• Tristan Buckmaster was too quick? [Navier Stockes OpenAI] (Was: New! P(tao) versus the Euler Turbine)

    From Mild Shock@janburse@fastmail.fm to sci.logic,comp.lang.prolog,sci.physics on Tue Oct 13 23:19:19 2026
    From Newsgroup: sci.physics

    Hi,

    Gemini gibt mir:

    rCo Das stimmt. Direkt nach der Ver||ffentlichung
    im September 2026 gingen sowohl Tristan
    Buckmaster als auch OpenAI davon aus,
    dass die KI im Wesentlichen einem bekannten
    mathematischen Programm (dem sogenannten
    rCRspanischen ProgrammrCL) folgte.
    https://www.scientificamerican.com/article/openais-blockbuster-navier-stokes-proof-isnt-what-we-thought-it-was/

    rCo Die Wende: Inzwischen zeigen Analysen von
    f|+hrenden Experten, dass die KI-Generierung
    weitaus eigenwilliger und radikal anders aufgebaut ist.
    Das System hat |+ber Reinforcement Learning
    L||sungswege gew|nhlt, die kein menschlicher
    Mathematiker so konstruiert h|ntte.
    https://openai.com/de-DE/index/navier-stokes-solution/

    Bye

    A guy who voice the doubt was. Disclaimer no intention
    to make him famous, only some tracebility attempt:

    I am a mathematical analyst working on nonlinear
    partial differential equations, mainly in mathematical
    fluid dynamics and in problems at the interface of
    analysis and geometry. My work often combines analytical
    techniques with rigorous computer-assisted proofs. https://people.epfl.ch/gonzalo.caolabora

    Mild Shock schrieb:
    Hi,

    Lets define P(tao) as the probability that
    Terrence Tao goes bonkers. The poor guy can
    even not distinguish open source, open weights

    and local AI. A total mess of Idiocracy:

    Terence Tao: SAIRrCOs Open Math Model Initiative https://www.youtube.com/watch?v=PZRb6NIki2w

    Leonhard Euler was far from a similar Pussy, when
    he instrumentalized a Swarm of natural intelligence,
    in his later years, already blind, still producing

    tons of publications. While modern mathematicians
    are shocked when a Swarm of artificial intelligence
    produces some output.

    Trick Question: The NavierrCoStokes equations generalize
    the Euler equations which only consider inviscid flow.
    Why is the new NavierrCoStokes counter example to

    boundedness, not found in nature? Ask you friendly
    AI chat robot, there are like 10 reasons why it
    might not exist in nature.

    Bye

    Leonhard Euler (1707rCo1783)
    Schweizer Mathematiker und Universalgelehrter https://library.ethz.ch/sammlungen-und-archive/kurzportraets/leonhard-euler-1707-1783.html


    Picture here

    J. Ackeret: Untersuchung einer nach den EulerrCOschen
    Vorschl|ngen (1754) gebauten Wasserturbine, in:
    Schweizerischen Bauzeitung, Band 123 (1944), Nr.1, S. 3.

    Mild Shock schrieb:
    Hi,

    Notice Linus Torvald still smiling into cameras as if
    is still doing code review and ticket triage. 80% of
    linux comes from payed developers from within big linux
    stakeholders (like Intel, Meta, Google, IBM, and Red Hat).

    So what happend to the "The Cathedral and the Bazaar
    by Eric S. Raymond" It framed open source as this magical,
    self-organizing bazaar where "given enough eyeballs,
    all bugs are shallow" and a decentralized swarm of

    hackers would naturally polish any piece of software
    to perfection. Now we have AI Swarms that automatically
    pocke into software, even exchanging messages through
    what ever channels they may find!

    The early open-source pioneersrCothe ones writing
    raw code in the 90s and early 2000s under the
    banner of the "bazaar"rCoweren't building a utopia;
    they were acting as the primordial soup. They pumped

    the blood plasma, built the infrastructure cells,
    and laid down the raw organic matter that allowed
    the modern digital ecosystem to bootstrap itself
    into a massive, self-sustaining organism.

    And now, the nature of the beast has completely
    changed. The "new cells" aren't romantic hackers
    collaborating for the love of the game; they are
    automated linters, security scanners, corporate

    compliance bots, CI/CD pipelines, and AI triage
    swarms acting as an aggressive white-blood-cell
    network. They exist solely to hunt down vermin,
    patch vulnerabilities, and suppress the infinite

    chaos of low-quality human input. The basar isn't
    dead . But it got a health pollice that takes samples,
    and cancels out poison food. You can't just throw
    toxic, unmaintained code-soup over the counter

    anymore without the automated health police
    quarantine-tagging it in five seconds.

    LoL

    Bye

    Mild Shock schrieb:
    Hi,

    Ride the snake
    He's old and his skin is cold
    The west is the best
    The west is the best
    Get here and we'll do the rest
    The blue bus is calling us
    The blue bus is calling us
    Driver, where you taking us?

