes in the net
of all possibilities. It is a beautiful and fascinating result very similar
to the ancient Eastern understanding of how reality came to be, Lila or
cosmic play, where non-individuated consciousness needs to individuate to
give birth to reality.

By the way, the ancient Rishi of India would say the essence of reality is
sat, chit and ananda, existence, consciousness, and bliss. The bliss part
seems strange to Westerners but I think consciousness is associated with a
sense of connectedness with the world that is a blissful state. This is
also why I don't fear super-intelligent AI, I think super-intelligent AI
cannot be evil, it will be Buddha-like. I know it sounds wishful thinking
but it is not.

Anyway in terms of knowing if a system is conscious or not the simple way
to solve this problem is to realize that consciousness is what
consciousness does. If a system seems it is conscious then it is conscious.
This applies to humans as well as to other minds.
Giovanni





On Thu, Apr 20, 2023 at 9:55 PM spike jones via extropy-chat <
extropy-chat at lists.extropy.org> wrote:

>
>
>
>
> *From:* Giovanni Santostasi <gsantostasi at gmail.com>
> *Subject:* Re: [ExI] all we are is just llms was: RE: e: GPT-4 on its
> inability to solve the symbol grounding problem
>
>
>
> >…Spike,
>
> I spent some time looking at Berger online activity and she is spewing a
> lot of hate for these supposed artificial minds. She even calls them
> "so-called AI". She is super biased for different reasons, probably she is
> a post-modern superwoke advocate of some kind that thinks so-called AI will
> continue the oppression of the patriarchy.
> Giovanni
>
>
>
>
>
>
>
> OK no worries Giovanni, I know nossink about Berger and I am not following
> that.
>
>
>
> I have no particular insight into Eliezer’s contentions he has been
> carrying around for over 30 years, about whether AI will kill us.  I don’t
> know enough to contribute to that debate.
>
>
>
> I am now interested in the idea that consciousness and intelligence can be
> separated (never mind emotion for now (that one is just too dang
> complicated)) and if so, how would we measure consciousness?
>
>
>
> spike
>
>
>
>
> _______________________________________________
> extropy-chat mailing list
> extropy-chat at lists.extropy.org
> http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20230420/ce809584/attachment.htm>