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

Groups > comp.lang.prolog > #15206 > unrolled thread

Good Morning, Vietnam! Sudoku 2026 Challenge

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

Contents

  Good Morning, Vietnam! Sudoku 2026 Challenge Mild Shock <janburse@fastmail.fm> - 2026-01-05 12:11 +0100
    Help! An AI found a Heuristic for me (Was: Good Morning, Vietnam! Sudoku 2026 Challenge) Mild Shock <janburse@fastmail.fm> - 2026-01-10 02:38 +0100
      Failed coping with AI: Moshe Vardi (Was: Help! An AI found a Heuristic for me) Mild Shock <janburse@fastmail.fm> - 2026-01-10 03:02 +0100
        A new proposal: The Gladiator Test (Was: Failed coping with AI: Moshe Vardi) Mild Shock <janburse@fastmail.fm> - 2026-01-10 03:20 +0100
    Chad Brown's Megalodon goes LLM (Was: Good Morning, Vietnam! Sudoku 2026 Challenge) Mild Shock <janburse@fastmail.fm> - 2026-01-10 17:19 +0100
      Magic square of squares [Richard Guy] (Re: Chad Brown's Megalodon goes LLM) Mild Shock <janburse@fastmail.fm> - 2026-01-22 01:40 +0100

#15206 — Good Morning, Vietnam! Sudoku 2026 Challenge

Hi,

Good Morning Vietnam, the HPC-AI Convergence
doesn't sleep. Here a friendly reminder of
the Sudoku leader board (Topn87 Challenge):

#1: jczsolve / Rust WASM
     Solving 87 Sudokus in 0.006 seconds (0 sec / Sudoku)
     https://emerentius.github.io/sudoku_web/

#2: Kudoku / JavaScript
     Solving 87 Sudokus in 0.043 seconds (0.0004 sec / Sudoku)
     https://attractivechaos.github.io/plb/kudoku.html

#3: Picat / import cp. solve([ff],L)
     CPU time 0.175 seconds
     https://picat-lang.org/

#4: Picat / import sat. solve(L)
     CPU time 0.373 seconds
     https://fmv.jku.at/kissat/

Tested on Windows 11, with a AMD Ryzen AI 350

Didn't test yet GNU Prolog, ECLiPSe Prolog or
Ciao Prolog. So whats next? Well beat #1 by
tapping into an NPU of Copilot+ PC.

Have Fun!

Its Winner Winner Chicken Dinner time again...

Bye

[toc] | [next] | [standalone]

#15219 — Help! An AI found a Heuristic for me (Was: Good Morning, Vietnam! Sudoku 2026 Challenge)

Hi,

No joke, I have a new CLP heuristic coded,
and made a blind test. Explaind to the AI
what I coded, but not how coded it,

omitted the crucial scoring function, and
the AI was exactly proposing this scoring
funcion out of the blue.

I was quite bamboozled. The AI could not tell
me how it recalled or halucinated the formula,
I spent like 1-2 hours trying to find a

precedent paper. Will blog about it later.
So whats going on . Why can AI suddently
propose Heuristics. Because they are now

differently trained:

Phase 1: Early LLMs (~2018–2020)
Phase 2: LLMs with Prompted Reasoning (~2020–2022)
Phase 3: Scaling & Reasoning Engineering (~2023–2025)

Training an AI to look for heuristics,
can help in self programming. The generative
part of the AI would generate some code,

which then the AI would use for itself. See also:

How it started:

AI for Low-Code for AI
Nikitha Rao, Jason Tsay, Kiran Kate,
Vincent J. Hellendoorn, Martin Hirzel - 2023
https://arxiv.org/abs/2305.20015

How its going:

Successor-Generator Planning with LLM-generated Heuristics
Alexander Tuisov1, Yonatan Vernik2,
Alexander Shleyfman - 2024
https://arxiv.org/abs/2501.18784v4