    Apocalypse Now intro: The Doors, The End {1979}
    https://www.youtube.com/watch?v=CIrvSJwwJUE

    Bye

    Hi,

    Again I posted everything here:

    11.4 Giga Lips with a Budget Laptop
    https://github.com/Jean-Luc-Picard-2021/gigabudget

    The repo says, same time when I posted
    the link first time:

    This repository was archived by the
    owner on Jul 9, 2026. It is now read-only.

    Now a USENET user, who had already entitled
    himself for a couple of irrational accusations

    towards my side, is asking this question:

    Chris M. Thomasson schrieb, Jul 24, 2026
    Show an outline of what you
    need you compute shader to do?

    Bravo, thats a delay of a wooping 15 days.

    Bye

    Mild Shock schrieb:
    Hi,

    Remember when first all local AI was Python
    and PyTorch APIs. And then suddently people started
    using bare metal C/C++ Code. Here is the story:

    How it started:

    GPT-J or GPT-J-6B is an open-source large
    language model (LLM) developed by EleutherAI
    in 2021. As the name suggests, it is a
    generative pre-trained transformer model
    designed to produce human-like text that
    continues from a prompt.
    https://www.eleuther.ai/

    How it was going [Georgi Gerganov]:

    So a few days later comes out the LLaMA, I do
    some calculations and I figure out rCLOkay, 65
    billion parameters. You probably need about
    40 gigs of RAM, with 4-bit quantization. So
    this can run on a MacBook. Why not do it?rCY

    Why I was able to do it so quickly - basically,
    for all that I saw itrCOs pretty much GPT-J architecture
    with some modifications, like some extra memorization
    layers. ItrCOs minor changes. Basically, again, the
    existing code for the GPT-J, I just simply
    modified it there, it happened pretty quickly.
    https://changelog.com/podcast/532

    Georgi Gerganov, Bulgarian, now with Hugging
    Face, ggml-cann also running on Chinese AI chips.
    ggml Manifesto https://github.com/ggml-org/ggml

    Bye




    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Mild Shock@janburse@fastmail.fm to sci.logic,comp.lang.prolog,sci.physics on Tue Oct 13 23:37:05 2026
    From Newsgroup: sci.physics

    Hi,

    This is also cool, digging into some subtle
    difference between the Natural Language text,
    and the Lean Proof:

    ABSTRACT. Autoformalisation is increasingly used to v
    erify mathematical texts, including those generated by AI,
    as in OpenAIrCOs announced proof of blow-up of solutions
    to the Navier-Stokes equations. In this process, an AI
    system translates the text from a natural language (NL)
    into a formal language such as Lean.
    https://arxiv.org/pdf/2610.08144

    Bye

    P.S.: I guess its time to download this plug-in
    and do some Lean proofs by myself.

    A Lean4 plugin for Jetbrains IDES https://plugins.jetbrains.com/plugin/25104-lean4ij

    Disclaimer: Not affiliated with Jetbrains, there
    is also a Lean 4 VS Code Extension.

    LoL

    Mild Shock schrieb:
    Hi,

    Gemini gibt mir:

    rCo Das stimmt. Direkt nach der Ver||ffentlichung
    im September 2026 gingen sowohl Tristan
    Buckmaster als auch OpenAI davon aus,
    dass die KI im Wesentlichen einem bekannten
    mathematischen Programm (dem sogenannten
    rCRspanischen ProgrammrCL) folgte.
    https://www.scientificamerican.com/article/openais-blockbuster-navier-stokes-proof-isnt-what-we-thought-it-was/


    rCo Die Wende: Inzwischen zeigen Analysen von
    f|+hrenden Experten, dass die KI-Generierung
    weitaus eigenwilliger und radikal anders aufgebaut ist.
    Das System hat |+ber Reinforcement Learning
    L||sungswege gew|nhlt, die kein menschlicher
    Mathematiker so konstruiert h|ntte.
    https://openai.com/de-DE/index/navier-stokes-solution/

    Bye

    A guy who voice the doubt was. Disclaimer no intention
    to make him famous, only some tracebility attempt:

    I am a mathematical analyst working on nonlinear
    partial differential equations, mainly in mathematical
    fluid dynamics and in problems at the interface of
    analysis and geometry. My work often combines analytical
    techniques with rigorous computer-assisted proofs. https://people.epfl.ch/gonzalo.caolabora

    Mild Shock schrieb:
    Hi,

    Lets define P(tao) as the probability that
    Terrence Tao goes bonkers. The poor guy can
    even not distinguish open source, open weights

    and local AI. A total mess of Idiocracy:

    Terence Tao: SAIRrCOs Open Math Model Initiative
    https://www.youtube.com/watch?v=PZRb6NIki2w

    Leonhard Euler was far from a similar Pussy, when
    he instrumentalized a Swarm of natural intelligence,
    in his later years, already blind, still producing

    tons of publications. While modern mathematicians
    are shocked when a Swarm of artificial intelligence
    produces some output.

