On 23 Feb 2017, at 21:12, Brent Meeker wrote:
On 2/23/2017 12:34 AM, Bruno Marchal wrote:
On 20 Feb 2017, at 16:33, Telmo Menezes wrote:
On Sat, Feb 18, 2017 at 1:19 AM, John Clark <[email protected]>
wrote:
On Wed, Feb 15, 2017 Telmo Menezes <[email protected]> wrote:
Dark Matter and Dark Energy remain complete mysteries.
As far as I can tell, what we have is a falsification of current
theories. They appear to be good enough approximations for many
things, but then they fail at predicting the expansion rate of the
universe right? Maybe it's dark matter, maybe it's something else,
They are 2 separate mysteries. Dark Matter is a mysterious
something that
makes up 28% of the universe and holds galaxies and clusters of
galaxies
together. Dark Energy is a even more mysterious something that
makes up 69%
of everything and causes the expansion of the entire universe to
accelerate.
And about 4% of the universe is made of the sort of normal matter
and energy
that until about 20 years ago was the only type we thought existed.
There is a straightforward extension of General Relativity and
Quantum
Mechanics that explains Dark Energy, however it gives a figure
that is
10^120 too large, it's been called the worse mismatch between
theory and
observation in the entire history of science. I think it's fair
to say we
really don't have a clue about Dark Energy, and Dark Matter is
almost as
confusing.
If science failed so far at explaining something, then it doesn't
matter?
Science has an explanation for consciousness that works
beautifully,
consciousness is the way information feels when it is being
processed
intelligently.
I know that your position is that information processing is
nonsensical without matter. Many times you invited Bruno to compete
with Intel, etc. So what you are saying is that "consciousness is
the
way matter feels when it participates in an intelligent
computation".
This "explanation" begs the question already.
Then there's the issue of defining "processed intelligently". What
does that even mean? Where do you draw the line between intelligent
and non-intelligent processing? Let me guess: intelligent processing
is the kind that generates consciousness.
Nobody ever came up with a way to test for the presence of
consciousness (probably because it's the wrong way to think about
it),
so there is no scientific theory about it. Zero. You make it worse
by
introducing ill-defined concepts.
What science doesn't yet have is a complete theory explaining
how to produce intelligence, but enormous progress has been made
in just the
last few years.
Not really. What is happening is that the artificial neural network
models from the 80s are finally paying off, because of the orders of
magnitude more computational power and training data that we have
now.
Progress is being made, but it has been very slow. It's a hard
problem.
I've worked in this field both in academia and industry, for what
it's worth.
The study of intelligence, now that's important!
That is a statement of faith. Gizmo worshiping.
At least 3 times a week for the last 5 years somebody on this
list has
accused me of being religious, apparently in the hope that I'll
burst into
tears and cry myself to sleep. It's not going to happen,
I can't talk for the others, but I have no interest in making you
feel bad.
I'm just pointing out dogmatic thinking.
Yes, it's important in
a sense. I too am interested in having medical breakthroughs,
freedom
from labour and all the nice things that AI can bring.
It's important even if you're only interested in philosophical
problems,
such as why did Evolution bother to make conscious animals at all.
Evolution is a theory on the origins of biological complexity. We
know
nothing about consciousness.
Do you agree that consciousness is a form of knowledge? That is:
consciousness requires some knowledge, and (genuine) knowledge
requires some conscious person)?
I don't think knowledge requires consciousness, much less a person.
Then we are talking about different things. Knowledge, in the sense of
cognitive science, or epistemology, requires a knower, which is
usually a person. Eventually, we need only to agree on the axioms, and
propose variant.
In the context of describing "consciousness", it is seen as a
particular "knowledge". (plausibly [1]<1>t)
Then do you agree with the S4 theory of rational knowledge, which
is that
A theory of "knowable" is not the same as a theory of knowledge. A
theory of knowledge has to include the fact that much less is known
than is knowable.
OK. It is the "omniscience problem". In the ideal context of the
arithmetically self-referentially correct machine, this is handled by
the difference between (x is a proof of y) and Ex(x is a prove of y).
We can use this to have still different modal variant, like proving-
this-day, (x is a shorter proof of y than z), etc. But the use of the
transitive "able" is simpler (in machine's "theology").
(knowable x) implies x
(knowable (x implies y)) implies ((knowable x) implies (knowable y))
(knowable x) implies (knowable (knowable x))
With the inference rules:
If I prove x I can deduce (knowable x)
+ modus ponens
Proof is realtive to premises. If you can prove x from true
premises, THEN it's knowable.
Yes. The premises have to be true, and the inference rules have to
preserve the truth, which can be proved by (transfinite) induction ...
by a machine having more induction power, which you can't prove to be
correct, except by adding induction, well, you get the transfinite etc.
But there is a miracle, we all feel that Robinson axioms(*) are true
about the natural numbers (the least inductive set). And we don't need
more to get the dreams, and, well, the math exist to see if those
dreams "cohere" enough to lead to the sharable quanta. Does
consciousness' stability requires deep histories, coupled with a
continuum?
(*) Robinson theory (above classical first order predicate calculus/
logic):
0 ≠ s(x)
s(x) = s(y) -> x = y
x = 0 v Ey(x = s(y))
x+0 = x
x+s(y) = s(x+y)
x*0=0
x*s(y)=(x*y)+x
The knower I am talking about, canonically attached to any Löbian
machine (universal machine knowing that they are universal, in a
technical sense I gave once), is any system which believes the axioms
above (or some "Turing"-equivalent), and believes in enough induction
axioms, notably on all formula first order (the variable denotes the
basic object we talk about, not set or functions applied to those
objects):
(F(0) & Ax(F(x) -> F(s(x))) -> AxF(x)
All self-referentially correct extensions, and even chains of
extensions, obeys to G and G*, and the knower is defined by the
theaetetus' idea applied on those beliefs, which by incompleteness do
not collapse into G. Instead, we get a logic of knowledge, or
knowledgeability if you prefer, and it can be used to understand that
the machine will be confused by her own understanding and
"experiencing" the difference between the first person view, "NOT a
number", and the third person view "Numbers and numbers relations".
Physics has to be invariant for the choice of the universal basic
system, when we assume mechanism. We can use any universal machine to
start with, we get the same "mind-body" problem, and the same solution
(laws of quanta and qualia, and their relations).
To use this approach in physics would be, today, non sensical, ---a
bit like using string theory to make a pizza. The goal is just to get
the hopefully less wrong big picture, and this in a way which does not
eliminate consciousness and persons.
Bruno
Brent
--
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 post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.
http://iridia.ulb.ac.be/~marchal/
--
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 post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/everything-list.
For more options, visit https://groups.google.com/d/optout.
