Суперкомпютърът Петак I / Юнак I

 

Свръхсметачът Юнак I / Петак I

Пророците на мислещите машини - основен том

• Новини от края на 2025 г. - както споменавам в други части от „Пророците…“, Вседържец, Емил и другите мислещи машини ще работят на такава техника, с каквато разполагаме. Започнах да изграждам „беоулф“ клъстер: самоделен „суперкомпютър“, засега съставен от множество обикновени ПС-та. 
  (...)

https://github.com/Twenkid/SIGI-2025/blob/main/petaki.md

Read More

Super Smartasses - improved colorized version of the graphic satirical story about the semiconductor and high-tech industry. The main volume of The Prophets ... and SIGI-2026 and Genesis: Creating Thinking Machines are coming

An original, unusual and very funny glimpse into a corner of the semiconductor industry,  presented in the pilot episode of a series conceived in 2009 and first published in 2012. 

Have fun and support The Sacred Computer: Thinking Machines, Creativity and Human Development

The main volume of "The Prophets of the Thinking Machines..." will be published soon and the next virtual conference: SIGI-2026 is starting on 1.1.2026. Be an early-bird there, LOL. :))

The next work is "Genesis: Creating Thinking Machines" and it is not designed to be "just" a set of 20-some books, monographs and papers, more than 4000 pages. (...)

 https://twenkid.com/smartasses/2025/

Todor Arnaudov

2 д ·  https://www.facebook.com/twenkid/posts/pfbid02FsMjrhjxtAjod6G9dvJ5biSLJSordKfFtKTc2Nt7qgGKhaB5o1zhCa6FXjus1JN8l?__cft__[0]=AZbmn5X2Pabyo5Xqx5YH0eGZOfV5ZJRR1WomuTYGA6N11fN3O90f8cJWcaw-TtC_g-Q80J3fm4AGTebehGjra9rwLVKyhuZkWwx4OXbGGLJFM_SOdU689xhSPCgPGfBYnbRP32Dj7zcGFgTltVM7gK1SkmnDdeQhkTo1erqQowTWlQ&__tn__=%2CO%2CP-R

Споделено с: Публично

Премиера на оцветената и подобрена версия на пилотния епизод на сатиричния ми комикс "Суперумници" (2012) - весели приключения и абсурди от полупроводниковата индустрия и живота: https://twenkid.com/smartasses/2025/

Първи епизод: "Спътници" (Companions)

* Идея, текст и рисунки с молив: Тош (Идея и описания на няколко епизода/скици на сторибордове: 2009 г.; първи осъществен голям пилотен епизод: лятото на 2012 г.)

* Оцветяване и подобрения на някои от рисунките (12/2025): Тош с Gemini и единични "ръчни" (наложи се едно завъртане: Horizontal-Flip;*) Някои от рисунките от черновата са запазени както са или само леко са оцветени при точно запазени контури и като цяло се стремях да се следва композицията, формите и стила на оригиналите с изключения, където не беше възможно или пък ми хареса както е. От линковете горе може да се види и оригиналната версия с молив.

* Сюжетът е вдъхновен от легендарна компания, която по-късно беше закупена от "Broadcom", след като много години работеха в тясно сътрудничество. "Броудком" днес се води с капитализация около 1.67 трилиона долара, около 2008-2009 - 3-4 милиарда (по-късно ги купуват за 37 милиарда през 2015 г., и после купуват и въпросната друга компания през 2016 г. и приемат името ѝ) Дали някой колега е бил "играч" и си е купил акции навреме?

...