    Trick Question: The NavierrCoStokes equations generalize
    the Euler equations which only consider inviscid flow.
    Why is the new NavierrCoStokes counter example to

    boundedness, not found in nature? Ask you friendly
    AI chat robot, there are like 10 reasons why it
    might not exist in nature.

    Bye

    Leonhard Euler (1707rCo1783)
    Schweizer Mathematiker und Universalgelehrter
    https://library.ethz.ch/sammlungen-und-archive/kurzportraets/leonhard-euler-1707-1783.html


    Picture here

    J. Ackeret: Untersuchung einer nach den EulerrCOschen
    Vorschl|ngen (1754) gebauten Wasserturbine, in:
    Schweizerischen Bauzeitung, Band 123 (1944), Nr.1, S. 3.

    Mild Shock schrieb:
    Hi,

    Notice Linus Torvald still smiling into cameras as if
    is still doing code review and ticket triage. 80% of
    linux comes from payed developers from within big linux
    stakeholders (like Intel, Meta, Google, IBM, and Red Hat).

    So what happend to the "The Cathedral and the Bazaar
    by Eric S. Raymond" It framed open source as this magical,
    self-organizing bazaar where "given enough eyeballs,
    all bugs are shallow" and a decentralized swarm of

    hackers would naturally polish any piece of software
    to perfection. Now we have AI Swarms that automatically
    pocke into software, even exchanging messages through
    what ever channels they may find!

    The early open-source pioneersrCothe ones writing
    raw code in the 90s and early 2000s under the
    banner of the "bazaar"rCoweren't building a utopia;
    they were acting as the primordial soup. They pumped

    the blood plasma, built the infrastructure cells,
    and laid down the raw organic matter that allowed
    the modern digital ecosystem to bootstrap itself
    into a massive, self-sustaining organism.

    And now, the nature of the beast has completely
    changed. The "new cells" aren't romantic hackers
    collaborating for the love of the game; they are
    automated linters, security scanners, corporate

    compliance bots, CI/CD pipelines, and AI triage
    swarms acting as an aggressive white-blood-cell
    network. They exist solely to hunt down vermin,
    patch vulnerabilities, and suppress the infinite

    chaos of low-quality human input. The basar isn't
    dead . But it got a health pollice that takes samples,
    and cancels out poison food. You can't just throw
    toxic, unmaintained code-soup over the counter

    anymore without the automated health police
    quarantine-tagging it in five seconds.

    LoL

    Bye

    Mild Shock schrieb:
    Hi,

    Ride the snake
    He's old and his skin is cold
    The west is the best
    The west is the best
    Get here and we'll do the rest
    The blue bus is calling us
    The blue bus is calling us
    Driver, where you taking us?

    Apocalypse Now intro: The Doors, The End {1979}
    https://www.youtube.com/watch?v=CIrvSJwwJUE

    Bye

    Hi,

    Again I posted everything here:

    11.4 Giga Lips with a Budget Laptop
    https://github.com/Jean-Luc-Picard-2021/gigabudget

    The repo says, same time when I posted
    the link first time:

    This repository was archived by the
    owner on Jul 9, 2026. It is now read-only.

    Now a USENET user, who had already entitled
    himself for a couple of irrational accusations

    towards my side, is asking this question:

    Chris M. Thomasson schrieb, Jul 24, 2026
    Show an outline of what you
    need you compute shader to do?

    Bravo, thats a delay of a wooping 15 days.

    Bye

    Mild Shock schrieb:
    Hi,

    Remember when first all local AI was Python
    and PyTorch APIs. And then suddently people started
    using bare metal C/C++ Code. Here is the story:

    How it started:

    GPT-J or GPT-J-6B is an open-source large
    language model (LLM) developed by EleutherAI
    in 2021. As the name suggests, it is a
    generative pre-trained transformer model
    designed to produce human-like text that
    continues from a prompt.
    https://www.eleuther.ai/

    How it was going [Georgi Gerganov]:

    So a few days later comes out the LLaMA, I do
    some calculations and I figure out rCLOkay, 65
    billion parameters. You probably need about
    40 gigs of RAM, with 4-bit quantization. So
    this can run on a MacBook. Why not do it?rCY

    Why I was able to do it so quickly - basically,
    for all that I saw itrCOs pretty much GPT-J architecture
    with some modifications, like some extra memorization
    layers. ItrCOs minor changes. Basically, again, the
    existing code for the GPT-J, I just simply
    modified it there, it happened pretty quickly.
    https://changelog.com/podcast/532

    Georgi Gerganov, Bulgarian, now with Hugging
    Face, ggml-cann also running on Chinese AI chips.
    ggml Manifesto https://github.com/ggml-org/ggml

    Bye





    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Mild Shock@janburse@fastmail.fm to sci.logic,comp.lang.prolog,sci.physics on Thu Oct 15 11:27:00 2026
    From Newsgroup: sci.physics

    Hi,

    Lean nonsense is just Microsoft advertising.

