Re: The code for AGI will be simple

[spudboy100 via Everything List]

This is only opinion here, but I agree with Telmo because we as a species 
require and roi, a return on investment, for a few good reasons. 1) We all like 
money or free stuff, or both.2) In case robots make everyone unemployed.3) We 
need longer term goals to ensure human survival. So we require better longer 
term projects. 
The vagueness of Alice & BoB will eventuate in something tasty someday, But not 
today. 


-Original Message-
From: Telmo Menezes 
To: John Clark ; 'Brent Meeker' via Everything List 

Sent: Fri, Sep 9, 2022 4:00 pm
Subject: Re: The code for AGI will be simple

#yiv6150247816 p.yiv6150247816MsoNormal, #yiv6150247816 
p.yiv6150247816MsoNoSpacing{margin:0;}#yiv6150247816 p.yiv6150247816MsoNormal, 
#yiv6150247816 p.yiv6150247816MsoNoSpacing{margin:0;}

Am Fr, 9. Sep 2022, um 13:26, schrieb John Clark:

On Fri, Sep 9, 2022 at 5:20 AM Telmo Menezes  wrote:




>> Even if you have an IQ of 200 and spend your entire life studying 
>> consciousness you will advance the field precisely as much as the entire 
>> human race has in the last thousand years. And that would be precisely zero. 
>> Isaac Newton must've had an IQ of about 200 and unfortunately he spent much 
>> more time studying theology than physics and mathematics put together, but 
>> despite that colossal effort he advanced the field of theology not at all,


> Unfortunately for you maybe, but perhaps it gave him joy and I bet that was 
> the main thing that mattered to Isaac Newton. Good for him, I would say. At 
> some point we will all be dead, and nothing will matter or be useful to us by 
> then.


I can only speak for myself but all else being equal if given the choice 
between a task that has a chance of accomplishing something and a task that can 
only lead to a dead end then I would choose the one that might actually achieve 
something, even if that achievement is just an abstract understanding of 
something with no practical value. 


If you choose, try to lose
for the loss of remain come and start
Start the game
I chi-chi, chi-chi I
chi-chi-chi, ka-ta-ko, choose to choose
choose to choose, choose to go


John K Clark    See what's on my new list at  Extropolis
npv


-- 
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/CAJPayv1KXcLo84RTCVSK50G1wUbWs0Mh2uC1WqgLtTFkYVYt%2BA%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/151dd21a-ce7e-4fd2-9cc8-212b30c21c39%40www.fastmail.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/823152313.929595.1662766730263%40mail.yahoo.com.

Re: The code for AGI will be simple

[Telmo Menezes]

Am Fr, 9. Sep 2022, um 13:26, schrieb John Clark:
> On Fri, Sep 9, 2022 at 5:20 AM Telmo Menezes  wrote:
> __
> 
>>> >> Even if you have an IQ of 200 and spend your entire life studying 
>>> >> consciousness you will advance the field precisely as much as the entire 
>>> >> human race has in the last thousand years. And that would be precisely 
>>> >> zero. Isaac Newton must've had an IQ of about 200 and unfortunately he 
>>> >> spent much more time studying theology than physics and mathematics put 
>>> >> together, but despite that colossal effort he advanced the field of 
>>> >> theology not at all,
>> 
>> *> Unfortunately for you maybe, but perhaps it gave him joy and I bet that 
>> was the main thing that mattered to Isaac Newton. Good for him, I would say. 
>> At some point we will all be dead, and nothing will matter or be useful to 
>> us by then.*
> 
> I can only speak for myself but all else being equal if given the choice 
> between a task that has a chance of accomplishing something and a task that 
> can only lead to a dead end then I would choose the one that might actually 
> achieve something, even if that achievement is just an abstract understanding 
> of something with no practical value. 

If you choose, try to lose
for the loss of remain come and start
Start the game
I chi-chi, chi-chi I
chi-chi-chi, ka-ta-ko, choose to choose
choose to choose, choose to go

> John K ClarkSee what's on my new list at  Extropolis 
> 
> npv
> 
> 
> 
> -- 
> 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/CAJPayv1KXcLo84RTCVSK50G1wUbWs0Mh2uC1WqgLtTFkYVYt%2BA%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/151dd21a-ce7e-4fd2-9cc8-212b30c21c39%40www.fastmail.com.

