Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.prolog > #14878 > unrolled thread

Vanishing Point . [Le Dot]

Back to article view | Back to comp.lang.prolog

Contents

  Vanishing Point . [Le Dot] Mild Shock <janburse@fastmail.fm> - 2025-10-05 00:51 +0200
    Lets compare AI data centers (Was: Vanishing Point . [Le Dot]) Mild Shock <janburse@fastmail.fm> - 2025-10-05 01:03 +0200
      USA domestic chips for the never ending crunch (Re: Lets compare AI data centers ) Mild Shock <janburse@fastmail.fm> - 2025-10-05 01:18 +0200
        Holy Schrimp Jesus: The Internet is Dead [Rossy Boy dethroned?] (Re: USA domestic chips for the never ending crunch) Mild Shock <janburse@fastmail.fm> - 2025-10-05 10:32 +0200
          Try https://mathstodon.xyz by Christian Lawson-Perfect (Re: Holy Schrimp Jesus: The Internet is Dead [Rossy Boy dethroned?] ) Mild Shock <janburse@fastmail.fm> - 2025-10-07 19:45 +0200
            Some statistics about Mastodon (Re: Try https://mathstodon.xyz by Christian Lawson-Perfect) Mild Shock <janburse@fastmail.fm> - 2025-10-07 19:55 +0200
        The joy of running an AI Model on your AI Laptop [Local AI] (Re: USA domestic chips for the never ending crunch) Mild Shock <janburse@fastmail.fm> - 2025-11-11 21:15 +0100
          Cute Racoon generated with Local AI (Re: The joy of running an AI Model on your AI Laptop [Local AI]) Mild Shock <janburse@fastmail.fm> - 2025-11-11 23:28 +0100
            What about Local AI? [Archimedes Plutonium AI Persona?] (Re: Cute Racoon generated with Local AI) Mild Shock <janburse@fastmail.fm> - 2025-11-11 23:36 +0100
              Ross Finlayson is completely Paranoid (Re: What about Local AI? [Archimedes Plutonium AI Persona?]) Mild Shock <janburse@fastmail.fm> - 2025-11-13 17:51 +0100
                How to make an AI Persona [With Local AI?] (Re: Ross Finlayson is completely Paranoid) Mild Shock <janburse@fastmail.fm> - 2025-11-13 17:54 +0100
    British writer Arthur C. Clarke (Re: Vanishing Point . [Le Dot]) Mild Shock <janburse@fastmail.fm> - 2025-10-05 19:40 +0200
      Could the AI Persona evolve over time? [With Local AI?] (Re: British writer Arthur C. Clarke) Mild Shock <janburse@fastmail.fm> - 2025-11-13 18:01 +0100
        Boris the Loris and Nazi-Retard Julio (Re: Could the AI Persona evolve over time? [With Local AI?]) Mild Shock <janburse@fastmail.fm> - 2025-11-13 18:08 +0100
    Knight Rider (2025/2026) will kill us all (Re: Vanishing Point . [Le Dot]) Mild Shock <janburse@fastmail.fm> - 2025-11-13 21:53 +0100
      The AGI Gambit: Hype vs. Reality (Re: Knight Rider (2025/2026) will kill us all) Mild Shock <janburse@fastmail.fm> - 2025-11-13 21:58 +0100
        Stable Diffusion with Control Net (Re: The AGI Gambit: Hype vs. Reality) Mild Shock <janburse@fastmail.fm> - 2025-11-13 23:05 +0100

#14878 — Vanishing Point . [Le Dot]

In a Tibetan lamasery, the monks seek to
list all of the names of God. They believe
the Universe was created for this purpose,
and that once this naming is completed, God
will bring the Universe to an end. Three

centuries ago, the monks created an alphabet
in which they calculated they could encode
all the possible names of God, numbering
about 9,000,000,000 ("nine billion") and
each having no more than nine characters.

riting the names out by hand, as they had
been doing, even after eliminating various
nonsense combinations, would take another
15,000 years; the monks wish to use modern
technology to finish this task in 100 days.

They rent a computer capable of printing all
the possible permutations, and hire two
Westerners to install and program the machine.
The computer operators are skeptical but
play along. After three months, as the job

nears completion, they fear that the monks
will blame the computer (and, by extension,
its operators) when nothing happens. The
Westerners leave slightly earlier than their
scheduled departure without warning the monks,

so that it will complete its final print run
shortly after they leave. On their way to the
airfield they pause on the mountain path. Under
a clear night sky they es timate that it must be
just about the time that the monks are pasting

the final printed names into their holy books.
Then they notice that "overhead, without any
fuss, the stars were going out."

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

[toc] | [next] | [standalone]

#14879 — Lets compare AI data centers (Was: Vanishing Point . [Le Dot])

Hi,

Interestingly there is also now a top-ten
for AI data centers, not only super computers.
We are talking about newly built AI data centers

that for the first time go into giga watts:

The New World’s Largest AI Supercluster
https://www.youtube.com/watch?v=RxuSvyOwVCI

Some confirmed picks:

xAI Colossus Memphis
Phase 1	100000	xAI	U.S.	Confirmed
Tesla Cortex Phase 1	50000	Tesla	U.S.	Confirmed
Lawrence Livermore NL
El Capitan Phase 2	44143	U.S. Department of
Energy	U.S.	Confirmed
Anonymized Chinese
System	30000	N/A	China	Confirmed
Meta GenAI 2024a	24576	Meta AI	U.S.	Confirmed
Meta GenAI 2024b	24576	Meta AI	U.S.	Confirmed
Jupiter, Jülich	23536	EuroHPC JU,
Jülich Supercomputing
Center	Germany	Confirmed
Anonymized Chinese
System	20000	N/A	China	Confirmed
Anonymized Chinese
System	20000	N/A	China	Confirmed

https://www.visualcapitalist.com/the-worlds-most-powerful-ai-supercomputers/

Bye

Mild Shock schrieb:
> 
> In a Tibetan lamasery, the monks seek to
> list all of the names of God. They believe
> the Universe was created for this purpose,
> and that once this naming is completed, God
> will bring the Universe to an end. Three
> 
> centuries ago, the monks created an alphabet
> in which they calculated they could encode
> all the possible names of God, numbering
> about 9,000,000,000 ("nine billion") and
> each having no more than nine characters.
> 
> riting the names out by hand, as they had
> been doing, even after eliminating various
> nonsense combinations, would take another
> 15,000 years; the monks wish to use modern
> technology to finish this task in 100 days.
> 
> They rent a computer capable of printing all
> the possible permutations, and hire two
> Westerners to install and program the machine.
> The computer operators are skeptical but
> play along. After three months, as the job
> 
> nears completion, they fear that the monks
> will blame the computer (and, by extension,
> its operators) when nothing happens. The
> Westerners leave slightly earlier than their
> scheduled departure without warning the monks,
> 
> so that it will complete its final print run
> shortly after they leave. On their way to the
> airfield they pause on the mountain path. Under
> a clear night sky they es timate that it must be
> just about the time that the monks are pasting
> 
> the final printed names into their holy books.
> Then they notice that "overhead, without any
> fuss, the stars were going out."
> 
> https://en.wikipedia.org/wiki/The_Nine_Billion_Names_of_God

[toc] | [prev] | [next] | [standalone]

#14880 — USA domestic chips for the never ending crunch (Re: Lets compare AI data centers )

Hi,

So the Tibetan lamasery had it right,
the lights go out. There is a never ending
hunger for crunch? Even requiring USA

to go into domestic chip production:

America’s Most Advanced Chip Factory Yet
https://www.youtube.com/watch?v=1VX3jNJmbcI

But chip production etc.. (embodied energy) is
only a fraction of the energy consumption:

 > Operational energy (running the data center
 > for 1 year at 1 GW) = 8,760 TWh Hardware
 > manufacturing energy = 0.75 TWh
 > So the energy to produce the hardware is
 > roughly 0.0085 (or 0.85%) of the annual
 > operational energy.