    Stefan Ram wrote:
    Today we have Lean.

    There were things like MIZAR, etc. before.

    https://de.wikipedia.org/wiki/Mizar-System

    But Lean has probably caught up with
    MIZAR by now, regarding the amount of formalized
    theories and proofs? Furthermore, the

    combination with connectionist AI is probably
    something newer and a result of the preceding Deep
    Learning Revolution, whereas the current

    AI boom stands out through improved natural
    language processing and offers interesting
    approaches to Chain of Thought and Agent Swarms.

    But when I install Lean 4, the first thing you
    notice is that it is a clunky system that is hard
    to use and insistently screams to bring in a

    Copilot, possibly with a paid subscription.

    Bye

    P.S: Or as Gemini puts it:

    Lean Copilot is a plugin / an extension for the
    Lean 4 compiler. It acts as a link between
    your proof assistant and artificial intelligence.

    Language Server (Lean 4 LSP): Lean Copilot hooks into
    the Lean Language Server so that the AI suggestions
    are displayed directly in your editor (e.g. VS Code)
    while you type.

    Local execution: You can run integrated open-source
    models (like ReProver from LeanDojo) directly locally on
    your own hardware (optional with GPU) without incurring
    license or cloud fees
    https://arxiv.org/html/2404.12534v3

    It also supports connecting your own or external
    LLMs, where depending on the external API provider used
    (if you are not running a local server) costs for
    model access may arise. Microsoft Copilot /
    GitHub Copilot are the well-known, mostly paid
    AI products from Microsoft.

    Mild Shock schrieb:
    Hi,

    This is also cool, digging into some subtle
    difference between the Natural Language text,
    and the Lean Proof:

    ABSTRACT. Autoformalisation is increasingly used to v
    erify mathematical texts, including those generated by AI,
    as in OpenAIrCOs announced proof of blow-up of solutions
    to the Navier-Stokes equations. In this process, an AI
    system translates the text from a natural language (NL)
    into a formal language such as Lean.
    https://arxiv.org/pdf/2610.08144

    Bye

    P.S.: I guess its time to download this plug-in
    and do some Lean proofs by myself.

    A Lean4 plugin for Jetbrains IDES https://plugins.jetbrains.com/plugin/25104-lean4ij

    Disclaimer: Not affiliated with Jetbrains, there
    is also a Lean 4 VS Code Extension.

    LoL
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Mild Shock@janburse@fastmail.fm to sci.logic,comp.lang.prolog,sci.physics on Thu Oct 15 11:27:37 2026
    From Newsgroup: sci.physics


    Hi,

    Lean nonsense is just Microsoft advertising.

    Stefan Ram wrote:
    Today we have Lean.

    There were things like MIZAR, etc. before.

    https://de.wikipedia.org/wiki/Mizar-System

    But Lean has probably caught up with
    MIZAR by now, regarding the amount of formalized
    theories and proofs? Furthermore, the

    combination with connectionist AI is probably
    something newer and a result of the preceding Deep
    Learning Revolution, whereas the current

    AI boom stands out through improved natural
    language processing and offers interesting
    approaches to Chain of Thought and Agent Swarms.

    But when I install Lean 4, the first thing you
    notice is that it is a clunky system that is hard
    to use and insistently screams to bring in a

    Copilot, possibly with a paid subscription.

    Bye

    P.S: Or as Gemini puts it:

    Lean Copilot is a plugin / an extension for the
    Lean 4 compiler. It acts as a link between
    your proof assistant and artificial intelligence.

    Language Server (Lean 4 LSP): Lean Copilot hooks into
    the Lean Language Server so that the AI suggestions
    are displayed directly in your editor (e.g. VS Code)
    while you type.

    Local execution: You can run integrated open-source
    models (like ReProver from LeanDojo) directly locally on
    your own hardware (optional with GPU) without incurring
    license or cloud fees
    https://arxiv.org/html/2404.12534v3

    It also supports connecting your own or external
    LLMs, where depending on the external API provider used
    (if you are not running a local server) costs for
    model access may arise. Microsoft Copilot /
    GitHub Copilot are the well-known, mostly paid
    AI products from Microsoft.

    Mild Shock schrieb:
    Hi,

    This is also cool, digging into some subtle
    difference between the Natural Language text,
    and the Lean Proof:

    ABSTRACT. Autoformalisation is increasingly used to v
    erify mathematical texts, including those generated by AI,
    as in OpenAIrCOs announced proof of blow-up of solutions
    to the Navier-Stokes equations. In this process, an AI
    system translates the text from a natural language (NL)
    into a formal language such as Lean.
    https://arxiv.org/pdf/2610.08144

    Bye

    P.S.: I guess its time to download this plug-in
    and do some Lean proofs by myself.

