date:20230402

Re: GPT4 + Mathematica

2023-04-02 Thread Brent Meeker

That's fascinating. I had not heard to the magnetic
stimulation/inhibition experiments. Have you a reference?

Brent

On 4/2/2023 2:58 PM, smitra wrote:
While GPT hooked up to Mathematica may greatly improve its
performance, I would not be impressed when judging that in terms of
approaching human level cognition.

A long time ago I was teaching a statistical mechanics class, the
students had to do a certain calculation for homework that involved
doing the series expansion of 1/sinh(x) to order x. I was shocked to
see that none of the students could do it. But some did come up with
the right answer, they had shown some failed attempts and then wrote
that they used Mathematica and Mathematica says that the answer is
1/x - x/6 ...

The students couldn't do it because they could not use the Taylor
expansion formula due to the singularity at x = 0. Most students had
not bothered to think about how to bypass that problem, otherwise at
least a few would have written down the correct derivation:

1/sinh(x) = 1/(x + x^3/6 +...) = 1/x 1/(1 + x^2/6 + ...) = 1/x (1 -
x^2/6 + ...) = 1/x - x/6 +...

Now, I do think that GPT is a great leap forward, I don't want to
downplay the progress made. But I'm quite skeptical about the idea
that such systems are approaching human level cognition. The human
brain is an enormously powerful system, but we don't have free access
to use the power of our brains to do whatever we want. For example,
most people cannot multiply two 5 digit numbers in their head, but a
simple pocket calculator has no problems with that task.

However, some autistic savants do have more of a privileged access to
use the power of the our brains to do arithmetic. Some of them can do
calculations that most other people cannot do. In certain tests
involving magnetic stimulation or inhibition of certain brain parts
performed on ordinary people, it has been shown that people can
temporarily gain certain abilities that they normally don't have. For
example, if you look at a screen with a few hundreds dots on it, can
you count the number of displayed dots in one second? Most people
can't, some autistic savants can do this. But in the
experiment with magnetic stimulation or inhibition, the test subjects
were also able to do this.

When we are consciously using our brains doing complicated things e.g.
mathematics, then we are using our brains in an extremely inefficient
way. If we could have an artificial brain similar to our brain but one
which is completely dedicated to doing mathematics instead of what our
brains are dedicated to do, then it would be enormously better at math
than we are. It's then likely that something like the brain of a
lizard that is fully dedicated to math would already completely
outclass the world's best mathematicians. Perhaps even the brain of
insects could perform at the same level of most mathematicians when
fully dedicated to doing math.

Saibal

On 28-03-2023 23:32, John Clark wrote:

Apparently I'm not the only one who has become obsessed with the
developments in AI that have occurred in the last few weeks, Stephen
Wolfram, the man who developed Mathematica, started Wolfram Alpha, and
wrote the book A New Kind Of Science about cellular automation, has
given two very interesting interviews on the subject. Wolfram wrote a
plug-in to connect GPT4 with Mathematica because calculation was the
one thing that GPT4 was not very good at but Mathematica is superb at
it; he describes the experience as "poking at an alien intelligence".

GPT + Wolfram: The Future of AI is Here! [1]

GPT, AI, and AGI with Stephen Wolfram [2]

John K Clark See what's on my new list at Extropolis [3]

9eq

--
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 everything-list+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/everything-list/CAJPayv2%3D_vDQ1fhCpoa3nv20TC66iWGLb1oiEtEmpGfb3YOQiA%40mail.gmail.com

[4].

Links:
--
[1] https://www.youtube.com/watch?v=z5WZhCBRDpU
[2] https://www.youtube.com/watch?v=szxiPMyuMGY
[3] https://groups.google.com/g/extropolis
[4]
https://groups.google.com/d/msgid/everything-list/CAJPayv2%3D_vDQ1fhCpoa3nv20TC66iWGLb1oiEtEmpGfb3YOQiA%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 everything-list+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/everything-list/1f979cc2-1c03-7e30-f704-1c0285a55bc2%40gmail.com.

Re: GPT4 + Mathematica

2023-04-02 Thread smitra

While GPT hooked up to Mathematica may greatly improve its performance,
I would not be impressed when judging that in terms of approaching
human level cognition.

A long time ago I was teaching a statistical mechanics class, the
students had to do a certain calculation for homework that involved
doing the series expansion of 1/sinh(x) to order x. I was shocked to see
that none of the students could do it. But some did come up with the
right answer, they had shown some failed attempts and then wrote that
they used Mathematica and Mathematica says that the answer is 1/x - x/6
...

1/sinh(x) = 1/(x + x^3/6 +...) = 1/x 1/(1 + x^2/6 + ...) = 1/x (1 -
x^2/6 + ...) = 1/x - x/6 +...