What are the projections? By 2030, global
power demand for data centers is projected to
reach approximately 220 GW, underscoring the

urgency and strategic importance of securing
power and hardware capacity:

$100B Bet on 10GW AI Infrastructure
https://51ai.substack.com/p/openai-nvidia-100b-bet-on-10gw-ai

Bye

Mild Shock schrieb:
> Hi,
> 
> Interestingly there is also now a top-ten
> for AI data centers, not only super computers.
> We are talking about newly built AI data centers
> 
> that for the first time go into giga watts:
> 
> The New World’s Largest AI Supercluster
> https://www.youtube.com/watch?v=RxuSvyOwVCI
> 
> Some confirmed picks:
> 
> xAI Colossus Memphis
> Phase 1    100000    xAI    U.S.    Confirmed
> Tesla Cortex Phase 1    50000    Tesla    U.S.    Confirmed
> Lawrence Livermore NL
> El Capitan Phase 2    44143    U.S. Department of
> Energy    U.S.    Confirmed
> Anonymized Chinese
> System    30000    N/A    China    Confirmed
> Meta GenAI 2024a    24576    Meta AI    U.S.    Confirmed
> Meta GenAI 2024b    24576    Meta AI    U.S.    Confirmed
> Jupiter, Jülich    23536    EuroHPC JU,
> Jülich Supercomputing
> Center    Germany    Confirmed
> Anonymized Chinese
> System    20000    N/A    China    Confirmed
> Anonymized Chinese
> System    20000    N/A    China    Confirmed
> 
> https://www.visualcapitalist.com/the-worlds-most-powerful-ai-supercomputers/ 
> 
> 
> Bye
> 
> Mild Shock schrieb:
>>
>> In a Tibetan lamasery, the monks seek to
>> list all of the names of God. They believe
>> the Universe was created for this purpose,
>> and that once this naming is completed, God
>> will bring the Universe to an end. Three
>>
>> centuries ago, the monks created an alphabet
>> in which they calculated they could encode
>> all the possible names of God, numbering
>> about 9,000,000,000 ("nine billion") and
>> each having no more than nine characters.
>>
>> riting the names out by hand, as they had
>> been doing, even after eliminating various
>> nonsense combinations, would take another
>> 15,000 years; the monks wish to use modern
>> technology to finish this task in 100 days.
>>
>> They rent a computer capable of printing all
>> the possible permutations, and hire two
>> Westerners to install and program the machine.
>> The computer operators are skeptical but
>> play along. After three months, as the job
>>
>> nears completion, they fear that the monks
>> will blame the computer (and, by extension,
>> its operators) when nothing happens. The
>> Westerners leave slightly earlier than their
>> scheduled departure without warning the monks,
>>
>> so that it will complete its final print run
>> shortly after they leave. On their way to the
>> airfield they pause on the mountain path. Under
>> a clear night sky they es timate that it must be
>> just about the time that the monks are pasting
>>
>> the final printed names into their holy books.
>> Then they notice that "overhead, without any
>> fuss, the stars were going out."
>>
>> https://en.wikipedia.org/wiki/The_Nine_Billion_Names_of_God
> 

[toc] | [prev] | [next] | [standalone]

#14881 — Holy Schrimp Jesus: The Internet is Dead [Rossy Boy dethroned?] (Re: USA domestic chips for the never ending crunch)

Hi,

Funny video that uses KI generated content
itself, to decry KI generated content. We
live in a paradox world:

"While we are still thinking about the
possible social effects of artificial
intelligence, the digital knowledge space
is already drowning into synthetic trash.

How could it get that far?  How could the
network, which had not been long ago as a
place of free knowledge and the open exchange
of information and entertainment, became
mechanically manufactured nonsensively
in record speed?"
KI: Der Tod des Internets - ARTE
https://www.youtube.com/watch?v=cGmVehWBdHI

Word of the year 2025:

"AI slop has been variously defined as "digital
clutter", "filler content [prioritizing] speed
and quantity over substance and quality",[6]
and "shoddy or unwanted AI content in social
media, art, books and [...] search results".
AI slop
https://en.wikipedia.org/wiki/AI_slop

Poor Rossy Boy, now he is completely jobless.
In the past he was the sole king of Human Slop.
Now AI Slop thanks to large scale remix, stable

diffusion and who knows what, is the new spam king.