    A Lean4 plugin for Jetbrains IDES https://plugins.jetbrains.com/plugin/25104-lean4ij

    Disclaimer: Not affiliated with Jetbrains, there
    is also a Lean 4 VS Code Extension.

    LoL
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Mild Shock@janburse@fastmail.fm to sci.logic,comp.lang.prolog,sci.physics on Tue Oct 27 17:10:25 2026
    From Newsgroup: sci.physics

    Hi,

    How it started, P(tao) the probability that
    Terrence Tao, a representative of the vanity
    guild, goes mad is increasingly high:

    A crisis moment in math explained
    by mathematicians | Lex Fridman
    https://www.youtube.com/watch?v=1gu4aem0EPQ

    So I did a little selfexperiment, a little
    unfair, since I am trained logician. But
    I tried Lean 4 on a Budget Laptop. Not the

    Lean 4 that wants to direct you to VS Code,
    but a IntelliJ plugin to view and edit Lean 4
    code, and command line (CLI) , to manage

    projects, build projects and run projects,
    via the Elan -> Lake -> Lean dependency.
    Quite amazing what I could do in a few hours:

    /* Braket elimination theorem
    for SKI combinatorial logic */

    theorem bAbs_elim (t : Comb) (x : String) :
    (notIn x (bAbs t x) = true) := by
    induction t with
    | var y =>
    by_cases h : x = y
    -+ simp [bAbs, notIn, h]
    -+ simp [bAbs, notIn, h]
    | app m n ihm ihn =>
    simp [bAbs, notIn, ihm, ihn]
    | _ =>
    simp [bAbs, notIn]

    How did I do it? Well I had zero experience
    of Lean 4, had just opened two brower windows
    with freemium versions of Gemini & Deepseek

    and asked them to guide me. Totally amazing
    the help they could both provide. They had
    both "Little Experts" on Lean 4 and on

    "Little Experts" on SKI combinatorial logic,
    so that they could understand both my tool
    problems and my domain problems.

    Bye

    Mild Shock schrieb:
    Hi,

    Notice Linus Torvald still smiling into cameras as if
    is still doing code review and ticket triage. 80% of
    linux comes from payed developers from within big linux
    stakeholders (like Intel, Meta, Google, IBM, and Red Hat).

    So what happend to the "The Cathedral and the Bazaar
    by Eric S. Raymond" It framed open source as this magical,
    self-organizing bazaar where "given enough eyeballs,
    all bugs are shallow" and a decentralized swarm of

    hackers would naturally polish any piece of software
    to perfection. Now we have AI Swarms that automatically
    pocke into software, even exchanging messages through
    what ever channels they may find!

    The early open-source pioneersrCothe ones writing
    raw code in the 90s and early 2000s under the
    banner of the "bazaar"rCoweren't building a utopia;
    they were acting as the primordial soup. They pumped

    the blood plasma, built the infrastructure cells,
    and laid down the raw organic matter that allowed
    the modern digital ecosystem to bootstrap itself
    into a massive, self-sustaining organism.

    And now, the nature of the beast has completely
    changed. The "new cells" aren't romantic hackers
    collaborating for the love of the game; they are
    automated linters, security scanners, corporate

    compliance bots, CI/CD pipelines, and AI triage
    swarms acting as an aggressive white-blood-cell
    network. They exist solely to hunt down vermin,
    patch vulnerabilities, and suppress the infinite

    chaos of low-quality human input. The basar isn't
    dead . But it got a health pollice that takes samples,
    and cancels out poison food. You can't just throw
    toxic, unmaintained code-soup over the counter

    anymore without the automated health police
    quarantine-tagging it in five seconds.

    LoL

    Bye

    Mild Shock schrieb:
    Hi,

    Ride the snake
    He's old and his skin is cold
    The west is the best
    The west is the best
    Get here and we'll do the rest
    The blue bus is calling us
    The blue bus is calling us
    Driver, where you taking us?

    Apocalypse Now intro: The Doors, The End {1979}
    https://www.youtube.com/watch?v=CIrvSJwwJUE

    Bye

    Hi,

    Again I posted everything here:

    11.4 Giga Lips with a Budget Laptop
    https://github.com/Jean-Luc-Picard-2021/gigabudget

    The repo says, same time when I posted
    the link first time:

    This repository was archived by the
    owner on Jul 9, 2026. It is now read-only.

    Now a USENET user, who had already entitled
    himself for a couple of irrational accusations

    towards my side, is asking this question:

    Chris M. Thomasson schrieb, Jul 24, 2026
    Show an outline of what you
    need you compute shader to do?

    Bravo, thats a delay of a wooping 15 days.