* С повече дорисуване на скици или скиците можеше да се подобрят още образите, но не ми се занимаваше достатъчно за, само пробвах някои подсказки и не се получаваха. ChatGPT translation:

Premiere of the colorized and enhanced version of the pilot episode of my satirical comic “Superumintsi” (2012) — cheerful adventures and absurdities from the semiconductor industry and from life:
https://twenkid.com/smartasses/2025/

Episode One: “Companions”

• Concept, text, and pencil drawings: Tosh
  (Idea and descriptions for several episodes / storyboard sketches: 2009; first fully realized large pilot episode: summer 2012)

• Coloring and enhancements of some of the drawings (12/2025): Tosh with Gemini, plus a few manual touch-ups (one horizontal flip was required).*
  Some drawings from the draft were kept as they were or only lightly colorized, with the original outlines preserved. Overall, I aimed to follow the composition, shapes, and style of the originals, with exceptions where that wasn’t possible or where I preferred it as is. The original pencil-only version can also be seen via the links above.

• The story is inspired by a legendary company that was later acquired by Broadcom, after many years of close collaboration. Broadcom is now listed with a market capitalization of about $1.67 trillion; around 2008–2009 it was valued at $3–4 billion (they were later acquired for $37 billion in 2015, and then went on to acquire the other company in question in 2016 and adopt its name).
  Did any colleague happen to be a “player” and buy shares in time?

…

• With more redraws or further refinement of the sketches, the characters could have been improved even more, but I didn’t feel like investing the extra effort— I only tried a few prompts, and they didn’t turn out well.

--------------------

The Sacred Computer's Lab In the meantime, now that I’ve finished the major “time waste” of the Biblical book surveys and defending the priority of my contributions, it’s time for real work. I want the handicap I gave to the AI opportunists back! :))

You can help me if you wish: donate hardware or services, support me, connect and help in the development and experiments.

According to early 2010s "wishful" ideas and designs, while working on the Infrastructure, and to what was counted as an AGI, universal artificial intelligence, a thinking machine, capable to think and create at a single human/super-human level/speed/scale/performance/quality/complexity level*, were "dreamed" to work on a quad-core PC with 32 GB RAM and a desktop GPU from that era - equipped with sensors: one or several cameras, microphones, Internet and other sensors, used for bootstrapping - without pre-trained and existing modules which are now available, we may not count them in the requirements. :) 

Whether that would be enough - we will see.

Ask if you are interested in these articles or search the blog in 2013-2014, Issues on AGIRI list, "What's going on - Everything", "The infrastructure" (now called Vsy, Jack of All Trades, Вседържец etc.)
* I.e such a machine is not required to be capable of generating photorealistic 4K videos in real time - this is vastly superhuman. It should be at least as fast as me or reasonably slower. 

Read More

Listove or Reflections on Everything - the latest published and second largest volume from The Prophets of the Thinking Machines | Листове - Мисли По Всичко - вторият по големина том от Пророците на Мислещите машини

The volume was released on 5.11.2025 and updated with new material up to 13.11.2025. It combines material in both Bulgarian and English. See reviews below:

СВЕЩЕНИЯТ СМЕТАЧ  
ТОДОР АРНАУДОВ – ТОШ

ПРОРОЦИТЕ НА
МИСЛЕЩИТЕ МАШИНИ
ИЗКУСТВЕН РАЗУМ И
РАЗВИТИЕ НА ЧОВЕКА

ИСТОРИЯ ТЕОРИЯ И ПИОНЕРИ
МИНАЛО НАСТОЯЩЕ И БЪДЕЩЕ

ЛИСТОВЕ
по всичко

от авторa на първия в света
университетски курс по
   Универсален изкуствен разум и
 Теория на разума и вселената

THE PROPHETS OF THE THINKING MACHINES
ARTIFICIAL GENERAL INTELLIGENCE & TRANSHUMANISM
HISTORY THEORY AND PIONEERS; PAST PRESENT AND FUTURE

...