Bye

Mild Shock schrieb:
> Hi,
> 
> So the Tibetan lamasery had it right,
> the lights go out. There is a never ending
> hunger for crunch? Even requiring USA
> 
> to go into domestic chip production:
> 
> America’s Most Advanced Chip Factory Yet
> https://www.youtube.com/watch?v=1VX3jNJmbcI
> 
> But chip production etc.. (embodied energy) is
> only a fraction of the energy consumption:
> 
>  > Operational energy (running the data center
>  > for 1 year at 1 GW) = 8,760 TWh Hardware
>  > manufacturing energy = 0.75 TWh
>  > So the energy to produce the hardware is
>  > roughly 0.0085 (or 0.85%) of the annual
>  > operational energy.
> 
> What are the projections? By 2030, global
> power demand for data centers is projected to
> reach approximately 220 GW, underscoring the
> 
> urgency and strategic importance of securing
> power and hardware capacity:
> 
> $100B Bet on 10GW AI Infrastructure
> https://51ai.substack.com/p/openai-nvidia-100b-bet-on-10gw-ai
> 
> Bye
> 
> Mild Shock schrieb:
>> Hi,
>>
>> Interestingly there is also now a top-ten
>> for AI data centers, not only super computers.
>> We are talking about newly built AI data centers
>>
>> that for the first time go into giga watts:
>>
>> The New World’s Largest AI Supercluster
>> https://www.youtube.com/watch?v=RxuSvyOwVCI
>>
>> Some confirmed picks:
>>
>> xAI Colossus Memphis
>> Phase 1    100000    xAI    U.S.    Confirmed
>> Tesla Cortex Phase 1    50000    Tesla    U.S.    Confirmed
>> Lawrence Livermore NL
>> El Capitan Phase 2    44143    U.S. Department of
>> Energy    U.S.    Confirmed
>> Anonymized Chinese
>> System    30000    N/A    China    Confirmed
>> Meta GenAI 2024a    24576    Meta AI    U.S.    Confirmed
>> Meta GenAI 2024b    24576    Meta AI    U.S.    Confirmed
>> Jupiter, Jülich    23536    EuroHPC JU,
>> Jülich Supercomputing
>> Center    Germany    Confirmed
>> Anonymized Chinese
>> System    20000    N/A    China    Confirmed
>> Anonymized Chinese
>> System    20000    N/A    China    Confirmed
>>
>> https://www.visualcapitalist.com/the-worlds-most-powerful-ai-supercomputers/ 
>>
>>
>> Bye
>>
>> Mild Shock schrieb:
>>>
>>> In a Tibetan lamasery, the monks seek to
>>> list all of the names of God. They believe
>>> the Universe was created for this purpose,
>>> and that once this naming is completed, God
>>> will bring the Universe to an end. Three
>>>
>>> centuries ago, the monks created an alphabet
>>> in which they calculated they could encode
>>> all the possible names of God, numbering
>>> about 9,000,000,000 ("nine billion") and
>>> each having no more than nine characters.
>>>
>>> riting the names out by hand, as they had
>>> been doing, even after eliminating various
>>> nonsense combinations, would take another
>>> 15,000 years; the monks wish to use modern
>>> technology to finish this task in 100 days.
>>>
>>> They rent a computer capable of printing all
>>> the possible permutations, and hire two
>>> Westerners to install and program the machine.
>>> The computer operators are skeptical but
>>> play along. After three months, as the job
>>>
>>> nears completion, they fear that the monks
>>> will blame the computer (and, by extension,
>>> its operators) when nothing happens. The
>>> Westerners leave slightly earlier than their
>>> scheduled departure without warning the monks,
>>>
>>> so that it will complete its final print run
>>> shortly after they leave. On their way to the
>>> airfield they pause on the mountain path. Under
>>> a clear night sky they es timate that it must be
>>> just about the time that the monks are pasting
>>>
>>> the final printed names into their holy books.
>>> Then they notice that "overhead, without any
>>> fuss, the stars were going out."
>>>
>>> https://en.wikipedia.org/wiki/The_Nine_Billion_Names_of_God
>>
> 

[toc] | [prev] | [next] | [standalone]

#14886 — Try https://mathstodon.xyz by Christian Lawson-Perfect (Re: Holy Schrimp Jesus: The Internet is Dead [Rossy Boy dethroned?] )

Hi,

Try this server:

https://mathstodon.xyz/about?lang=en

It says:

 > Mathstodon.xyz is a Mastodon instance for
 > people who love maths. This instance is for
 > people who love maths and we hope there’ll be
 > lots of maths chat, but any topic of conversation
 > following the code of conduct and the principle
 > of getting along together is OK.

It also says:

 > Administered by:
 > Christian Lawson-Perfect
 > @christianp

Bye

P.S.: Sorry Julio Di Egidio, might push your Nazi
Retard buttons again. Mastodon was created and is
led by Eugen Rochko, a German software developer.

Ross Finlayson schrieb:
 > On 03/12/2024 10:08 AM, Ross Finlayson wrote:
 >> On 03/07/2024 08:09 AM, Ross Finlayson wrote:
 >>> On 02/29/2024 07:55 PM, Ross Finlayson wrote:
 >>>> On 02/20/2024 07:47 PM, Ross Finlayson wrote:
 >>>>> About a "dedicated little OS" to run a "dedicated little service".
 >>>>> [... tons of gibberish ...]

[toc] | [prev] | [next] | [standalone]

#14887 — Some statistics about Mastodon (Re: Try https://mathstodon.xyz by Christian Lawson-Perfect)

Hi,

Some statistics (2019 pre Corona):

Number of instances by country:
- Japan hosts ~ 25.5% of all Mastodon instances.
- The United States hosts ~ 21.4% of instances.
- France, Germany etc. also have significant shares,
   but less than those top two.
Tweets are caled Toots:

Users / Content concentration (“where the users / toots are”):
- Even though Japan has ~25.5% of instances,
   it accounts for about 41% of users.
- Many “toots” (posts) are concentrated in a few countries.
   For example, the top three countries host a majority of toots.

See also:

“Challenges in the Decentralised Web: The Mastodon Case.”
De Domenico et al. (2019),
https://www.researchgate.net/publication/336661890

Bye

Mild Shock schrieb:
> Hi,
> 
> Try this server:
> 
> https://mathstodon.xyz/about?lang=en
> 
> It says:
> 
>  > Mathstodon.xyz is a Mastodon instance for
>  > people who love maths. This instance is for
>  > people who love maths and we hope there’ll be
>  > lots of maths chat, but any topic of conversation
>  > following the code of conduct and the principle
>  > of getting along together is OK.
> 
> It also says:
> 
>  > Administered by:
>  > Christian Lawson-Perfect
>  > @christianp
> 
> Bye
> 
> P.S.: Sorry Julio Di Egidio, might push your Nazi
> Retard buttons again. Mastodon was created and is
> led by Eugen Rochko, a German software developer.
> 
> Ross Finlayson schrieb:
>  > On 03/12/2024 10:08 AM, Ross Finlayson wrote:
>  >> On 03/07/2024 08:09 AM, Ross Finlayson wrote:
>  >>> On 02/29/2024 07:55 PM, Ross Finlayson wrote:
>  >>>> On 02/20/2024 07:47 PM, Ross Finlayson wrote:
>  >>>>> About a "dedicated little OS" to run a "dedicated little service".
>  >>>>> [... tons of gibberish ...]

[toc] | [prev] | [next] | [standalone]

#14983 — The joy of running an AI Model on your AI Laptop [Local AI] (Re: USA domestic chips for the never ending crunch)

Hi,

Now that was a little unexpected that Intel beats AMD.
Could get my hands on two AI Laptops for testing.
The have both DDR5X Ram.

AI Laptop 1: Acer_Swift_SFG16-61
AMD Ryzen 7 350
Radeon Graphics
AMD Ryzen AI Boost, up to 50 TOPS

AI Laptop 2: LENOVO_83JQ
Intel Core Ultra 7 258V
Intel Arc Graphics 140V
Intel AI Boost, up to 47 TOPS

I couldn't test the AI Boost yet. Besides the nice
CPUs, the iGPU and the fast memory, in this case
32 GB, seem to give quite some bang:

Here some test with LMStudio, I picked the
model Qwen3 Coder 30B, with Q4_K_M encoding,
which is around 16 GB, and fits into my GPU memory:
https://lmstudio.ai/models/qwen/qwen3-coder-30b

AI Laptop 1: 18 Tokens/s

AI Laptop 2: 6 Tokens/s

Now I am waiting for drivers and applications
that make use of the AI Boost. Will say
I only tested the iGPU, not yet the NPU.

For the donwload of the AI model it is
required to have fiber and WiFi 6, you can
then make it in less than a minute, even

if you are far away from the router.