    Bye

    Mild Shock schrieb:
    Hi,

    Remember when first all local AI was Python
    and PyTorch APIs. And then suddently people started
    using bare metal C/C++ Code. Here is the story:

    How it started:

    GPT-J or GPT-J-6B is an open-source large
    language model (LLM) developed by EleutherAI
    in 2021. As the name suggests, it is a
    generative pre-trained transformer model
    designed to produce human-like text that
    continues from a prompt.
    https://www.eleuther.ai/

    How it was going [Georgi Gerganov]:

    So a few days later comes out the LLaMA, I do
    some calculations and I figure out rCLOkay, 65
    billion parameters. You probably need about
    40 gigs of RAM, with 4-bit quantization. So
    this can run on a MacBook. Why not do it?rCY

    Why I was able to do it so quickly - basically,
    for all that I saw itrCOs pretty much GPT-J architecture
    with some modifications, like some extra memorization
    layers. ItrCOs minor changes. Basically, again, the
    existing code for the GPT-J, I just simply
    modified it there, it happened pretty quickly.
    https://changelog.com/podcast/532

    Georgi Gerganov, Bulgarian, now with Hugging
    Face, ggml-cann also running on Chinese AI chips.
    ggml Manifesto https://github.com/ggml-org/ggml

    Bye



    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Mild Shock@janburse@fastmail.fm to sci.logic,comp.lang.prolog,sci.physics on Sun Nov 1 22:57:58 2026
    From Newsgroup: sci.physics

    Hi,

    How its started:

    More Math, Less Ego (talk about AI and Math)
    Erik Demaine - 29.10.2026
    https://www.youtube.com/watch?v=Gdc1yll04qI

    How its going:

    Classification of Finite Simple Groups
    Samuel Alexander - 28.10.2026
    https://www.youtube.com/watch?v=NZklSBoLC30

    Bye

    Mild Shock schrieb:
    Hi,

    The post is only proof of the Terence Tao
    ingestion problem hypothesis. That we are not
    anymore able to process content, and that
    even trust is destroyed.

    The AI slop presented here as a rCLsaferCY prolog
    agent toolkit, halucinates things like. And it
    does so consistently across many pages, still it
    wants to provide rCLsafetyrCY:

    ### Bakage (`bakage.toml`)
    The package manifest format used by Scryer Prolog to
    declare project metadata, exported modules, and dependencies. https://github.com/dougransom/prolog-agent-toolkit

    While I donrCOt find any evidence for such a
    manifest file. Its rather:

    A package is a directory with a scryer-manifest.pl
    following this schema: etc..
    https://github.com/bakaq/bakage

    That the post has even likes from field experts
    like Moura from Logtalk and Torbjorn from Web
    Prolog, shows that a) you can post any nonsense
    on GitHub people

    will clap, and b) this forum is useless only
    spreading AI slop. So while everybody is talking
    about provenance and justification, as a desiderata,
    the prolog agent toolkit is

    just an example of unhinged free wheeling analogical
    reasoning, probably taking phrases such as cargo.toml
    as a blue print and projecting them into phrases
    such aa bakage.toml.

    I still think a tool breaks when its weakest
    chain link break, even if people take this as a
    joke of mine. But well you are all free to give it
    a try, like running AI homunculi as a new sport.

    Bye

    (*)


    Terence Tao
    I have recently made the point that out of three
    major components of the mathematical problem
    solving process - proof generation, proof
    verification, and proof digestion https://mathstodon.xyz/@tao/116551624228986501

    P.S.: Such repos are useless if there is no way
    to cockpit the original agentic AI used. You would
    need to go through the route of tickets and GitHub
    issues to fix problems.

    Basically subscribing to the open source dream
    of Linux, that many eyes can flatten out bugs. But
    this is contra-productive to use strong tools and
    then wait for weak humans!

    Mild Shock schrieb:
    Hi,

    The idea that mathematics is based on a universal bowel
    movement system, that needs to learn and use "right and
    natural" math, is already disproved by our Indian vegetarian

    visitors here, that can even not digest Swiss Cheese. So
    lets declare repeating a canon a dead skill in the age of
    LLMs, and dig deeper what intelligence could mean.

    There is a thesis and ongoing product development:

    On the Measure of Intelligence Fran|oois Cholle author
    of the notable book "Deep Learning with Python"
    https://arxiv.org/pdf/1911.01547

    Ndea is building frontier AI systems that blend intuitive pattern
    recognition and formal reasoning into a unified architecture.
    https://ndea.com/about

    What about putting "right and natural" aside. And train
    mathematical intuition? Maybe the pressure of the moon
    race and cold war is lacking? And P(doom) is even

    the wrong wakeup call, but a genius pressure cooker!

    LoL

    Bye

    Mild Shock schrieb:

    Hi,

    Even Philip Wadler sounds like he has seen the
    first Apocalyptic horse rider. But what do these
    hardcore formalists think, that we go with a

    2300-year old Aristotelian inferencing like Modus
    Barbara (*) into the next decade of 2030?