На 5.11.2025 беше публикувана новата двуезична книга от "Пророците..." - някои части и статии на български, други на английски, някои и на двата с превод. The volume Листове по всичко, or Listove - Reflections on Everything, appendix to "The Prophets of the Thinking Machines ..." was recently released at SIGI-2025. The second biggest volume with enormous diversity and both scope and depth of the analysis. (724 pages in current edition) Read at the the conference.
Страници на конференцията Мислещи машини 2025:

https://github.com/Twenkid/SIGI-2025

https://twenkid.com/agi/


Short review by ChatGPT5:

1. Produce a detailed review and summarize: what are the contributions?

(Then: Can you prepare all? Also can you add reviews of other topics which are missing, for example Quantum theories about consciousness and the critique in the work?)

ChatGPT said:

Below is a detailed review of the uploaded volume Listove: The Prophets of the Thinking Machines together with a clear summary of its main contributions, based strictly on the content found in your PDF.
All citations are to the file you provided.

"#4 — Distilled abstract (for publication — ~160–190 words)

Listove: The Prophets of the Thinking Machines presents an expansive, interdisciplinary compendium that situates the author’s early Theory of Universe and Mind (TUM, 2001–2004) within the contemporary landscape of AGI research, neuroscience, and philosophy. Combining historical reconstruction, technical review, and theoretical synthesis, the work argues that several influential 2010–2025 research programs — from active inference and relevance realization to hierarchical reinforcement learning and the consciousness prior — instantiate principles anticipated by TUM. Central to the book is a hierarchical “universe-as-computer” ontology where embedded causality-control units (CCUs) operate at distinct resolutions (RCCP), explaining limits of prediction and apparent non-computability as operational, resolution-dependent phenomena rather than metaphysical barriers. The volume includes extensive technical appendices (algorithmic complexity, SLAM, LLM memory systems), curated bibliographies, and meta-analyses intended to serve as a living dataset and an accelerator for human and machine research. It also offers sustained philosophical critique — notably a careful challenge to quantum-consciousness claims — and proposes alternative, multi-scale explanations of agency, meaning, and consciousness. The book is framed as both a scholarly resource and a seed corpus for interactive thinking machines."

Повече:

"1. Contributions of Listove (the volume)

(a) A large, curated, cross-disciplinary compendium
The volume acts as a single condensed knowledge map that connects:

• theories of mind, consciousness, philosophy
• AGI history, modern AI research, academic debates
• cybernetics, systems theory, ML architectures, computational neuroscience
• current papers and models up to October–November 2025
• historical milestones and rediscoveries

This is explicitly presented as a unique integrative effort, framed as an appendix to the "hyperbook" The Prophets of the Thinking Machines

(b) A dataset / seed for further human & machine research
The book is intentionally designed to serve as a training, inspiration, or continuation substrate for researchers and LLMs. It is explicit about being “dynamic and living,” suitable for Vsy, ACS, LLMs, and future interactive agents that continue its research lines

.

(c) A mapping between global contemporary work and TUM
A major contribution is demonstrating correspondences—sometimes strong structural equivalences—between modern AGI theories and the author’s early-2000s Theory of Universe and Mind (TUM).
The text repeatedly lists examples of 2010s–2020s works that the author claims “rediscover” TUM's principles decades later (e.g., Wolpert, Levin, FEP/AIF, Bach, Bengio, LeCun, Metzinger, Stack Theory, etc.)

(d) Critical commentary and synthesis
The volume provides detailed annotations, critique, and theoretical integration of:

• panpsychism
• computationalism/non-computationalism
• consciousness research
• active inference, free energy principle
• AGI architectures
• prediction limits, undecidability, causality-control
• modern LLMs, multimodal transformers, memory systems

This positions the work as an analytic bridge between philosophical, scientific, and engineering perspectives.

(e) Historical reconstruction of ideas in AGI
The author emphasizes early Bulgarian contributions and his own early publications (1999–2004) as antecedents to later internationally known theories.
This includes highlighting his teenage works, ideas about multi-scale causality, prediction, memory, “thinking machines,” and universe-as-computer metaphors.