Re: The code for AGI will be simple

[Jesse Mazer]

On Fri, Sep 9, 2022 at 8:26 AM smitra  wrote:

> So, I think insect-level AGI will cause a rapid transition to a machine
> civilization. This will lead to a new biology of machines with insect
> level intelligence ending up wiping out all life on Earth due to
> pollution, similar to the great oxygenation event:
>

Are you assuming insect-level AGI would also be small like insects and
could self-replicate just as rapidly using commonly-found materials as
"nutrients"? If we had insect-level AGI but they were larger and easier to
spot, and also took much longer than an insect to self-replicate (and
perhaps required external infrastructure or uncommon materials to do so),
it seems hard to imagine a scenario in which humanity wouldn't be able to
prevent them from going into runaway self-replication mode.

I think the possibility of relatively "dumb" self-replicating machines,
even if large and relatively slow like Eric Drexler's concept of a
"clanking replicator" (see
http://wfmh.org.pl/enginesofcreation/EOC_Chapter_4.html ), could disrupt
society for a different reason--they could spell the end of capitalism, or
at least radically change its nature. If there were commercially available
machines that could replicate themselves, those who owned them could make
copies for just the cost of raw materials and energy, and if they were
competing to sell them, competition would tend to drive the cost down to
materials/energy cost or barely above it, basically destroying profits for
any good that isn't forced into artificial scarcity by intellectual
property laws. This would likewise go for any other goods the machines are
capable of replicating. If self-replicating machines could also extract
resources (fully automated mining facilities, say), then profit would still
be possible if raw materials returned > raw materials invested (akin to
'energy return on energy invested' in energy economics), but if companies
were making profits by just setting up mining machines and then sitting
back and doing nothing, this would probably cause political instability,
both in democracies and autocratic systems, where either the people or the
politicians would likely prefer to be the ones reliably getting back more
than their initial investment with no work needed. Perhaps instead of
totally ending capitalism, we might end up with a hybrid system where some
sort of intellectual property laws would still be in place so companies and
individuals could still profit from those, but actual production machinery
would mostly be publicly owned, and people (along with retail companies)
could order up any good from a database of designs, receiving something
like a basic income in raw materials and energy (funded by mining and
energy generation facilities which could also be publicly owned).

Arthur C. Clarke, in his 1962 nonfiction book Profiles of the Future,
commented about how a self-replicating machine which could also replicate
other goods, which he just called a "Replicator", would disrupt our current
economic system:

"The advent of the Replicator would mean the end of all factories, and
perhaps all transportation of raw materials and all farming. The entire
structure of industry and commerce, as it is now organized, would cease to
exist. Every family would produce all that it needed on the spot — as,
indeed, it has had to do throughout most of human history. The present
machine era of mass-production would then be seen as a brief interregnum
between two far longer periods of self-sufficiency, and the only valuable
item of exchange would be matrices, or recordings, which had to be inserted
into the Replicator to control its creations.

"No one who has read thus far will, I hope, argue that the Replicator would
itself be so expensive that nobody could possibly afford it. The prototype,
it is true, is hardly likely to cost less than £1,000,000,000,000 spread
over a few centuries of time. The second model would cost nothing, because
the Replicator's first job would be to produce other Replicators. It is
perhaps relevant to point out that in 1951 the great mathematician, John
von Neumann, established the important principle that a machine could
always be designed to build any describable machine -- including itself.
The human race has squalling proof of this several hundred thousand times a
day.

"A society based on the Replicator would be so completely different from
ours that the present debate between Capitalism and Communism would become
quite meaningless. All material possessions would be literally cheap as
dirt. Soiled handkerchiefs, diamond tiaras, Mona Lisas totally
indistinguishable from the original, once-worn mink stoles, half-consumed
bottles of the most superb champagnes – all would go back into the hopper
when they were no longer required. Even the furniture in the house of the
future might cease to exist when it was not actually in use.”

Probably this book was a major influence on Gene Roddenberry's vision of a
post-scarcity 

Re: The code for AGI will be simple

[smitra]

This is something that fits in with what I wrote here some time ago 
about insect-level AI taking over from us.