Bye

P.S.: The 18 Tokens/s are quite fluid,
the 6 Tokens/s are a little annoying. But
on both machines I could converse with the

model, and go from primes to twin primes.
It didn't use any RAG, fully produce via
Local AI, also no remote AI access after

download of the AI Model:

----------------- cut here ---------------------

Here's a Prolog implementation to find
twin prime pairs (pairs of primes that differ by 2):

% Generate twin prime pairs up to N
twin_primes_up_to(N, TwinPairs) :-
     findall((P, Q),
             (between(3, N, P),
              is_prime(P),
              Q is P + 2,
              is_prime(Q)),
             TwinPairs).

% Check if a number is prime
is_prime(2) :- !.
is_prime(3) :- !.
is_prime(N) :-
     N > 3,
     N mod 2 =\= 0,          % Even numbers > 2 are not prime
     \+ has_factor(N, 3).    % Check if N has any factors starting from 3

% Check if N has a factor starting from F
has_factor(N, F) :-
     F * F =< N,
     (N mod F =:= 0 ; has_factor(N, F + 2)).

Example usage:

?- twin_primes_up_to(20, Pairs).
Pairs = [(3, 5), (5, 7), (11, 13), (17, 19)].

?- twin_primes_up_to(50, Pairs).
Pairs = [(3, 5), (5, 7), (11, 13), (17, 19), (29, 31), (41, 43)].

This finds pairs of primes (P, Q) where Q = P + 2,
which are called twin primes. The findall/3 predicate
collects all such pairs where both P and P+2 are
prime numbers and P is less than or equal to N.

----------------- cut here ---------------------

Mild Shock schrieb:
> Hi,
> 
> So the Tibetan lamasery had it right,
> the lights go out. There is a never ending
> hunger for crunch? Even requiring USA
> 
> to go into domestic chip production:
> 
> America’s Most Advanced Chip Factory Yet
> https://www.youtube.com/watch?v=1VX3jNJmbcI
> 
> But chip production etc.. (embodied energy) is
> only a fraction of the energy consumption:
> 
>  > Operational energy (running the data center
>  > for 1 year at 1 GW) = 8,760 TWh Hardware
>  > manufacturing energy = 0.75 TWh
>  > So the energy to produce the hardware is
>  > roughly 0.0085 (or 0.85%) of the annual
>  > operational energy.
> 
> What are the projections? By 2030, global
> power demand for data centers is projected to
> reach approximately 220 GW, underscoring the
> 
> urgency and strategic importance of securing
> power and hardware capacity:
> 
> $100B Bet on 10GW AI Infrastructure
> https://51ai.substack.com/p/openai-nvidia-100b-bet-on-10gw-ai
> 
> Bye
> 
> Mild Shock schrieb:
>> Hi,
>>
>> Interestingly there is also now a top-ten
>> for AI data centers, not only super computers.
>> We are talking about newly built AI data centers
>>
>> that for the first time go into giga watts:
>>
>> The New World’s Largest AI Supercluster
>> https://www.youtube.com/watch?v=RxuSvyOwVCI
>>
>> Some confirmed picks:
>>
>> xAI Colossus Memphis
>> Phase 1    100000    xAI    U.S.    Confirmed
>> Tesla Cortex Phase 1    50000    Tesla    U.S.    Confirmed
>> Lawrence Livermore NL
>> El Capitan Phase 2    44143    U.S. Department of
>> Energy    U.S.    Confirmed
>> Anonymized Chinese
>> System    30000    N/A    China    Confirmed
>> Meta GenAI 2024a    24576    Meta AI    U.S.    Confirmed
>> Meta GenAI 2024b    24576    Meta AI    U.S.    Confirmed
>> Jupiter, Jülich    23536    EuroHPC JU,
>> Jülich Supercomputing
>> Center    Germany    Confirmed
>> Anonymized Chinese
>> System    20000    N/A    China    Confirmed
>> Anonymized Chinese
>> System    20000    N/A    China    Confirmed
>>
>> https://www.visualcapitalist.com/the-worlds-most-powerful-ai-supercomputers/ 
>>
>>
>> Bye
>>
>> Mild Shock schrieb:
>>>
>>> In a Tibetan lamasery, the monks seek to
>>> list all of the names of God. They believe
>>> the Universe was created for this purpose,
>>> and that once this naming is completed, God
>>> will bring the Universe to an end. Three
>>>
>>> centuries ago, the monks created an alphabet
>>> in which they calculated they could encode
>>> all the possible names of God, numbering
>>> about 9,000,000,000 ("nine billion") and
>>> each having no more than nine characters.
>>>
>>> riting the names out by hand, as they had
>>> been doing, even after eliminating various
>>> nonsense combinations, would take another
>>> 15,000 years; the monks wish to use modern
>>> technology to finish this task in 100 days.
>>>
>>> They rent a computer capable of printing all
>>> the possible permutations, and hire two
>>> Westerners to install and program the machine.
>>> The computer operators are skeptical but
>>> play along. After three months, as the job
>>>
>>> nears completion, they fear that the monks
>>> will blame the computer (and, by extension,
>>> its operators) when nothing happens. The
>>> Westerners leave slightly earlier than their
>>> scheduled departure without warning the monks,
>>>
>>> so that it will complete its final print run
>>> shortly after they leave. On their way to the
>>> airfield they pause on the mountain path. Under
>>> a clear night sky they es timate that it must be
>>> just about the time that the monks are pasting
>>>
>>> the final printed names into their holy books.
>>> Then they notice that "overhead, without any
>>> fuss, the stars were going out."
>>>
>>> https://en.wikipedia.org/wiki/The_Nine_Billion_Names_of_God
>>
> 

[toc] | [prev] | [next] | [standalone]

#14984 — Cute Racoon generated with Local AI (Re: The joy of running an AI Model on your AI Laptop [Local AI])

Hi,

There is even more joy of running an AI Model
on your AI Laptop. All you need to put "AI Slob"
into the wild is Stable Diffusion Software and Models.

Cute Racoon generated with Local AI
https://www.youtube.com/shorts/QlS0MqMh_bM

Have Fun!

Bye

P.S.: Was using this gadget:

Create AI art, with the latest Stable
Diffusion models, on your own PC.
https://www.amuse-ai.com/