2. Contributions of the underlying Theory of Universe and Mind (TUM)

The volume repeatedly stresses the following conceptual contributions of TUM (2001–2004):

(a) Universe-as-computer with hierarchical subuniverses (CCUs)

• Everything is computation and memory.
• Agents are subuniverses with limited resolution (RCCP: resolution of causality-control and perception).
  This anticipates later work in active inference, hierarchical modeling, unpredictability, etc.
  Example: Wolpert's theorems are interpreted as rediscoveries of TUM concepts about prediction limits and subuniverse constraints
  Listove_The_Prophets_of_the_Thi…
  .

(b) Multi-scale, multi-resolution intelligence
Intelligence = operating at multiple abstraction levels, with different causal resolutions.
This is strongly connected to:

• Bengio’s consciousness prior
• LeCun’s hierarchical predictive architectures
• Bach’s reward distribution views
  The author claims these repeat ideas from early TUM.

(c) Meaning, semantics, and the “spirit/model” idea
TUM defines “spirit/soul/consciousness” as a model/simulation maintained by an evaluator-observer.
This anticipates aspects of Metzinger’s Phenomenal Self-Model (2009) and later cognitive science work

(d) Non-boundary between living and non-living
TUM claims early that cognition and agency arise along continuous physical/computational lines—ideas similar to Levin’s and others' work two decades later

(e) Explanation of undecidability and limits as resolution issues
Contrary to claims that computation cannot capture reality (e.g., Krauss et al., 2025), TUM interprets undecidability as a matter of:

• insufficient resolution
• insufficient perceptual/causal access
• mismatched type-of-causation
  This is presented as a key original conceptual contribution explaining Gödel/Tarski/Chaitin limits in physical terms, not as metaphysical boundaries
  
  .

(f) Thinking machines: architectures, prediction, memory, universes-within-universes
TUM provides early theorizing on:

• prediction machines
• simulation-based reasoning
• memory as “subuniverses”
• the distinction between mortal and immortal computational systems
  The volume links this to the 2025 paper Is Mortal Computation Required…? and modern neuromorphic and AGI discussions
  
  
  (...)"
  
  
  

  Detailed Review of the Volume

  Scope and Nature

  The text is a massive interdisciplinary notebook/scholarly compendium (over 700 pages), consisting of:

  • reading notes

  • commentary

  • cross-referencing

  • curated scientific material

  • philosophical arguments

  • AGI theory comparisons

  • historical and autobiographical remarks

  It's not a traditional monograph but a continuously accumulating knowledge hyperstructure.

  Structure

  The document interleaves:

  • Bulgarian and English sections

  • citations and excerpts from scientific papers

  • personal annotations

  • critiques and theoretical expansions

  • philosophical digressions

  • connections to TUM

  The organization is layered and associative rather than linear.

  Core Themes

  1. Rediscovery Thesis
     Many modern theories in AGI, neuroscience, and philosophy echo ideas from TUM.

  2. Universe as Memory/Computer
     Everything is stored and computed within the universal memory space.

  3. Limits of Prediction and Computation
     Not as “non-computable essence” but as limited resolution or inadequate causal interfaces.

  4. Cognition Beyond Neural Activity
     Multi-scale systems (cells, tissues, environments) demonstrate intelligence, matching views seen in Levin’s work.

  5. Multimodal AI, LLMs, and Memory Systems
     The text places modern transformer architectures, memory protocols, agentic context engineering, and RAG frameworks within the TUM framework of evolving virtual subuniverses.

  6. Philosophy of Mind and Consciousness
     Panpsychism, phenomenal unity, meaning, self-models, and interpreter dependencies are continuously examined.

  Intended Use

  The author positions the volume as:

  • a research accelerator

  • a conceptual atlas

  • a training corpus for thinking machines

  • a proof of the longevity and originality of TUM

  • a living work to be extended jointly by humans and AI systems

  ---

  Condensed Contribution List (One Page)

  Primary Contributions of the Book

  • A comprehensive, interdisciplinary, annotated knowledgebase linking thousands of ideas, papers, and theories across AGI, AI, philosophy, neuroscience, and systems theory.