A system with AGI doesn't have to be all that intelligent for it to be 
extremely useful. Today we cannot build a remotely controlled spider 
that could survive in Nature. That little intelligence a spider has is 
the GI it needs to take on the challenges of surviving. If we have 
something similar, say spider level AGI then that's good enough to fully 
automatize our entire economy with. The reason why you can't replace all 
factory worker with machines is due to a lack of even a minimal amount 
of AGI.


So, I think insect-level AGI will cause a rapid transition to a machine 
civilization. This will lead to a new biology of machines with insect 
level intelligence ending up wiping out all life on Earth due to 
pollution, similar to the great oxygenation event:


https://en.wikipedia.org/wiki/Great_Oxidation_Event

And as I pointed out earlier, I think this is a universal phenomena that 
all intelligent life is subject to. The whole point of being intelligent 
is to let as much of the work be done for you by entities that are 
dumber than you. But in that process that leads to faster and faster 
economic growth, its inevitable that at some point you are going to 
crate autonomous systems that will grow exponentially. The point where 
the transition to artificial life starts is going to be close to the 
minimum intelligence level needed for exponential growth.


If you make it hotter and hotter in some closed space, a fire will break 
out, this is going to happen close to the minimum required temperature 
for ignition, not at some extremely high value for the temperature. 
Nature shows us that the minimum amount of intelligence required for 
efficient self-maintenance and reproduction that yields exponential 
growth is very low.


Saibal





On 08-09-2022 14:09, John Clark wrote:

This is an interview of the great computer programmer John Carmack, he
thinks the time when computers can do everything, not just some
things, as good or better than humans is much closer than most people
believe, he thinks there is a 60% chance it will happen by 2030. Like
me Carmack is much more interested in intelligence than consciousness
and has no interest in the "philosophical zombie" argument. As far as
the future history of the human race is concerned the following
quotation is particularly relevant:

"___It seems to me this is the highest leverage moment for a single
individual potentially_ _in the history of the world._ [...]   _I am
not a mad man in saying that the code for artificial General
intelligence is going to be tens of thousands of lines of code, not
millions of lines of code. This is code that conceivably one
individual could write, unliker writing a new web browser or operating
system._"

The code for AGI will be simple [1]

John K ClarkSee what's on my new list at  Extropolis [2]

b30

 --
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/CAJPayv3ZEbXXVjs803%3Dutjc2pvkCgpZGA%2Bad_OWBhue-5kxDJQ%40mail.gmail.com
[3].


Links:
--
[1] https://www.youtube.com/watch?v=xLi83prR5fg
[2] https://groups.google.com/g/extropolis
[3]
https://groups.google.com/d/msgid/everything-list/CAJPayv3ZEbXXVjs803%3Dutjc2pvkCgpZGA%2Bad_OWBhue-5kxDJQ%40mail.gmail.com?utm_medium=email_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/d4b54074fe283e5c198ff6a6d709b143%40zonnet.nl.

Re: The code for AGI will be simple

[John Clark]

On Fri, Sep 9, 2022 at 5:20 AM Telmo Menezes  wrote:

>> Even if you have an IQ of 200 and spend your entire life studying
>> consciousness you will advance the field precisely as much as the entire
>> human race has in the last thousand years. And that would be precisely
>> zero. Isaac Newton must've had an IQ of about 200 and unfortunately he
>> spent much more time studying theology than physics and mathematics put
>> together, but despite that colossal effort he advanced the field of
>> theology not at all,
>
>
> *> Unfortunately for you maybe, but perhaps it gave him joy and I bet that
> was the main thing that mattered to Isaac Newton. Good for him, I would
> say. At some point we will all be dead, and nothing will matter or be
> useful to us by then.*
>

I can only speak for myself but all else being equal if given the choice
between a task that has a chance of accomplishing something and a task that
can only lead to a dead end then I would choose the one that might actually
achieve something, even if that achievement is just an abstract
understanding of something with no practical value.

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

npv

-- 
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/CAJPayv1KXcLo84RTCVSK50G1wUbWs0Mh2uC1WqgLtTFkYVYt%2BA%40mail.gmail.com.

Re: The code for AGI will be simple

[Telmo Menezes]

Am Do, 8. Sep 2022, um 17:00, schrieb John Clark:
> On Thu, Sep 8, 2022 at 8:19 AM Telmo Menezes  wrote:
> __
> 
>>>  >> Like me Carmack is much more interested in intelligence than 
>>> consciousness and has no interest in the "philosophical zombie" argument.
>> 
>> *> It is possible to be highly interested in both. Why not?*
> 
> Because one is a useful activity and the other is not.

