Sequoia Supercomputer Towers above the Rest

[Daniel Shawul]

the neurons maybe, but not consciousness.
cmputers are deterministic, a humans brain is not.
at least mine is not or at least i believe its not.

That is still being argued. Logical reasoning both can do however.

irrespective of my writing style.
which i learned from Blik(C) btw on ICC

Just to be clear, that was a joke which you picked on immediately
I write worse than that most of the time.
cheers

[jefk]

[quote="Daniel Shawul"][quote]
the neurons maybe, but not consciousness.
That is still being argued[quote]

yes indeed and its not such an easy problem.
but to put it simple: artificial intelligence
depends on the software which is running on such
a comp no matter how fast, and how parallel or not.
and guess what, so far most of our software is
deterministic unless we build in some random
generators of course, and yes then we get
the problem of so-called free will. to see how far current thinking
goes you might like to have a look at eg:
http://brainvoyage.com/RBC/2Ed.php
(and btw i also gave some input to these guys;
and i still disagree a bit about the title but for the
rest the content is interesting although i'm more into
string theory then into quantumconscious stuff and
so on; but then there still are things physics is missing)

irrespective of my writing style.
[/quote]
Just to be clear, that was a joke which you picked on immediately
>I write worse than that most of the time.[/quote]

ah ok well your writing style is clear if people know who
a bit more about who you are i guess, something which often
the case in communication btw, also something which
computer software really hardly cant do; they even cannot
translate properly yet, let alone try to understand people.

if it theoretically would be possible ?
well a biological and philosophical question again.
chess might well be solved up to the level that
1.Nh3 and 1.g4 are losing, but before we have thinking
robots we are at least thousand years further imho.
luckily because of the ethical implications.
which goes a bit further than cheating quesions
for the Fide whether the players or their toilets
have houdini chips implanted.


jef

[ZirconiumX]

I had a very weird and off topic thought, reading this article: my first thought was "wow, this thing must cost in one second the amount of electricity I would use in one year!"

Actually, Blue Gene nodes are designed to be as power efficient as possible. So you have two sucky PPC chips @ 1.6GHz per node. But because ibm chose quantity over quality you need a LOT of them.

Matthew:out

[ZirconiumX]

Jef,

The problem of 'free will' or rampancy as they called it was explored in a game called Marathon.

First you have the problem of hardware. You would need a large cloud of computers, because a limitation of hardware is a limitation of growth.

I'll put this into perspective. Run fruit on the command line and give it a go infinite command. You will eventually reach a stage where nps and cpuload are maxed out. This is the limitation of growth due to limitation of hardware. And this is the same problem for AI.

Second you have the problem of rampancy itself. No amount of software restraint could be able to contain an AI that is rampant, because it is intelligent, and can experiment. Experiment several billion times a second.

Matthew:out

[rbarreira]

Turns out article I read is quite old. Rat's and Cat's brain have been solved.

the neurons maybe, but not consciousness.
cmputers are deterministic, a humans brain is not.
at least mine is not or at least i believe its not.
irrespective of my writing style.
which i learned from Blik(C) btw on ICC

Are you saying that your brain does not follow the laws of physics? If it does follow them then it's very hard to argue that your consciousness / intelligence / anything else happening in your brain is not reproducible.

As for being deterministic or not, that doesn't matter, not all computers are deterministic anyway.

[jefk]

[quote="rbarreira"][quote="Daniel Shawul"]Turns out article I read is quite old. Rat's and Cat's brain have been solved.

Are you saying that your brain does not follow the laws of physics? If it does follow them then it's very hard to argue that your consciousness / intelligence / anything else happening in your brain is not reproducible.
As for being deterministic or not, that doesn't matter, not all computers are deterministic anyway.[/quote]

no, i don't say the brain doesn't follow the laws of physics, but i say
that it doesnt work like a computer. computers which are not
deterministic ? what do you mean ? chaos theory or so ?
still deterministic, but not predictable.

as for the brain, not all laws of physics are know in its details,
in astrophysics we don't know much about dark matter
and dark energy, in physics we don't know everything about
mult-component quantum-physics for example.
example if two photons can seem to communicate at a distance
like in the EPR experiment than larger scale information structures
could also be coupled or transmitting info in some way
which we don't know of yet. might be the case in biology

[hgm]

Have you ever tried running games between SMP engines?

Non-determinism has nothing to do with consciouness. Nor with free will.

[jefk]

[quote="hgm"]Have you ever tried running games between SMP engines?
Non-determinism has nothing to do with consciouness. Nor with free will.[/quote]

depends on the definition of determinism and consciousness.
and free will as well.
A guy like Roger Penrose writes a lot about such things
(Hofstadter as well) in more words than you do Harm.

for the rest i assume you know that a computer is nothing
more than a bunch of logical nand and nor gates and so on
or putting it simply a binary machine crunching
hexadecimal numbers. bits and bytes.
and running with a rather constant clock cycle
although you might have intel turbo boost or whatever.
now you might even build in some feedback loops
and some different sort of memory but i suspect you will
have a difficult job in trying to build 'consciousness' and
'intelligence' (another word which needs a definition)
into such a machine.
PS even some neuroscientists as Dutch Diederik Swaab may
think, or at least suggest in their writings that consciousness
and free will is an illusion, well i don't.
There are many things in biology which physicists -nor biologists-
thus science itself can explain yet, how do pigeons find
their way, earth magnetism ? or other things.
Do they have consciousness ?
Once a physicist from Amsterdam Lagendijk wrote a book
about the arrogance of the average physicist. Might
apply to information sciences as well.

[jefk]

as i hadnt replied to the first statement, quoting below:

[quote="hgm"]Have you ever tried running games between SMP engines?

[/quote]

yes, often.
But you mean on one and the same computer i suppose ?
well it looks pretty deterministic if you ask me.
oh and this neuroscientist is Dick S (not Diederik S, thats someone else
PS and the opposite view of Penrose is held by someone like Minsky,
which while applying Occams razor is an interesting hypothesis
yet one which fail in the longer run as future will show imho.

[Uri]

I would be interested to know how much it cost to build this computer.

My estimate is that something like this costs something like 600 million dollars to build and maintain.

Bill Gates could still buy something like this because Gates has a networth of about 65 billion dollars, maybe even more.