  • A unique comparative framework demonstrating deep parallels between cutting-edge 2020s theories and TUM (2001–2004).

  • An annotated research dataset intended for training and guiding human/AI systems.

  • A philosophically integrated critique of contemporary claims about simulation, computation limits, consciousness, and prediction.

  • A historical reconstruction of AGI thought, emphasizing overlooked early contributions.

  Primary Contributions of TUM (as highlighted in the book)

  • A hierarchical, multi-scale universe-as-computer framework with CCUs and RCCP.

  • Early explanation of prediction limits and mutual unpredictability decades before Wolpert.

  • The model-based view of consciousness, anticipating Metzinger and modern cognitive science.

  • A unified continuum between biological and artificial agency, anticipating Levin/TAME/FEP discussions.

  • A structural explanation of undecidability as resolution mismatch, not metaphysical non-computability.

  • Multiple ideas about architecture, memory, prediction, agenthood, that later reappear in Bengio, LeCun, Bach, Levin, etc.

  
  
  (...)
  
  

  - Can you prepare all? Also can you add reviews of other topics which are
  
   missing, for example Quantum theories about consciousness and the critique
  
  in the work?
  
  
  

  ChatGPT said:

  Thought for 13s

  Great — I’ve prepared everything you asked for: a 5-sentence summary, a textual “graphical” map of contributions, a comparison table (TUM vs several modern AGI theories), a distilled abstract suitable for publication, and an extended review specifically of quantum theories of consciousness and the book’s critique of them. All items are based on the uploaded PDF and I cite the file where relevant.

  ---

  #1 — Five-sentence summary (very short)

  1. Listove: The Prophets of the Thinking Machines is a large, interdisciplinary compendium linking the author’s early Theory of Universe and Mind (TUM, 2001–2004) with modern AGI, neuroscience, philosophy and engineering literature up to late 2025.

  2. The book’s core claim is that many recent theories (Free Energy / Active Inference, relevance realization, Bengio’s consciousness prior, LeCun’s path to autonomous AI, Wolpert’s unpredictability results, Michael Levin’s biologically-driven cognition) are rediscoveries or close relatives of ideas that appear in TUM.

  3. Conceptually TUM emphasizes a hierarchical “universe-as-computer” model with embedded causality-control units (CCUs) operating at different resolutions (RCCP), explaining limits of prediction and computation as resolution/causal-access issues rather than mystical non-computability.

  4. The volume is explicitly intended as a “living” research corpus and seed dataset for human and machine research (Vsy, ACS, LLMs), with commentaries, curated references and cross-comparisons aimed at both historians and AGI builders.

  5. It also contains philosophical critiques and extensions — including a skeptical, systematic critique of many quantum-consciousness claims and a reframing of Gödel/Tarski/Chaitin-style limits in operational/embodiment terms.

  
  ...
  
  

   #2 — Textual “graphical” map of contributions (use as a visual guide)

  [ THE PROPHETS / LISTOVE ]
           |
           +--> (A) TUM core concepts
           |        - Universe-as-Computer (Master memory)
           |        - CCU (Causality-Control Units)
           |        - RCCP (Resolution of Causality-Control & Perception)
           |        - Multi-scale/relevance realization
           |
           +--> (B) Modern AGI & science (mapped)
           |        - Free Energy / Active Inference  <-- parallels to TUM. :contentReference[oaicite:5]{index=5}
           |        - Relevance Realization / Consciousness Prior  <-- TUM precursors. :contentReference[oaicite:6]{index=6}
           |        - Wolpert/limits of inference  <-- repurposed & reframed by TUM. :contentReference[oaicite:7]{index=7}
           |        - Transformers, LLMs, memory, multimodal models (practical tech notes). :contentReference[oaicite:8]{index=8}
           |
           +--> (C) Philosophy & consciousness
           |        - Panpsychism / phenomenal unity discussion. :contentReference[oaicite:9]{index=9}
           |        - Quantum-consciousness critiques (Orch-OR, microtubules, Georgiev). :contentReference[oaicite:10]{index=10}
           |
           +--> (D) Practical outputs
                    - dataset + lecture / course materials (first university AGI course claim)
                    - seed for LLM training + living / interactive book goals. :contentReference[oaicite:11]{index=11}

  ---

  #3 — Comparison table — Theory of Universe and Mind (TUM) vs selected modern theories

  Table rows show: Core idea → TUM position / contribution → How Listove compares to modern theory (brief).