If there is one thing one cannot doubt, John, is that you are thoroughly 
American :)

> Even if you have an IQ of 200 and spend your entire life studying 
> consciousness you will advance the field precisely as much as the entire 
> human race has in the last thousand years. And that would be precisely zero. 
> Isaac Newton must've had an IQ of about 200 and unfortunately he spent much 
> more time studying theology

Unfortunately for you maybe, but perhaps it gave him joy and I bet that was the 
main thing that mattered to Isaac Newton. Good for him, I would say. At some 
point we will all be dead, and nothing will matter or be useful to us by then.

Telmo

> than physics and mathematics put together, but despite that colossal effort 
> he advanced the field of theology not at all, and nobody else has managed to 
> do any better. The same is true with consciousness. 
> 
>> 
>>> 
>>> >> "***It seems to me this is the highest leverage moment for a single 
>>> >> individual potentially** **in the history of the world.* [...]   *I am 
>>> >> not a mad man in saying that the code for artificial General 
>>> >> intelligence is going to be tens of thousands of lines of code, not 
>>> >> millions of lines of code. This is code that conceivably one individual 
>>> >> could write, unliker writing a new web browser or operating system.**"*
>> 
>> *> In a sense, I agree. But remember that, even with code, we are sitting on 
>> the shoulders of giants. A few lines of code in contemporary Python mobilize 
>> decades upon decades of the blood sweat and tears of the programmers that 
>> came before, who built all of this amazing infrastructure. How many lines in 
>> the Linux kernel?*
> 
> That's why I disagree with those who say Moore's law only applies to hardware 
> and not to software.  Imagine if there were no modern software tools and you 
> had to program everything in machine language using nothing but 0 and 1. 
> Fortunately we don't have to do that because machines have been able to help 
> us write computer programs for many decades. 
> 
> John K ClarkSee what's on my new list at  Extropolis 
> 
> stc
> 
> 
> 
> -- 
> 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/CAJPayv3Oo3GJW%2BXU%2BgO0_h0cKXf0EwpEooOiEZwk%3DnqR_EmNdg%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/37b3f864-41d5-4f58-8b0c-c3628af5a813%40www.fastmail.com.

Re: The code for AGI will be simple

[Henrik Ohrstrom]

Could research into unconsciousness be somewhat easier?
When are a brain unconscious?
During anesthesia you might say.
Alas, as everything else, it is not that simple.

If we look at what's actually required for a satisfied patient, a truly
unconscious brain ai'nt that.
Especially as long as both lawyers and patients agree that impaired memory
are enough for satisfaction.
Absence of objections are not evidence of good enough.
A well and sufficiently (not over) medicated patient have a quite active
brain during anesthesia. Conscious? Probably not. Memory retention?
Hopefully not . Pain reactions to surgery? Actually possible to measure, se
www.med-storm.com most patients with classic standard anesthesia are
severly as in cognitive impairment dose level overmedicated during surgery.
No Pain.

So a patient who officially are the definition of unconscious, aint and
should not be.

So what is unconscious then? I would wager that it is difficult to prove
anything about a regular stone, so lets call that one unconscious.
/Henrik


Den fre 9 sep. 2022 00:44John Clark  skrev:

> On Thu, Sep 8, 2022 at 6:31 PM Brent Meeker  wrote:
>
>    Like me Carmack is much more interested in intelligence than
> consciousness and has no interest in the "philosophical zombie" argument.


 *>>>  It is possible to be highly interested in both. Why not?*

>>>
>>> * >> If one were unconscious to whom would it be useful?*
>>
>>
>
> > Because one is a useful activity and the other is not.
>
>
> I meant that research into the nature of intelligence is useful, that is
> to say it leads somewhere, but consciousness research leads nowhere and
> accomplishes nothing.
>
>  John K ClarkSee what's on my new list at  Extropolis
> 
>
> kkq
>
>
>
>
> --
> 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/CAJPayv13bweRRmHnBk4BHB7ASwCgwTSWmgucqeCNevQrahB2EA%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/CAF0GBnhczN%2B0K7jcS9upCSLU2Pz4nwnoRT741e5WEYdGZA8GAw%40mail.gmail.com.