Mild Shock schrieb:
> Hi,
> 
> Now that was a little unexpected that Intel beats AMD.
> Could get my hands on two AI Laptops for testing.
> The have both DDR5X Ram.
> 
> AI Laptop 1: Acer_Swift_SFG16-61
> AMD Ryzen 7 350
> Radeon Graphics
> AMD Ryzen AI Boost, up to 50 TOPS
> 
> AI Laptop 2: LENOVO_83JQ
> Intel Core Ultra 7 258V
> Intel Arc Graphics 140V
> Intel AI Boost, up to 47 TOPS
> 
> I couldn't test the AI Boost yet. Besides the nice
> CPUs, the iGPU and the fast memory, in this case
> 32 GB, seem to give quite some bang:
> 
> Here some test with LMStudio, I picked the
> model Qwen3 Coder 30B, with Q4_K_M encoding,
> which is around 16 GB, and fits into my GPU memory:
> https://lmstudio.ai/models/qwen/qwen3-coder-30b
> 
> AI Laptop 1: 18 Tokens/s
> 
> AI Laptop 2: 6 Tokens/s
> 
> Now I am waiting for drivers and applications
> that make use of the AI Boost. Will say
> I only tested the iGPU, not yet the NPU.
> 
> For the donwload of the AI model it is
> required to have fiber and WiFi 6, you can
> then make it in less than a minute, even
> 
> if you are far away from the router.
> 
> Bye
> 
> P.S.: The 18 Tokens/s are quite fluid,
> the 6 Tokens/s are a little annoying. But
> on both machines I could converse with the
> 
> model, and go from primes to twin primes.
> It didn't use any RAG, fully produce via
> Local AI, also no remote AI access after
> 
> download of the AI Model:
> 
> ----------------- cut here ---------------------
> 
> Here's a Prolog implementation to find
> twin prime pairs (pairs of primes that differ by 2):
> 
> % Generate twin prime pairs up to N
> twin_primes_up_to(N, TwinPairs) :-
>      findall((P, Q),
>              (between(3, N, P),
>               is_prime(P),
>               Q is P + 2,
>               is_prime(Q)),
>              TwinPairs).
> 
> % Check if a number is prime
> is_prime(2) :- !.
> is_prime(3) :- !.
> is_prime(N) :-
>      N > 3,
>      N mod 2 =\= 0,          % Even numbers > 2 are not prime
>      \+ has_factor(N, 3).    % Check if N has any factors starting from 3
> 
> % Check if N has a factor starting from F
> has_factor(N, F) :-
>      F * F =< N,
>      (N mod F =:= 0 ; has_factor(N, F + 2)).
> 
> Example usage:
> 
> ?- twin_primes_up_to(20, Pairs).
> Pairs = [(3, 5), (5, 7), (11, 13), (17, 19)].
> 
> ?- twin_primes_up_to(50, Pairs).
> Pairs = [(3, 5), (5, 7), (11, 13), (17, 19), (29, 31), (41, 43)].
> 
> This finds pairs of primes (P, Q) where Q = P + 2,
> which are called twin primes. The findall/3 predicate
> collects all such pairs where both P and P+2 are
> prime numbers and P is less than or equal to N.
> 
> ----------------- cut here ---------------------

[toc] | [prev] | [next] | [standalone]

#14985 — What about Local AI? [Archimedes Plutonium AI Persona?] (Re: Cute Racoon generated with Local AI)

Hi,

Come on old boy, why not try Local AI?
Elaborate your paleo stone throwing theory
with a Local AI (An AI Laptop and AI Models).

Actually I think they should train an AI
model on Archimedes Plutonium. And then
use it as a USENET AI Slob geberator.

An AI Slob generator that will be fed
with the newest news about Trump, Putin,
Math, etc.. in its conext window,

Instructed do give AP style reflections.

LoL

Bye

Ross Finlayson schrieb:
 > The status of excluded middle vs. inverse
 > The "sufficient and thorough reason" vs. traditional PSR
 > The "paleo-classical post-modern" character
 > The synthesis of traditions (Pythagorean, Cantorian, etc.)
 > The de-fragmentation claim

[toc] | [prev] | [next] | [standalone]

#14991 — Ross Finlayson is completely Paranoid (Re: What about Local AI? [Archimedes Plutonium AI Persona?])

Hi,

If you take identity as E-mail address, which you
should see with a suitable news reader. I don't
use randomly generated E-mail adresses, or invalid

E-mail address. Its always the same E-mail adress.
At the time of Google+ it was maybe a different E-mail
address to access Google+. So its always the same

"identity". What changes is my nickname sometimes.
But not from post to post, I am not a nickname shapeshifter,
I only change it every decade or so. So you can take

your false accusations, and shove it up your ass.

Fucking paranoid moron you are.

Bye

Ross Finlayson schrieb:
 > Then, since for example there's nothing funny about it,
 > I'm not interested in your plan of mal-journalism.
 > Furthermore all your sockpuppets are attached to
 > your identity.

Mild Shock schrieb:
> Hi,
> 
> Come on old boy, why not try Local AI?
> Elaborate your paleo stone throwing theory
> with a Local AI (An AI Laptop and AI Models).
> 
> Actually I think they should train an AI
> model on Archimedes Plutonium. And then
> use it as a USENET AI Slob geberator.
> 
> An AI Slob generator that will be fed
> with the newest news about Trump, Putin,
> Math, etc.. in its conext window,
> 
> Instructed do give AP style reflections.
> 
> LoL
> 
> Bye
> 
> Ross Finlayson schrieb:
>  > The status of excluded middle vs. inverse
>  > The "sufficient and thorough reason" vs. traditional PSR
>  > The "paleo-classical post-modern" character
>  > The synthesis of traditions (Pythagorean, Cantorian, etc.)
>  > The de-fragmentation claim

[toc] | [prev] | [next] | [standalone]

#14992 — How to make an AI Persona [With Local AI?] (Re: Ross Finlayson is completely Paranoid)

Hi,

There are different ways to make an AI Persona.
For example Twitter aka X, has a grok account.
But grok is not trimmed towards a particular

profile, maybe an alter ego of Elon Musk was
speculated. What could be interesting would be
to go towards character.ai :

"Character.ai was established in November 2021.
The company's co-founders, Noam Shazeer and
Daniel de Freitas, were both engineers from
Google. They both worked on AI-related projects:
Shazeer was a lead author on a paper that
Business Insider reported in April 2023 "has
been widely cited as key to today's chatbots",
and Freitas was the lead designer of an
experimental AI at Google initially called Meena,
which later became known as LaMDA.
https://en.wikipedia.org/wiki/Character.ai

Maybe this can be even done with Local AI. By for
example populating a context. Or creating a kind
of LangChain memory. ChatGPT calls it persistent context.

So yes you could "spam" an AI Archimedes Plutonium
probably from your desktop with an AI Laptop, maybe
already in 2025, or later in 2026.

Bye

Mild Shock schrieb:
> Hi,
> 
> If you take identity as E-mail address, which you
> should see with a suitable news reader. I don't
> use randomly generated E-mail adresses, or invalid
> 
> E-mail address. Its always the same E-mail adress.
> At the time of Google+ it was maybe a different E-mail
> address to access Google+. So its always the same
> 
> "identity". What changes is my nickname sometimes.
> But not from post to post, I am not a nickname shapeshifter,
> I only change it every decade or so. So you can take
> 
> your false accusations, and shove it up your ass.
> 
> Fucking paranoid moron you are.
> 
> Bye
> 
> Ross Finlayson schrieb:
>  > Then, since for example there's nothing funny about it,
>  > I'm not interested in your plan of mal-journalism.
>  > Furthermore all your sockpuppets are attached to
>  > your identity.
> 
> Mild Shock schrieb:
>> Hi,
>>
>> Come on old boy, why not try Local AI?
>> Elaborate your paleo stone throwing theory
>> with a Local AI (An AI Laptop and AI Models).
>>
>> Actually I think they should train an AI
>> model on Archimedes Plutonium. And then
>> use it as a USENET AI Slob geberator.
>>
>> An AI Slob generator that will be fed
>> with the newest news about Trump, Putin,
>> Math, etc.. in its conext window,
>>
>> Instructed do give AP style reflections.
>>
>> LoL
>>
>> Bye
>>
>> Ross Finlayson schrieb:
>>  > The status of excluded middle vs. inverse
>>  > The "sufficient and thorough reason" vs. traditional PSR
>>  > The "paleo-classical post-modern" character
>>  > The synthesis of traditions (Pythagorean, Cantorian, etc.)
>>  > The de-fragmentation claim
> 

