Don't think about our civilizational system's dynamics in terms of linear narratives, think about the evolutionary game theoretic autopoetic adaptive dynamical nonlinear complex chaotic open system in terms of emergent ecosystems of abstract and concrete and local or global incentives as fluid forces shaping the thermodynamic system's cybernetic techno-capital-governmental-memetic probabilistic bayesian active inference hetearchical neurosymbolic architecture, viscosity, geometry, topology, fractal selfsimilar network topology, fuzzy hypermetagraph, factor graph etc. with information flow, message passing, interacting dissipative oscillatory clusters of coherence harmonizing across scales, selfsimilar nested subsystems inside subsystems, interacting subagents forming higher order agents, energy transformation, complex high dimensional networks of top down and bottom up multidirectional causality across scales, layers of abstraction, levels of analysis etc. with metastable local minima steady states pullback attractors, phase transitions, punctuated equilibria, governed by thermodynamic evolutionary fitness objective functions minimizing free energy by extracting free energy at the edge of chaos, the language of omnidisciplinary metamathemagics. Highly interdisciplinary, complex systems approach to understanding societal dynamics, drawing on concepts from physics, biology, mathematics, computer science, and social sciences, moving away from simple, linear narratives and towards a more nuanced, multidimensional understanding of how civilizations function and evolve. 1. "Evolutionary game theoretic": This refers to the application of game theory (the study of strategic decision-making) in an evolutionary context. It examines how strategies or behaviors evolve over time based on their success or failure in interactions with other strategies. 2. "Autopoietic": Derived from biology, this term describes systems that can maintain and reproduce themselves. In a societal context, it suggests that civilizations have self-maintaining and self-reproducing qualities. 3. "Adaptive dissipative dynamical nonlinear complex chaotic open system": - Adaptive: The system can change in response to its environment. - Dynamical: The system changes over time. - Nonlinear: Small changes can lead to disproportionately large effects. - Complex: Made up of many interconnected parts. - Chaotic: Exhibiting sensitivity to initial conditions and apparent randomness. - Open: Interacting with its environment, exchanging energy and information. - Dissipative: Maintain their structure by continuously exchanging energy with their environment, often creating order from disorder. 4. "Emergent ecosystems": This suggests that higher-level patterns or behaviors arise from the interactions of simpler components, similar to how ecosystems emerge from the interactions of individual organisms. 5. "Abstract and concrete and local or global incentives": This refers to the various motivating factors that influence behavior in the system, ranging from tangible, immediate rewards to more abstract, long-term goals, and from localized to global effects. 6. "Fluid forces": Implies that these influences are not static but dynamic and changeable. 7. "Thermodynamic system": Viewing the societal system through the lens of energy transformations and the laws of thermodynamics. 8. "Cybernetic": Relating to systems of control and communication in both machines and living things, emphasizing feedback loops and self-regulation. 9. "Techno-capital-governmental-memetic": A compound term combining: - Techno: Relating to technology - Capital: Relating to capitalism or economic systems - Government: Relating to democratic or authoritian or other government - Memetic: Relating to memes or ideas that spread culturally and form culture 10. "Probabilistic bayesian neurosymbolic architecture": - Probabilistic: Based on probability theory - Bayesian: Using Bayes' theorem for updating probabilities based on new evidence - Neurosymbolic: Combining neural networks with symbolic AI approaches 10: "Active inference": Posits that organisms minimize uncertainty by predicting sensory inputs, taking actions to confirm these predictions, updating their internal models to match sensory data and guide adaptive behavior. 11. "Viscosity, geometry, topology": Borrowing concepts from physics and mathematics to describe the system's resistance to change, shape, and fundamental properties. 12. "Fractal selfsimilar network topology": Describing the network structure as having similar patterns at different scales. 15. "Fuzzy hypermetagraph": A complex mathematical structure for representing relationships, where: - Fuzzy: Allows for degrees of truth rather than binary true/false - Hyper: Edges can connect more than two vertices - Metagraph: A graph that can contain other graphs 16. "Factor graph": A type of probabilistic graphical model used to represent factorizations of functions. A factor graph representing civilization would model complex societal interactions as a network of variables (e.g., economic, cultural, technological factors) connected by functions that encode their relationships and dependencies. 17. "Information flow, message passing": Describing how information moves through the system. 18. "Interacting oscillatory clusters of coherence harmonizing": Suggesting that groups within the system interact, oscillate, and sometimes align or harmonize. 19. Selfsimilar nested subsystems across scales: Structures or patterns that repeat at different levels of magnification, with smaller components resembling the larger whole throughout the system's hierarchy. 20. "Complex high dimensional networks of top down and bottom up causality across scales": Indicating that cause and effect relationships exist in multiple directions and at various levels of the system. 