Bye

Mild Shock schrieb:
> Hi,
> 
> Good Morning Vietnam, the HPC-AI Convergence
> doesn't sleep. Here a friendly reminder of
> the Sudoku leader board (Topn87 Challenge):
> 
> #1: jczsolve / Rust WASM
>      Solving 87 Sudokus in 0.006 seconds (0 sec / Sudoku)
>      https://emerentius.github.io/sudoku_web/
> 
> #2: Kudoku / JavaScript
>      Solving 87 Sudokus in 0.043 seconds (0.0004 sec / Sudoku)
>      https://attractivechaos.github.io/plb/kudoku.html
> 
> #3: Picat / import cp. solve([ff],L)
>      CPU time 0.175 seconds
>      https://picat-lang.org/
> 
> #4: Picat / import sat. solve(L)
>      CPU time 0.373 seconds
>      https://fmv.jku.at/kissat/
> 
> Tested on Windows 11, with a AMD Ryzen AI 350
> 
> Didn't test yet GNU Prolog, ECLiPSe Prolog or
> Ciao Prolog. So whats next? Well beat #1 by
> tapping into an NPU of Copilot+ PC.
> 
> Have Fun!
> 
> Its Winner Winner Chicken Dinner time again...
> 
> Bye
> 

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

#15220 — Failed coping with AI: Moshe Vardi (Was: Help! An AI found a Heuristic for me)

Hi,

How to spot a failed coping with AI?
Very easy, if somebody starts some gibberish
about consciousness you can be sure he is clueless:

Are AI minds genuine minds?
A genuine mind is typically associated with
consciousness, self-awareness, intentionality,
and the capacity to experience mental states such as emotions.
https://events.rice.edu/event/416255-moshe-vardi-world-logic-day-lecture

What will be Moshe Vardi conclusion nowadays,
that self programming is a first form of
inner monologue? Well I guess the GPUs have

termal activity either way, who cares?

LoL

Bye

P.S.: Just observe how clever Alan Turing was:
He writes near the beginning of the paper:

“I propose to consider the question, ‘Can machines
think?’ This should begin with definitions of the
meaning of the terms ‘machine’ and ‘think.’”

To avoid endless philosophical debate, he replaces
it with the imitation game.

Key points:
- Word usage — Turing does not use the word “consciousness”
in the paper as a criterion for the test.

- Intent — He wanted to sidestep debates about
inner experience and focus on observable behavior.

- Later interpretations — Many later philosophers
and AI critics have connected the Turing Test to
questions of consciousness, but this is a re-
interpretation or extension of Turing’s original proposal.

So, while the Turing Test is often brought into
discussions of machine consciousness, Turing himself
did not mention it in that context in his 1950 paper.

Mild Shock schrieb:
> Hi,
> 
> No joke, I have a new CLP heuristic coded,
> and made a blind test. Explaind to the AI
> what I coded, but not how coded it,
> 
> omitted the crucial scoring function, and
> the AI was exactly proposing this scoring
> funcion out of the blue.
> 
> I was quite bamboozled. The AI could not tell
> me how it recalled or halucinated the formula,
> I spent like 1-2 hours trying to find a
> 
> precedent paper. Will blog about it later.
> So whats going on . Why can AI suddently
> propose Heuristics. Because they are now
> 
> differently trained:
> 
> Phase 1: Early LLMs (~2018–2020)
> Phase 2: LLMs with Prompted Reasoning (~2020–2022)
> Phase 3: Scaling & Reasoning Engineering (~2023–2025)
> 
> Training an AI to look for heuristics,
> can help in self programming. The generative
> part of the AI would generate some code,
> 
> which then the AI would use for itself. See also:
> 
> How it started:
> 
> AI for Low-Code for AI
> Nikitha Rao, Jason Tsay, Kiran Kate,
> Vincent J. Hellendoorn, Martin Hirzel - 2023
> https://arxiv.org/abs/2305.20015
> 
> How its going:
> 
> Successor-Generator Planning with LLM-generated Heuristics
> Alexander Tuisov1, Yonatan Vernik2,
> Alexander Shleyfman - 2024
> https://arxiv.org/abs/2501.18784v4
> 
> Bye
> 
> Mild Shock schrieb:
>> Hi,
>>
>> Good Morning Vietnam, the HPC-AI Convergence
>> doesn't sleep. Here a friendly reminder of
>> the Sudoku leader board (Topn87 Challenge):
>>
>> #1: jczsolve / Rust WASM
>>      Solving 87 Sudokus in 0.006 seconds (0 sec / Sudoku)
>>      https://emerentius.github.io/sudoku_web/
>>
>> #2: Kudoku / JavaScript
>>      Solving 87 Sudokus in 0.043 seconds (0.0004 sec / Sudoku)
>>      https://attractivechaos.github.io/plb/kudoku.html
>>
>> #3: Picat / import cp. solve([ff],L)
>>      CPU time 0.175 seconds
>>      https://picat-lang.org/
>>
>> #4: Picat / import sat. solve(L)
>>      CPU time 0.373 seconds
>>      https://fmv.jku.at/kissat/
>>
>> Tested on Windows 11, with a AMD Ryzen AI 350
>>
>> Didn't test yet GNU Prolog, ECLiPSe Prolog or
>> Ciao Prolog. So whats next? Well beat #1 by
>> tapping into an NPU of Copilot+ PC.
>>
>> Have Fun!
>>
>> Its Winner Winner Chicken Dinner time again...
>>
>> Bye
>>
> 