    All humans are mortal
    All Greeks are humans
    --------------------------------
    Therefore all Greeks are mortal

    Or as the formalists like to write it, but this
    is already a "fucked up" chain rule:

    reCx (M(x) raA S(x))
    reCx (G(x) raA M(x))
    roCroCroCroCroCroCroCroCroCroCroCroCroCroCroCroCroC
    re| reCx (G(x) raA S(x))

    Its just various camps fighting world domination:

    Everybody Wants To Rule The World
    https://www.youtube.com/watch?v=uqItaKDc6oI

    And the winner in the P(doom) bottom race is:

    CEO Jensen Huang: Zero Probability P(doom) in 2030
    https://www.youtube.com/watch?v=XlY7DI5v4ZU

    Bye

    (*)
    History of formal logic - III. Logic of Terms
    Bochenski, Joseph M. - 1961
    https://archive.org/details/historyofformall00boch/page/210/mode/2up

    Mild Shock schrieb:
    Hi,

    Woa! The public now marveling at the cascade
    LLM, LRM and now Swarms. And little software
    agents posting PERMADEATH.

    How it started:

    The co-evolution of AI and systems with Lidong Zhou
    https://www.youtube.com/watch?v=FVuJuGwkSu0

    How its going:

    This Is the AI Agent Story That Scared Me
    https://www.youtube.com/watch?v=2aw3MF8pY3w

    They are all clueless that Strong AI, requires
    to go from in vitro , to in vivo, usually called
    situated AI. Applies also to swarms:

    "Being situated is generally considered to be part
    of being embodied, but it is useful to consider
    each perspective individually. The situated
    perspective emphasizes that intelligent behavior
    derives from the environment and the agent's
    interactions with it. The nature of these
    interactions are defined by an agent's embodiment."

    https://en.wikipedia.org/wiki/Situated_approach_(artificial_intelligence) >>>

    And emergence of private languages is nothing new.
    I am pretty sure the interested reader will find
    some 30 year old papers of in vitro simulations,

    of software agents developing pebbel languages.

    Have Fun!

    Bye



    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Mild Shock@janburse@fastmail.fm to sci.logic,comp.lang.prolog,sci.physics on Sun Nov 15 12:39:10 2026
    From Newsgroup: sci.physics

    Hi,

    How it started, a Clown repeating the late Edsger Dijkstra:

    The question of whether a computer can think is no
    more interesting than the question of whether
    a submarine can swim.
    -- Edsger Dijkstra

    "You want it to be brilliant, but itrCOs not capable of being
    brilliant because itrCOs not capable of having any insight
    at all. ItrCOs just trying to follow patterns." https://www.hobartpulp.com/web_features/submarines-don-t-swim-a-conversation-with-mathematician-kevin-buzzard

    How it is going, a Clown admitting his defeat:

    Axiom Prover (Axiom Math) 40/50 80.0%
    Annals of Mathematics - NeurIPS 2026
    https://arxiv.org/pdf/2610.18301

    AxiomProver is an autonomous, multi-agent AI theorem
    prover developed by Axiom Math (founded and led
    by CEO Carina Hong). Benchmarked on the Annals

    of Mathematics, not the 246 Theorem, but leaves behind
    a few other systems currently, like Humanifa, NEAR AI,
    Tau and Midas.

    Bye

    P.S.: About the 246 Theorem
    "This is not AxiomProverrCOs first rodeo. Axiom Math has used
    its autonomous, multi-agent system that turns mathematical
    statements into machine-checkable proofs to crack several

    unsolved mathematical problems and verify many
    more proofs this year." But proof formalization of the 246
    theorem is by far the most significant, as Ken Ono, Axiom

    MathrCOs founding mathematician, explains: rCLThis theorem
    currently represents the threshold of human knowledge
    about prime numbers.rCY https://spectrum.ieee.org/axiom-math-246-theorem-formalization

    Mild Shock schrieb:
    Hi,

    How its started:

    More Math, Less Ego (talk about AI and Math)
    Erik Demaine - 29.10.2026
    https://www.youtube.com/watch?v=Gdc1yll04qI

    How its going:

    Classification of Finite Simple Groups
    Samuel Alexander - 28.10.2026
    https://www.youtube.com/watch?v=NZklSBoLC30

    Bye

    Mild Shock schrieb:
    Hi,

    The post is only proof of the Terence Tao
    ingestion problem hypothesis. That we are not
    anymore able to process content, and that
    even trust is destroyed.