[toc] | [prev] | [next] | [standalone]

#14882 — British writer Arthur C. Clarke (Re: Vanishing Point . [Le Dot])

I dunno. Was trying to find another story,
of a scientiest who studies the mind, and
then goes slowly crazy when he discovers

how the mind works. But what I posted
is a **Plot Summmary** of a short story:

 > "The Nine Billion Names of God" is a
 > 1953 science fiction short story by
 > British writer Arthur C. Clarke.
 > https://en.wikipedia.org/wiki/The_Nine_Billion_Names_of_God

Maybe he used some existing Asian lore,
I don't know. Who is an expert in this matter?

WM schrieb:
 > On 05.10.2025 00:52, Mild Shock wrote:
 >>
 >> In a Tibetan lamasery, the monks seek to
 >> list all of the names of God. They believe
 >> the Universe was created for this purpose,
 >> and that once this naming is completed, God
 >> will bring the Universe to an end. Three
 >
 > A very old story. I read it as a child.
 >
 > Regards, WM

Mild Shock schrieb:
> 
> In a Tibetan lamasery, the monks seek to
> list all of the names of God. They believe
> the Universe was created for this purpose,
> and that once this naming is completed, God
> will bring the Universe to an end. Three
> 
> centuries ago, the monks created an alphabet
> in which they calculated they could encode
> all the possible names of God, numbering
> about 9,000,000,000 ("nine billion") and
> each having no more than nine characters.
> 
> riting the names out by hand, as they had
> been doing, even after eliminating various
> nonsense combinations, would take another
> 15,000 years; the monks wish to use modern
> technology to finish this task in 100 days.
> 
> They rent a computer capable of printing all
> the possible permutations, and hire two
> Westerners to install and program the machine.
> The computer operators are skeptical but
> play along. After three months, as the job
> 
> nears completion, they fear that the monks
> will blame the computer (and, by extension,
> its operators) when nothing happens. The
> Westerners leave slightly earlier than their
> scheduled departure without warning the monks,
> 
> so that it will complete its final print run
> shortly after they leave. On their way to the
> airfield they pause on the mountain path. Under
> a clear night sky they es timate that it must be
> just about the time that the monks are pasting
> 
> the final printed names into their holy books.
> Then they notice that "overhead, without any
> fuss, the stars were going out."
> 
> https://en.wikipedia.org/wiki/The_Nine_Billion_Names_of_God

[toc] | [prev] | [next] | [standalone]

#14993 — Could the AI Persona evolve over time? [With Local AI?] (Re: British writer Arthur C. Clarke)

Hi,

With RAG it could react on news. RAG (Retrieval-
Augmented Generation) and other techniques can
make a local AI evolve dynamically — both in knowledge

and personality. Lets say I put such a AI Persona in
a News Group loop. I could feedback the posts there,
to make it also evolve. One can collect community

reactions (upvotes, replies, sentiment). Or apply automatic
evaluation — e.g., sentiment analysis or engagement
score. This becomes your fitness function.

You can even run multiple personas in the same loop
Different viewpoints (e.g., “optimist”, “skeptic”,
“data scientist”) They interact and debate posts.

This is ideal for dead open source projects, such as
SWI-Prolog, to pretend user participation, very important
in our new marketing world that is totally

engagement focus. The quality(*) of content doesn't count.

LoL

Bye

(*) Same for research paper mills. BTW, ChatGPT suggests
me the following tech stack:

LLM: Ollama (LLaMA 3, Mistral, Gemma, etc.)
LangChain / LlamaIndex: context + memory
ChromaDB / SQLite: memory store
Feedparser / API client: ingest news or group posts
Transformers / VADER: sentiment scoring
Simple scheduler or cron job: to run the loop daily

Mild Shock schrieb:
> 
> I dunno. Was trying to find another story,
> of a scientiest who studies the mind, and
> then goes slowly crazy when he discovers
> 
> how the mind works. But what I posted
> is a **Plot Summmary** of a short story:
> 
>  > "The Nine Billion Names of God" is a
>  > 1953 science fiction short story by
>  > British writer Arthur C. Clarke.
>  > https://en.wikipedia.org/wiki/The_Nine_Billion_Names_of_God
> 
> Maybe he used some existing Asian lore,
> I don't know. Who is an expert in this matter?
> 
> WM schrieb:
>  > On 05.10.2025 00:52, Mild Shock wrote:
>  >>
>  >> In a Tibetan lamasery, the monks seek to
>  >> list all of the names of God. They believe
>  >> the Universe was created for this purpose,
>  >> and that once this naming is completed, God
>  >> will bring the Universe to an end. Three
>  >
>  > A very old story. I read it as a child.
>  >
>  > Regards, WM
> 
> Mild Shock schrieb:
>>
>> In a Tibetan lamasery, the monks seek to
>> list all of the names of God. They believe
>> the Universe was created for this purpose,
>> and that once this naming is completed, God
>> will bring the Universe to an end. Three
>>
>> centuries ago, the monks created an alphabet
>> in which they calculated they could encode
>> all the possible names of God, numbering
>> about 9,000,000,000 ("nine billion") and
>> each having no more than nine characters.
>>
>> riting the names out by hand, as they had
>> been doing, even after eliminating various
>> nonsense combinations, would take another
>> 15,000 years; the monks wish to use modern
>> technology to finish this task in 100 days.
>>
>> They rent a computer capable of printing all
>> the possible permutations, and hire two
>> Westerners to install and program the machine.
>> The computer operators are skeptical but
>> play along. After three months, as the job
>>
>> nears completion, they fear that the monks
>> will blame the computer (and, by extension,
>> its operators) when nothing happens. The
>> Westerners leave slightly earlier than their
>> scheduled departure without warning the monks,
>>
>> so that it will complete its final print run
>> shortly after they leave. On their way to the
>> airfield they pause on the mountain path. Under
>> a clear night sky they es timate that it must be
>> just about the time that the monks are pasting
>>
>> the final printed names into their holy books.
>> Then they notice that "overhead, without any
>> fuss, the stars were going out."
>>
>> https://en.wikipedia.org/wiki/The_Nine_Billion_Names_of_God
> 

[toc] | [prev] | [next] | [standalone]

#14994 — Boris the Loris and Nazi-Retard Julio (Re: Could the AI Persona evolve over time? [With Local AI?])

Hi,

My favorite AI personas would be:

- Boris the Loris
   Agent provocateur that keeps the quality
   of posts low, only dumb open source consumers
   allowed, no Fuzzy Logic, etc..

- Nazi-Retard Julio
   Agent provocateur that pretends to know
   some subject, but doesn't know it at all,
   this is to compensate for the dump open source
   consumers, that populate the group now.