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

#15221 — A new proposal: The Gladiator Test (Was: Failed coping with AI: Moshe Vardi)

Hi,

Now replace "machine" by "bioform", and "think"
by "life" in the Turing Test, and you get the
Gladiator Test. The judgement is not by a single

counterpart behind a teletype, but a crowd,
we could invoke the image of a Hollywood thumbs-up or
thumbs-down gesture as signalling a yes or no.

Source:

Adrian Tchaikovsky
Dogs of War
https://en.wikipedia.org/wiki/Dogs_of_War_(2017_novel)

How Tchaikovsky’s deepens the Turing Analogy

- From "Can it think?" to "Does it suffer? Does
it deserve to live?" Turing’s test is about cognitive
performance. The bioforms’ struggle is about moral
standing. Their "Turing Test" isn’t to prove
intelligence (they are clearly intelligent), but
to prove consciousness, sentience, and intrinsic
value to a society that legally defines them as "equipment."

- The Crowd is the Global Public & Legal System
In the novel, the "crowd" isn't just a Roman audience—it's
the court of public opinion, international law, media,
and corporate boardrooms. The question is whether this
"crowd" will show mercy (granting personhood) or
condemn them to remain weapons.

Etc.. Etc..

Have Fun!

Bye

Mild Shock schrieb:
> Hi,
> 
> How to spot a failed coping with AI?
> Very easy, if somebody starts some gibberish
> about consciousness you can be sure he is clueless:
> 
> Are AI minds genuine minds?
> A genuine mind is typically associated with
> consciousness, self-awareness, intentionality,
> and the capacity to experience mental states such as emotions.
> https://events.rice.edu/event/416255-moshe-vardi-world-logic-day-lecture
> 
> What will be Moshe Vardi conclusion nowadays,
> that self programming is a first form of
> inner monologue? Well I guess the GPUs have
> 
> termal activity either way, who cares?
> 
> LoL
> 
> Bye
> 
> P.S.: Just observe how clever Alan Turing was:
> He writes near the beginning of the paper:
> 
> “I propose to consider the question, ‘Can machines
> think?’ This should begin with definitions of the
> meaning of the terms ‘machine’ and ‘think.’”
> 
> To avoid endless philosophical debate, he replaces
> it with the imitation game.
> 
> Key points:
> - Word usage — Turing does not use the word “consciousness”
> in the paper as a criterion for the test.
> 
> - Intent — He wanted to sidestep debates about
> inner experience and focus on observable behavior.
> 
> - Later interpretations — Many later philosophers
> and AI critics have connected the Turing Test to
> questions of consciousness, but this is a re-
> interpretation or extension of Turing’s original proposal.
> 
> So, while the Turing Test is often brought into
> discussions of machine consciousness, Turing himself
> did not mention it in that context in his 1950 paper.
> 
> Mild Shock schrieb:
>> Hi,
>>
>> No joke, I have a new CLP heuristic coded,
>> and made a blind test. Explaind to the AI
>> what I coded, but not how coded it,
>>