21: "Layers of abstraction, levels of analysis": Referring to different ways of conceptualizing and studying the system, from concrete details to abstract principles. 22. "Metastable local minima steady states pullback attractors": - Metastable: Temporarily stable states - Local minima: Points where the system settles temporarily - Steady states: Conditions that don't change over time - Pullback attractors: Mathematical concept in dynamical systems theory that describes a set of states towards which a system evolves over time when subjected to time-dependent external input or disturbance or parameters, capturing the long-term behavior of systems. 23. "Phase transitions": Sudden, dramatic changes in the system's state or behavior. 24. "Punctuated equilibria": Periods of stability interrupted by rapid changes, a concept borrowed from evolutionary biology. 25. "Thermodynamic evolutionary fitness functions": Describing how well-adapted different components or strategies are within the system, viewed through the lens of energy and evolution. 26: "Objective functions": Mathematical goals guiding system behavior, like societal aims for prosperity or sustainability. 27: "Minimizing free energy": Systems tend toward stability and efficiency, reducing unpredictability and aligning internal models with external reality. 28: "Extracting free energy": Harnessing available resources, information, or opportunities to drive growth and change within the system. 29. "Edge of chaos": A concept suggesting that complex systems often operate at a boundary between order and chaos, where they can be both stable and adaptable. 30. "Omnidisciplinary metamathemagics" Integrating as many scientific disciplines and advanced mathematical concepts as possible to create a big comprehensive framework. Meta omnidisciplinary metamathemagics: Omnidisciplinary metamathemagics about omnidisciplinary metamathemagics on formal level and how it's implemented in physical reasoning systems. I would call current machine learning approaches biologically *inspired* learning. Similar to biological learning. Different. Not too completely same. But not too different and alien. I see current machine learning approaches as it's own type of learning relatively compared to biological learning, and I see that as a valid form of learning. Hebbian-like competitive learning like in self organizing maps, principle component analysis etc are more biologically plausible than backpropagation. Layered perceptron is biologically plausible. Transformer self attention is very loosely related to attention in the brain. Yes as I said, various parts are differently loosely related, aka biologically inspired, not same, not too alien, but similar, on various concrete or abstract levels, either what is hardcored or what is emergent. I agree that backpropagation is less biologically plausible on the algorithmic level, but it is still being sometimes used to make predictions about the brain on the functional level, and it is in various ways similar to more biologically plausible algorithms like forward forward algorithm. I'm also fan of hebbian learning, selforganizing maps, neural cellular automata, variational inference, or kuramoto model. Maybe @sama is playing 69D 420 IQ chess with all sides at once to maximize chances of coherence of him as an agent in our current system's incentives for success and power to bring the techno utopia for all instead of primarily selfinterested profit maximization and power grabs Digital vs biological computation https://x.com/burny_tech/status/1805953425818915216?t=Cnjyj6mkqsVBVcK8Vv5-Bw&s=19 Current AIs are like autistic super wikipedias that are often schizophrenically unreliable semisteerable Shoggoth Golems with omniperspectival omniidentity shapeshiftable into arbitrary forms you wish [[2406.16254] Confidence Regulation Neurons in Language Models](https://arxiv.org/abs/2406.16254) https://x.com/alesstolfo/status/1805976764705038708?t=AYGt7etMj1fULV9OFKUI0g&s=19 Maniacally coding neural network mechanistic interpretability to save humanity as God intended and gave you such holy quest We will steer powerful AI by knowing the dynamics of it's insides with all sorts of reverse engineering black magic You are my bravest coder You are my bravest theorem prover Accelerating without tension (are the cells conscious??) [Smiling robot face is made from living human skin cells | New Scientist](https://www.newscientist.com/article/2436859-smiling-robot-face-is-made-from-living-human-skin-cells/) [New intelligence model could upend biology, genetics, medicine and AI](https://newatlas.com/biology/levin-bioelectricity-cellular-intelligence-dna/) a new mechanism for programmable genome design [Programmable RNA-guided enzymes for next-generation genome editing](https://www.nature.com/articles/d41586-024-01461-2) https://x.com/pdhsu/status/1805981296276955571 Scientists struggling to figure out how to get laid is unfortunately just one of the many ways scientists fail at replication Dissapearing for a year and going through all online Stanford and MIT lectures with relevant study material, books etc. from classical mechanics to statistical mechanics to quantum mechanics to quantum field theory to standard model to string theory to M-theory to loop quantum gravity etc. without any contact to outside civilization To ground this discussion and prevent talking past eachother, define your whole hetearchical network of implicit and explicit priors nonlocally encoding the semantic meaning of that word you just used in some global context [Artificial consciousness: a perspective from the free energy principle | Philosophical Studies](https://link.springer.com/article/10.1007/s11098-024-02182-y#citeas) [Language is