Bye

Mild Shock schrieb:
> Hi,
> 
> With RAG it could react on news. RAG (Retrieval-
> Augmented Generation) and other techniques can
> make a local AI evolve dynamically — both in knowledge
> 
> and personality. Lets say I put such a AI Persona in
> a News Group loop. I could feedback the posts there,
> to make it also evolve. One can collect community
> 
> reactions (upvotes, replies, sentiment). Or apply automatic
> evaluation — e.g., sentiment analysis or engagement
> score. This becomes your fitness function.
> 
> You can even run multiple personas in the same loop
> Different viewpoints (e.g., “optimist”, “skeptic”,
> “data scientist”) They interact and debate posts.
> 
> This is ideal for dead open source projects, such as
> SWI-Prolog, to pretend user participation, very important
> in our new marketing world that is totally
> 
> engagement focus. The quality(*) of content doesn't count.
> 
> LoL
> 
> Bye
> 
> (*) Same for research paper mills. BTW, ChatGPT suggests
> me the following tech stack:
> 
> LLM: Ollama (LLaMA 3, Mistral, Gemma, etc.)
> LangChain / LlamaIndex: context + memory
> ChromaDB / SQLite: memory store
> Feedparser / API client: ingest news or group posts
> Transformers / VADER: sentiment scoring
> Simple scheduler or cron job: to run the loop daily
> 
> Mild Shock schrieb:
>>
>> I dunno. Was trying to find another story,
>> of a scientiest who studies the mind, and
>> then goes slowly crazy when he discovers
>>
>> how the mind works. But what I posted
>> is a **Plot Summmary** of a short story:
>>
>>  > "The Nine Billion Names of God" is a
>>  > 1953 science fiction short story by
>>  > British writer Arthur C. Clarke.
>>  > https://en.wikipedia.org/wiki/The_Nine_Billion_Names_of_God
>>
>> Maybe he used some existing Asian lore,
>> I don't know. Who is an expert in this matter?
>>
>> WM schrieb:
>>  > On 05.10.2025 00:52, Mild Shock wrote:
>>  >>
>>  >> In a Tibetan lamasery, the monks seek to
>>  >> list all of the names of God. They believe
>>  >> the Universe was created for this purpose,
>>  >> and that once this naming is completed, God
>>  >> will bring the Universe to an end. Three
>>  >
>>  > A very old story. I read it as a child.
>>  >
>>  > Regards, WM
>>
>> Mild Shock schrieb:
>>>
>>> In a Tibetan lamasery, the monks seek to
>>> list all of the names of God. They believe
>>> the Universe was created for this purpose,
>>> and that once this naming is completed, God
>>> will bring the Universe to an end. Three
>>>
>>> centuries ago, the monks created an alphabet
>>> in which they calculated they could encode
>>> all the possible names of God, numbering
>>> about 9,000,000,000 ("nine billion") and
>>> each having no more than nine characters.
>>>
>>> riting the names out by hand, as they had
>>> been doing, even after eliminating various
>>> nonsense combinations, would take another
>>> 15,000 years; the monks wish to use modern
>>> technology to finish this task in 100 days.
>>>
>>> They rent a computer capable of printing all
>>> the possible permutations, and hire two
>>> Westerners to install and program the machine.
>>> The computer operators are skeptical but
>>> play along. After three months, as the job
>>>
>>> nears completion, they fear that the monks
>>> will blame the computer (and, by extension,
>>> its operators) when nothing happens. The
>>> Westerners leave slightly earlier than their
>>> scheduled departure without warning the monks,
>>>
>>> so that it will complete its final print run
>>> shortly after they leave. On their way to the
>>> airfield they pause on the mountain path. Under
>>> a clear night sky they es timate that it must be
>>> just about the time that the monks are pasting
>>>
>>> the final printed names into their holy books.
>>> Then they notice that "overhead, without any
>>> fuss, the stars were going out."
>>>
>>> https://en.wikipedia.org/wiki/The_Nine_Billion_Names_of_God
>>
> 

[toc] | [prev] | [next] | [standalone]

#14997 — Knight Rider (2025/2026) will kill us all (Re: Vanishing Point . [Le Dot])

Hi,

Now I have the feeling we are minutes away
from robotic AI. Wondering why all the AI Laptops
have now GPUs with ray tracing hardware.

Combining LRM (Large Reasoning Models) with a
3D-Worlds modality could do the job. One could
use genetic algorithms to produce synthetic

training data. This might challenge the
connotation behind the holy grail of AI,
called "Embodiment":

"Der Schweizer Informatiker und Robotik-Experte
Rolf Pfeifer nimmt in diesem Kontext den
Standpunkt ein, dass Intelligenz ausschließlich
verkörperten Agenten, d. h. realen physischen
Systemen, deren Verhalten in der Interaktion
mit der Umwelt beobachtbar ist, zugeschrieben
werden kann.",
- How the body shapes the way we think. A new
   view of intelligence, Pfeifer & Bongard, 2007

Bye

Mild Shock schrieb:
> 
> In a Tibetan lamasery, the monks seek to
> list all of the names of God. They believe
> the Universe was created for this purpose,
> and that once this naming is completed, God
> will bring the Universe to an end. Three
> 
> centuries ago, the monks created an alphabet
> in which they calculated they could encode
> all the possible names of God, numbering
> about 9,000,000,000 ("nine billion") and
> each having no more than nine characters.
> 
> riting the names out by hand, as they had
> been doing, even after eliminating various
> nonsense combinations, would take another
> 15,000 years; the monks wish to use modern
> technology to finish this task in 100 days.
> 
> They rent a computer capable of printing all
> the possible permutations, and hire two
> Westerners to install and program the machine.
> The computer operators are skeptical but
> play along. After three months, as the job
> 
> nears completion, they fear that the monks
> will blame the computer (and, by extension,
> its operators) when nothing happens. The
> Westerners leave slightly earlier than their
> scheduled departure without warning the monks,
> 
> so that it will complete its final print run
> shortly after they leave. On their way to the
> airfield they pause on the mountain path. Under
> a clear night sky they es timate that it must be
> just about the time that the monks are pasting
> 
> the final printed names into their holy books.
> Then they notice that "overhead, without any
> fuss, the stars were going out."
> 
> https://en.wikipedia.org/wiki/The_Nine_Billion_Names_of_God

[toc] | [prev] | [next] | [standalone]

#14998 — The AGI Gambit: Hype vs. Reality (Re: Knight Rider (2025/2026) will kill us all)

Hi,

There are quite some signs of LWM (Large World
Model) AI on the horizon. Like for example:

Controllable World Models are HERE
Explore a playable world model, Marble, from World Labs,
now available for free. This multimodal AI generates 3D
environments from various inputs, including text and images.
Users can navigate, edit, and export these worlds
in multiple formats.
https://www.youtube.com/watch?v=-QbM6_BSdS0

