An operational test for intelligence requires that ability to act in the
world to achieve goals. LLM's are intelligent in that they act to
satisfy prompts. If you went to the beach and you said to a crab,
"Write in the sand a short poem about waves." and the crab scratched out:
Born
of wind and
earth's embrace
an ocean's memory of
storms beyond the horizon
its undulating information uselessly inscribed
in the meandering sand
finds voice at last
its fall a sigh a
single syllable
of surf
You'd think the crab was pretty smart. An LLM could do that. The only
reason for us thinking it is not intelligent is that we know how the LLM
does it. When I first took a class in AI fifty years ago at UCLA. The
Professor explained that the definition of intelligence was "whatever
computers can't do yet."
Brent
On 3/12/2023 9:29 PM, spudboy100 via Everything List wrote:
Does it really display intelligent behavior, and now you will need to
define intelligent behavior with testable parameters. Do humans
display intelligent behavior? How often, and how consistently? Many
say crows do.
This was part of my reason in hectoring JC on this. A neurobiologist
tells, via research how parts of the human brain behave, and thus
produce intelligence and (non-philosophically) consciousness. What
part of the server farm enacts the illusion (I think) of intelligence?
-----Original Message-----
From: Stathis Papaioannou <[email protected]>
To: [email protected]
Sent: Sun, Mar 12, 2023 9:29 pm
Subject: Re: ChatGPT's rebuttal to Chomsky
On Mon, 13 Mar 2023 at 12:18, Jason Resch <[email protected]> wrote:
On Sun, Mar 12, 2023, 8:57 PM Stathis Papaioannou
<[email protected]> wrote:
On Mon, 13 Mar 2023 at 04:12, Lawrence Crowell
<[email protected]> wrote:
Remember no matter haw complex an algorithm it is
ultimately a lot of Boolean switching acting on what ever
data is dumped into it.
That means that if the entity displays intelligent behaviour,
Boolean switching acting on whatever data is dumped into it
can generate intelligent behaviour.
Boolean logic gates (and, or, not) are universal. There's no
finitely describable behavior that can't be replicated by the
repeated application of boolean logic gates paired with a memory.
Yes. But more generally, even from a position of ignorance, if you
don’t believe that something made from electrical circuitry (or
whatever) can display intelligent behaviour, and something made from
electrical circuitry does in fact display intelligent behaviour, that
means you were wrong.
--
Stathis Papaioannou
--
You received this message because you are subscribed to the Google
Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send
an email to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/everything-list/CAH%3D2ypXZLOOeKndehY1w3iseRPbrbF1HGG1dAk9rDh%3DvOtfVFA%40mail.gmail.com
<https://groups.google.com/d/msgid/everything-list/CAH%3D2ypXZLOOeKndehY1w3iseRPbrbF1HGG1dAk9rDh%3DvOtfVFA%40mail.gmail.com?utm_medium=email&utm_source=footer>
.
--
You received this message because you are subscribed to the Google
Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send
an email to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/everything-list/922538855.1640289.1678681763911%40mail.yahoo.com
<https://groups.google.com/d/msgid/everything-list/922538855.1640289.1678681763911%40mail.yahoo.com?utm_medium=email&utm_source=footer>.
--
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/everything-list/a38a6774-b5db-b940-d232-1d7f7aa5df3a%40gmail.com.