>> omitted the crucial scoring function, and
>> the AI was exactly proposing this scoring
>> funcion out of the blue.
>>
>> I was quite bamboozled. The AI could not tell
>> me how it recalled or halucinated the formula,
>> I spent like 1-2 hours trying to find a
>>
>> precedent paper. Will blog about it later.
>> So whats going on . Why can AI suddently
>> propose Heuristics. Because they are now
>>
>> differently trained:
>>
>> Phase 1: Early LLMs (~2018–2020)
>> Phase 2: LLMs with Prompted Reasoning (~2020–2022)
>> Phase 3: Scaling & Reasoning Engineering (~2023–2025)
>>
>> Training an AI to look for heuristics,
>> can help in self programming. The generative
>> part of the AI would generate some code,
>>
>> which then the AI would use for itself. See also:
>>
>> How it started:
>>
>> AI for Low-Code for AI
>> Nikitha Rao, Jason Tsay, Kiran Kate,
>> Vincent J. Hellendoorn, Martin Hirzel - 2023
>> https://arxiv.org/abs/2305.20015
>>
>> How its going:
>>
>> Successor-Generator Planning with LLM-generated Heuristics
>> Alexander Tuisov1, Yonatan Vernik2,
>> Alexander Shleyfman - 2024
>> https://arxiv.org/abs/2501.18784v4
>>
>> Bye
>>
>> Mild Shock schrieb:
>>> Hi,
>>>
>>> Good Morning Vietnam, the HPC-AI Convergence
>>> doesn't sleep. Here a friendly reminder of
>>> the Sudoku leader board (Topn87 Challenge):
>>>
>>> #1: jczsolve / Rust WASM
>>>      Solving 87 Sudokus in 0.006 seconds (0 sec / Sudoku)
>>>      https://emerentius.github.io/sudoku_web/
>>>
>>> #2: Kudoku / JavaScript
>>>      Solving 87 Sudokus in 0.043 seconds (0.0004 sec / Sudoku)
>>>      https://attractivechaos.github.io/plb/kudoku.html
>>>
>>> #3: Picat / import cp. solve([ff],L)
>>>      CPU time 0.175 seconds
>>>      https://picat-lang.org/
>>>
>>> #4: Picat / import sat. solve(L)
>>>      CPU time 0.373 seconds
>>>      https://fmv.jku.at/kissat/
>>>
>>> Tested on Windows 11, with a AMD Ryzen AI 350
>>>
>>> Didn't test yet GNU Prolog, ECLiPSe Prolog or
>>> Ciao Prolog. So whats next? Well beat #1 by
>>> tapping into an NPU of Copilot+ PC.
>>>
>>> Have Fun!
>>>
>>> Its Winner Winner Chicken Dinner time again...
>>>
>>> Bye
>>>
>>
> 

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

#15227 — Chad Brown's Megalodon goes LLM (Was: Good Morning, Vietnam! Sudoku 2026 Challenge)

Hi,

This is a brief description of a project that
has already autoformalized a large portion of
the general topology from the Munkres textbook
(which has in total 241 pages in 7 chapters
and 39 sections).

The proof checker is Chad Brown's higher-order
set theory system Megalodon, and the core library
is Brown's formalization of basic set theory
and surreal numbers (including reals, etc).
The rest is some prompt engineering and
technical choices which we describe here.

130k Lines of Formal Topology in Two Weeks
https://arxiv.org/abs/2601.03298

Is this the end to the mess, of the same
theorem proved with proof assistant A, but
using set theory, and also proved with proof
assistant B, but using type theory.

How malleable are LLM generated proofs?