    The AI slop presented here as a rCLsaferCY prolog
    agent toolkit, halucinates things like. And it
    does so consistently across many pages, still it
    wants to provide rCLsafetyrCY:

    ### Bakage (`bakage.toml`)
    The package manifest format used by Scryer Prolog to
    declare project metadata, exported modules, and dependencies.
    https://github.com/dougransom/prolog-agent-toolkit

    While I donrCOt find any evidence for such a
    manifest file. Its rather:

    A package is a directory with a scryer-manifest.pl
    following this schema: etc..
    https://github.com/bakaq/bakage

    That the post has even likes from field experts
    like Moura from Logtalk and Torbjorn from Web
    Prolog, shows that a) you can post any nonsense
    on GitHub people

    will clap, and b) this forum is useless only
    spreading AI slop. So while everybody is talking
    about provenance and justification, as a desiderata,
    the prolog agent toolkit is

    just an example of unhinged free wheeling analogical
    reasoning, probably taking phrases such as cargo.toml
    as a blue print and projecting them into phrases
    such aa bakage.toml.

    I still think a tool breaks when its weakest
    chain link break, even if people take this as a
    joke of mine. But well you are all free to give it
    a try, like running AI homunculi as a new sport.

    Bye

    (*)


    Terence Tao
    I have recently made the point that out of three
    major components of the mathematical problem
    solving process - proof generation, proof
    verification, and proof digestion
    https://mathstodon.xyz/@tao/116551624228986501

    P.S.: Such repos are useless if there is no way
    to cockpit the original agentic AI used. You would
    need to go through the route of tickets and GitHub
    issues to fix problems.

    Basically subscribing to the open source dream
    of Linux, that many eyes can flatten out bugs. But
    this is contra-productive to use strong tools and
    then wait for weak humans!

    Mild Shock schrieb:
    Hi,

    The idea that mathematics is based on a universal bowel
    movement system, that needs to learn and use "right and
    natural" math, is already disproved by our Indian vegetarian

    visitors here, that can even not digest Swiss Cheese. So
    lets declare repeating a canon a dead skill in the age of
    LLMs, and dig deeper what intelligence could mean.

    There is a thesis and ongoing product development:

    On the Measure of Intelligence Fran|oois Cholle author
    of the notable book "Deep Learning with Python"
    https://arxiv.org/pdf/1911.01547

    Ndea is building frontier AI systems that blend intuitive pattern
    recognition and formal reasoning into a unified architecture.
    https://ndea.com/about

    What about putting "right and natural" aside. And train
    mathematical intuition? Maybe the pressure of the moon
    race and cold war is lacking? And P(doom) is even

    the wrong wakeup call, but a genius pressure cooker!

    LoL

    Bye

    Mild Shock schrieb:

    Hi,

    Even Philip Wadler sounds like he has seen the
    first Apocalyptic horse rider. But what do these
    hardcore formalists think, that we go with a

    2300-year old Aristotelian inferencing like Modus
    Barbara (*) into the next decade of 2030?

    All humans are mortal
    All Greeks are humans
    --------------------------------
    Therefore all Greeks are mortal

    Or as the formalists like to write it, but this
    is already a "fucked up" chain rule:

    reCx (M(x) raA S(x))
    reCx (G(x) raA M(x))
    roCroCroCroCroCroCroCroCroCroCroCroCroCroCroCroCroC
    re| reCx (G(x) raA S(x))

    Its just various camps fighting world domination:

    Everybody Wants To Rule The World
    https://www.youtube.com/watch?v=uqItaKDc6oI

    And the winner in the P(doom) bottom race is:

    CEO Jensen Huang: Zero Probability P(doom) in 2030
    https://www.youtube.com/watch?v=XlY7DI5v4ZU

    Bye

    (*)
    History of formal logic - III. Logic of Terms
    Bochenski, Joseph M. - 1961
    https://archive.org/details/historyofformall00boch/page/210/mode/2up

    Mild Shock schrieb:
    Hi,

    Woa! The public now marveling at the cascade
    LLM, LRM and now Swarms. And little software
    agents posting PERMADEATH.

    How it started:

    The co-evolution of AI and systems with Lidong Zhou
    https://www.youtube.com/watch?v=FVuJuGwkSu0

    How its going:

    This Is the AI Agent Story That Scared Me
    https://www.youtube.com/watch?v=2aw3MF8pY3w

    They are all clueless that Strong AI, requires
    to go from in vitro , to in vivo, usually called
    situated AI. Applies also to swarms:

    "Being situated is generally considered to be part
    of being embodied, but it is useful to consider
    each perspective individually. The situated
    perspective emphasizes that intelligent behavior
    derives from the environment and the agent's
    interactions with it. The nature of these
    interactions are defined by an agent's embodiment."

    https://en.wikipedia.org/wiki/Situated_approach_(artificial_intelligence) >>>>

    And emergence of private languages is nothing new.
    I am pretty sure the interested reader will find
    some 30 year old papers of in vitro simulations,

    of software agents developing pebbel languages.

    Have Fun!

    Bye




    --- Synchronet 3.22a-Linux NewsLink 1.2