Grok 5 by Elon Musk — AGI Is Closer Than You Think
Musk's claim that Grok 5 has a "10% and rising" chance
of achieving AGI demands scrutiny. He's gone further,
stating "Grok 5 will be AGI or something indistinguishable
from AGI"—a prediction that would make it the first
system to achieve human-level general intelligence.
https://www.youtube.com/watch?v=Afsl4qUsfdw

Bye

Mild Shock schrieb:
> Hi,
> 
> Now I have the feeling we are minutes away
> from robotic AI. Wondering why all the AI Laptops
> have now GPUs with ray tracing hardware.
> 
> Combining LRM (Large Reasoning Models) with a
> 3D-Worlds modality could do the job. One could
> use genetic algorithms to produce synthetic
> 
> training data. This might challenge the
> connotation behind the holy grail of AI,
> called "Embodiment":
> 
> "Der Schweizer Informatiker und Robotik-Experte
> Rolf Pfeifer nimmt in diesem Kontext den
> Standpunkt ein, dass Intelligenz ausschließlich
> verkörperten Agenten, d. h. realen physischen
> Systemen, deren Verhalten in der Interaktion
> mit der Umwelt beobachtbar ist, zugeschrieben
> werden kann.",
> - How the body shapes the way we think. A new
>    view of intelligence, Pfeifer & Bongard, 2007
> 
> Bye
> 
> Mild Shock schrieb:
>>
>> In a Tibetan lamasery, the monks seek to
>> list all of the names of God. They believe
>> the Universe was created for this purpose,
>> and that once this naming is completed, God
>> will bring the Universe to an end. Three
>>
>> centuries ago, the monks created an alphabet
>> in which they calculated they could encode
>> all the possible names of God, numbering
>> about 9,000,000,000 ("nine billion") and
>> each having no more than nine characters.
>>
>> riting the names out by hand, as they had
>> been doing, even after eliminating various
>> nonsense combinations, would take another
>> 15,000 years; the monks wish to use modern
>> technology to finish this task in 100 days.
>>
>> They rent a computer capable of printing all
>> the possible permutations, and hire two
>> Westerners to install and program the machine.
>> The computer operators are skeptical but
>> play along. After three months, as the job
>>
>> nears completion, they fear that the monks
>> will blame the computer (and, by extension,
>> its operators) when nothing happens. The
>> Westerners leave slightly earlier than their
>> scheduled departure without warning the monks,
>>
>> so that it will complete its final print run
>> shortly after they leave. On their way to the
>> airfield they pause on the mountain path. Under
>> a clear night sky they es timate that it must be
>> just about the time that the monks are pasting
>>
>> the final printed names into their holy books.
>> Then they notice that "overhead, without any
>> fuss, the stars were going out."
>>
>> https://en.wikipedia.org/wiki/The_Nine_Billion_Names_of_God
> 

[toc] | [prev] | [next] | [standalone]

#14999 — Stable Diffusion with Control Net (Re: The AGI Gambit: Hype vs. Reality)

Hi,

So the "Amuse" App on my AI Laptops is full of Stable
Diffusion with Control Net. Robots may use Stable
Diffusion with Control Net. Like in this GitHub project:

Adding Conditional Control to Text-to-Image Diffusion Models
https://github.com/lllyasviel/ControlNet

The idea is already being used to control robots in new
research. Like Diffusion Policies (Stanford, 2023–2024),
Robots use a diffusion model to generate action

trajectories instead of images. Trajectory Diffusion, QDiffusion,
SE(3) Diffusion, UniDexGrasp, GenAug, etc... The field is exploding.
many labs have shown diffusion models beating RL in:

- grasping
- manipulation
- locomotion
- navigation

Have Fun!

Bye

Mild Shock schrieb:
> Hi,
> 
> There are quite some signs of LWM (Large World
> Model) AI on the horizon. Like for example:
> 
> Controllable World Models are HERE
> Explore a playable world model, Marble, from World Labs,
> now available for free. This multimodal AI generates 3D
> environments from various inputs, including text and images.
> Users can navigate, edit, and export these worlds
> in multiple formats.
> https://www.youtube.com/watch?v=-QbM6_BSdS0
> 
> Grok 5 by Elon Musk — AGI Is Closer Than You Think
> Musk's claim that Grok 5 has a "10% and rising" chance
> of achieving AGI demands scrutiny. He's gone further,
> stating "Grok 5 will be AGI or something indistinguishable
> from AGI"—a prediction that would make it the first
> system to achieve human-level general intelligence.
> https://www.youtube.com/watch?v=Afsl4qUsfdw
> 
> Bye
> 
> Mild Shock schrieb:
>> Hi,
>>
>> Now I have the feeling we are minutes away
>> from robotic AI. Wondering why all the AI Laptops
>> have now GPUs with ray tracing hardware.
>>
>> Combining LRM (Large Reasoning Models) with a
>> 3D-Worlds modality could do the job. One could
>> use genetic algorithms to produce synthetic
>>
>> training data. This might challenge the
>> connotation behind the holy grail of AI,
>> called "Embodiment":
>>
>> "Der Schweizer Informatiker und Robotik-Experte
>> Rolf Pfeifer nimmt in diesem Kontext den
>> Standpunkt ein, dass Intelligenz ausschließlich
>> verkörperten Agenten, d. h. realen physischen
>> Systemen, deren Verhalten in der Interaktion
>> mit der Umwelt beobachtbar ist, zugeschrieben
>> werden kann.",
>> - How the body shapes the way we think. A new
>>    view of intelligence, Pfeifer & Bongard, 2007
>>
>> Bye
>>
>> Mild Shock schrieb:
>>>
>>> In a Tibetan lamasery, the monks seek to
>>> list all of the names of God. They believe
>>> the Universe was created for this purpose,
>>> and that once this naming is completed, God
>>> will bring the Universe to an end. Three
>>>
>>> centuries ago, the monks created an alphabet
>>> in which they calculated they could encode
>>> all the possible names of God, numbering
>>> about 9,000,000,000 ("nine billion") and
>>> each having no more than nine characters.
>>>
>>> riting the names out by hand, as they had
>>> been doing, even after eliminating various
>>> nonsense combinations, would take another
>>> 15,000 years; the monks wish to use modern
>>> technology to finish this task in 100 days.
>>>
>>> They rent a computer capable of printing all
>>> the possible permutations, and hire two
>>> Westerners to install and program the machine.
>>> The computer operators are skeptical but
>>> play along. After three months, as the job
>>>
>>> nears completion, they fear that the monks
>>> will blame the computer (and, by extension,
>>> its operators) when nothing happens. The
>>> Westerners leave slightly earlier than their
>>> scheduled departure without warning the monks,
>>>
>>> so that it will complete its final print run
>>> shortly after they leave. On their way to the
>>> airfield they pause on the mountain path. Under
>>> a clear night sky they es timate that it must be
>>> just about the time that the monks are pasting
>>>
>>> the final printed names into their holy books.
>>> Then they notice that "overhead, without any
>>> fuss, the stars were going out."
>>>
>>> https://en.wikipedia.org/wiki/The_Nine_Billion_Names_of_God
>>
> 

[toc] | [prev] | [standalone]

Back to top | Article view | comp.lang.prolog

csiph-web