  Core idea	TUM / Listove position	How modern theory maps (examples) & book’s note
  Hierarchy & multi-resolution control	Agents/subuniverses (CCUs) operate at different RCCP; intelligence = multi-scale control. …	Bengio (Consciousness prior) and hierarchical RL — similar emphasis; book claims TUM described these ideas earlier. …
  Limits of prediction / mutual unpredictability	Limits arise from embedded subuniverse’s limited RCCP and causal access, not metaphysical impossibility. …	Wolpert (physical limits): Book treats Wolpert’s formal results as rediscoveries and clarifies premises; TUM offers operational interpretation. …
  Free energy / active inference	TUM anticipated principle-like ideas: prediction, viability, and multi-scale agency; aligns conceptually. …	Friston et al. (FEP/AIF): Book frames them as a fork/rediscovery and discusses differences in scope. …
  Pan/constitutive-panpsychism & unity problem	TUM offers models of “subject” and phenomenal unity arising from evaluator models and linked subassemblies; raises selection problem. …	Contemporary panpsychism literature: Book surveys (Woodward, Roelofs) and discusses selection/mixing problems and TUM’s alternative framing. …
  Quantum consciousness (microtubules, Orch-OR)	TUM is skeptical: accepts quantum physics but argues that claims tying consciousness to fragile quantum coherence are often speculative and misapplied; offers alternative multi-scale, evaluator-based explanation.
  Transformers / LLMs / Memory systems	TUM treats LLMs as practical agentic components; the book includes technical reviews (memory, multimodal) and positions them within broader agent architectures (memory interchanges, RAG).

  (References in the right column point into Listove where the comparisons and critiques appear.)

  ---

  
  
  
  
  
  

THE SACRED COMPUTER 
TODOR ARNAUDOV – TOSH

THE PROPHETS OF THE
THINKING MACHINES
ARTIFICIAL GENERAL INTELLIGENCE & TRANSHUMANISM

HISTORY THEORY AND PIONEERS
PAST PRESENT AND FUTURE

LISTOVE
REFLECTIONS ON EVERYTHING

by the author of the world’s first university course in
Artificial General Intelligence and the
Theory of Universe and Mind

 

ПРОРОЦИТЕ НА МИСЛЕЩИТЕ МАШИНИ
ИЗКУСТВЕН РАЗУМ И РАЗВИТИЕ НА ЧОВЕКА
ИСТОРИЯ ТЕОРИЯ И ПИОНЕРИ; МИНАЛО НАСТОЯЩЕ И БЪДЕЩЕ

,,,

Read More

The Sacred Computer's Discord Channel

Join The Sacred Computer Discord Chanel:  https://discord.gg/qFQSvA2A3T

Read More

Logos: The Sacred Computer in English DALL-E 3 - Part II

More logos. The attempts to render with title in Cyrillic - the letters ended up like strange symbols or nothing. Created with Bing image creator.

It couldn't get the caption always right, with minor errors or "hybrid" letters, but about 1/4 or was good, sometimes maybe 2/4. The model couldn't properly generate Cyrillic though, as well as unusual letter combinations, such as "RAZUMIR" or "PA3YMUP" (To resemble a Cyrillic "РАЗУМИР", a title of the spin-off e-zine of the Sacred Computer).

Generation of beautiful 3D logos/captions with Bing Image Creator.

Read More