Bye

Mild Shock schrieb:
> Hi,
> 
> Good Morning Vietnam, the HPC-AI Convergence
> doesn't sleep. Here a friendly reminder of
> the Sudoku leader board (Topn87 Challenge):
> 
> #1: jczsolve / Rust WASM
>      Solving 87 Sudokus in 0.006 seconds (0 sec / Sudoku)
>      https://emerentius.github.io/sudoku_web/
> 
> #2: Kudoku / JavaScript
>      Solving 87 Sudokus in 0.043 seconds (0.0004 sec / Sudoku)
>      https://attractivechaos.github.io/plb/kudoku.html
> 
> #3: Picat / import cp. solve([ff],L)
>      CPU time 0.175 seconds
>      https://picat-lang.org/
> 
> #4: Picat / import sat. solve(L)
>      CPU time 0.373 seconds
>      https://fmv.jku.at/kissat/
> 
> Tested on Windows 11, with a AMD Ryzen AI 350
> 
> Didn't test yet GNU Prolog, ECLiPSe Prolog or
> Ciao Prolog. So whats next? Well beat #1 by
> tapping into an NPU of Copilot+ PC.
> 
> Have Fun!
> 
> Its Winner Winner Chicken Dinner time again...
> 
> Bye
> 

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

#15349 — Magic square of squares [Richard Guy] (Re: Chad Brown's Megalodon goes LLM)

Hi,

How about this constraint solving problem:

"For 4×4 magic squares, it’s unknown whether a
magic square of squares exists. This is the most
famous open case. No one has found one, and no
proof exists that one cannot exist."
https://en.wikipedia.org/wiki/Magic_square_of_squares

Well the 4x4 problem fits well, I am currently
working on a CLP(FD) that can detect quadratic forms.
Something to explore as part of the Railgun CLP(FD)

project. Just joking. I saw the explosion of a
4x4 magic square today, by mistake I changed L ins
1..4 into L ins 1..9, and boom, no more solutions

obtainable in short time.

Bye

Mild Shock schrieb:
> Hi,
> 
> This is a brief description of a project that
> has already autoformalized a large portion of
> the general topology from the Munkres textbook
> (which has in total 241 pages in 7 chapters
> and 39 sections).
> 
> The proof checker is Chad Brown's higher-order
> set theory system Megalodon, and the core library
> is Brown's formalization of basic set theory
> and surreal numbers (including reals, etc).
> The rest is some prompt engineering and
> technical choices which we describe here.
> 
> 130k Lines of Formal Topology in Two Weeks
> https://arxiv.org/abs/2601.03298
> 
> Is this the end to the mess, of the same
> theorem proved with proof assistant A, but
> using set theory, and also proved with proof
> assistant B, but using type theory.
> 
> How malleable are LLM generated proofs?
> 
> Bye
> 
> Mild Shock schrieb:
>> Hi,
>>
>> Good Morning Vietnam, the HPC-AI Convergence
>> doesn't sleep. Here a friendly reminder of
>> the Sudoku leader board (Topn87 Challenge):
>>
>> #1: jczsolve / Rust WASM
>>      Solving 87 Sudokus in 0.006 seconds (0 sec / Sudoku)
>>      https://emerentius.github.io/sudoku_web/
>>
>> #2: Kudoku / JavaScript
>>      Solving 87 Sudokus in 0.043 seconds (0.0004 sec / Sudoku)
>>      https://attractivechaos.github.io/plb/kudoku.html
>>
>> #3: Picat / import cp. solve([ff],L)
>>      CPU time 0.175 seconds
>>      https://picat-lang.org/
>>
>> #4: Picat / import sat. solve(L)
>>      CPU time 0.373 seconds
>>      https://fmv.jku.at/kissat/
>>
>> Tested on Windows 11, with a AMD Ryzen AI 350
>>
>> Didn't test yet GNU Prolog, ECLiPSe Prolog or
>> Ciao Prolog. So whats next? Well beat #1 by
>> tapping into an NPU of Copilot+ PC.
>>
>> Have Fun!
>>
>> Its Winner Winner Chicken Dinner time again...
>>
>> Bye
>>
> 

[toc] | [prev] | [standalone]

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

csiph-web