Now, I do think that GPT is a great leap forward, I don't want to
downplay the progress made. But I'm quite skeptical about the idea that
such systems are approaching human level cognition. The human brain is
an enormously powerful system, but we don't have free access to use the
power of our brains to do whatever we want. For example, most people
cannot multiply two 5 digit numbers in their head, but a simple pocket
calculator has no problems with that task.

However, some autistic savants do have more of a privileged access to
use the power of the our brains to do arithmetic. Some of them can do
calculations that most other people cannot do. In certain tests
involving magnetic stimulation or inhibition of certain brain parts
performed on ordinary people, it has been shown that people can
temporarily gain certain abilities that they normally don't have. For
example, if you look at a screen with a few hundreds dots on it, can you
count the number of displayed dots in one second? Most people can't,
some autistic savants can do this. But in the
experiment with magnetic stimulation or inhibition, the test subjects
were also able to do this.

When we are consciously using our brains doing complicated things e.g.
mathematics, then we are using our brains in an extremely inefficient
way. If we could have an artificial brain similar to our brain but one
which is completely dedicated to doing mathematics instead of what our
brains are dedicated to do, then it would be enormously better at math
than we are. It's then likely that something like the brain of a lizard
that is fully dedicated to math would already completely outclass the
world's best mathematicians. Perhaps even the brain of insects could
perform at the same level of most mathematicians when fully dedicated to
doing math.

Saibal

On 28-03-2023 23:32, John Clark wrote:

GPT + Wolfram: The Future of AI is Here! [1]

GPT, AI, and AGI with Stephen Wolfram [2]

John K ClarkSee what's on my new list at Extropolis [3]

9eq

--
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 everything-list+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/everything-list/d853615bdddbd0b993040c282514ae15%40zonnet.nl.

Re: It's too late to stop GPT4 now

2023-04-02 Thread spudboy100 via Everything List

Maybe someday we humans can do the same? 

-Original Message-
From: John Clark 
To: 'Brent Meeker' via Everything List 
Sent: Sun, Apr 2, 2023 3:35 pm
Subject: It's too late to stop GPT4 now

This video is a summary of several technical papers that have come out in the 
last 72 hours, apparently GPT4 can now improve itself without human help by 
self-reflecting on its errors and can even design better hardware for itself. 
GPT 4 Can Improve Itself by self reflection 
John K Clark    See what's on my new list at  Extropolis
3zi-- 
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 everything-list+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/everything-list/CAJPayv1Zdxw4fhV7Vq%3DWHDL5SOUNbnyCNgFmjuEo8%3DqOAC%2Busg%40mail.gmail.com.

-- 
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 everything-list+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/everything-list/119213892.2325970.1680469555398%40mail.yahoo.com.

Re: It's too late to stop GPT4 now

2023-04-02 Thread Jason Resch

"Let an ultraintelligent machine be defined as a machine that can far
surpass all the intellectual activities of any man however clever. Since
the design of machines is one of these intellectual activities, an
ultraintelligent machine could design even better machines; there would
then unquestionably be an 'intelligence explosion,' and the intelligence of
man would be left far behind... Thus the first ultraintelligent machine is
the last invention that man need ever make, provided that the machine is
docile enough to tell us how to keep it under control. It is curious that
this point is made so seldom outside of science fiction. It is sometimes
worthwhile to take science fiction seriously."
-- I.J. Good

On Sun, Apr 2, 2023, 3:35 PM John Clark  wrote:

> This video is a summary of several technical papers that have come out in
> the last 72 hours, apparently GPT4 can now improve itself without human
> help by self-reflecting on its errors and can even design better hardware
> for itself.
>
> GPT 4 Can Improve Itself by self reflection
> 
>
> John K ClarkSee what's on my new list at  Extropolis
> 
> 3zi
>
> --
> 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 everything-list+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/everything-list/CAJPayv1Zdxw4fhV7Vq%3DWHDL5SOUNbnyCNgFmjuEo8%3DqOAC%2Busg%40mail.gmail.com
> 
> .
>

-- 
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 everything-list+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/everything-list/CA%2BBCJUhkHk6Apmq%3D%2BzUYMpX1tNjc3TzYUwyJW1x_WthQy0p7BA%40mail.gmail.com.

It's too late to stop GPT4 now

2023-04-02 Thread John Clark

This video is a summary of several technical papers that have come out in
the last 72 hours, apparently GPT4 can now improve itself without human
help by self-reflecting on its errors and can even design better hardware
for itself.

GPT 4 Can Improve Itself by self reflection


John K ClarkSee what's on my new list at  Extropolis

3zi

-- 
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 everything-list+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/everything-list/CAJPayv1Zdxw4fhV7Vq%3DWHDL5SOUNbnyCNgFmjuEo8%3DqOAC%2Busg%40mail.gmail.com.

Re: GPT4 + Mathematica

Re: GPT4 + Mathematica

Re: It's too late to stop GPT4 now

Re: It's too late to stop GPT4 now

It's too late to stop GPT4 now

5 matches

Site Navigation

Mail list logo

Footer information