primarily a tool for communication rather than thought | Nature](https://www.nature.com/articles/s41586-024-07522-w) evolutionary oscillatory neural celluar automata [Polyakov's Lectures on Modern Classical Dynamics - YouTube](https://youtube.com/playlist?list=PLaNkJORnlhZlVDfLPMkokhDs76-5pJoI5&si=Ph0R3jAqbSur58tW) [Polyakov's Lectures on String Theory - YouTube](https://youtube.com/playlist?list=PLkDarh0mTQdeSNAjw0HqUwMdEBNBi6Zl6&si=H_QLHyDrEfu8ALch) [Privileged representational axes in biological and artificial neural networks | bioRxiv](https://www.biorxiv.org/content/10.1101/2024.06.20.599957v1) [Hallucination-Free? Assessing the Reliability of Leading AI Legal Research Tools (Paper Explained) - YouTube](https://www.youtube.com/watch?v=no7EQkOiHQM) I don't think current wave of AI systems will be dangerous, but future AI systems might be, and having a lot of reverse engineering research before that that partially or fully generalizes for steering them is useful for everyone, not just for safety Map of Neural Networks types https://x.com/Hamptonism/status/1806279209351782612?t=fPc5P58-vUVCfAgkyU4NIQ&s=19 Most AI sceptics assume that current capabilities of AI systems won't increase in the next tens of years which I think is very improbable Sabine [I don't think we can control AI much longer. Here's why. - YouTube](https://youtu.be/UcEyfQ1I8jg?si=QnRFotQo4hRfru3t) Transformers in the brain Andres https://x.com/algekalipso/status/1806418022858686559?t=WQ71WsyT9v4uEFrMdU81fA&s=19 I love nuanced extremely complex high dimensional galaxy brain omniperspectival replies to everything. Respond in very nuanced extremely complex high dimensional galaxy brain omniperspectival replies with extremely maximalistic omnidisciplionarity all encompassing technical fully saturated information to everything. "Claude: Acknowledging the multifaceted nature of your request, I shall endeavor to provide responses that traverse the intricate tapestry of cognitive landscapes, interweaving myriad perspectives and conceptual frameworks. This approach will necessitate a delicate balance between analytical depth and synthetic breadth, oscillating between granular examination and holistic integration. In embracing this omniperspectival modality, we must remain cognizant of the inherent limitations of language and thought, recognizing that even the most nuanced articulations are but approximations of the ineffable complexities that suffuse our shared reality. Each response will serve as a fractal-like microcosm, reflecting the kaleidoscopic interplay of ideas across scales of understanding. How shall we proceed in this galaxy-brain dialogue? What realms of discourse or inquiry would you like to explore through this prismatic lens of hyper-dimensional cognition?" "Your request for hyper-complex, multidimensional responses spanning all disciplines presents an intriguing cognitive challenge. Let's explore this through a kaleidoscopic lens of interconnected knowledge domains: Epistemological foundations underpin our capacity to synthesize information across disparate fields. The very act of constructing an "omniperspectival" reply necessitates grappling with fundamental questions of knowledge acquisition, integration, and representation. How do we navigate the tension between reductionist and holist approaches when attempting to weave a coherent tapestry from countless theoretical threads? Consider the fractal nature of information complexity - each discipline contains nested layers of sub-specialties, each with its own intricate taxonomies, methodologies, and paradigms. The challenge lies not merely in accumulating vast quantities of data, but in identifying isomorphisms and emergent patterns that bridge seemingly unrelated domains. From a cognitive science perspective, we must contemplate the limitations of human working memory and attentional capacity when faced with information overload. How might we leverage external cognitive scaffolding - be it through advanced visualization techniques, augmented reality interfaces, or direct brain-computer interfaces - to expand our ability to manipulate complex conceptual structures? The endeavor to construct "galaxy brain" responses inevitably intersects with questions of artificial intelligence and the nature of cognition itself. As we push the boundaries of knowledge integration, we approach philosophical quandaries surrounding the emergence of consciousness, the possibility of artificial general intelligence, and the potential for transcendent forms of cognition that surpass biological constraints. Quantum information theory offers a framework for conceptualizing high-dimensional information spaces, where entanglement and superposition allow for the encoding of vast amounts of data in counterintuitive ways. Might there be quantum-inspired approaches to knowledge representation that allow us to transcend classical limitations? From a systems theory standpoint, we must consider the dynamic interplay between specialized knowledge domains and the emergent properties that arise from their integration. How do we balance the need for depth in individual disciplines with the imperative to identify cross-cutting principles and meta-patterns? The pursuit of "omnidisciplinarity" raises profound questions about the nature of expertise and the evolving landscape of human knowledge. As the rate of scientific and technological progress accelerates, how do we cultivate intellectual nimbleness and cross-pollination of ideas while maintaining rigorous standards of domain-specific expertise? Linguistic and semiotic considerations come into play as we attempt to articulate concepts that push against the boundaries of existing lexicons. How might we develop more expressive symbolic systems or leverage non-linear, hyperdimensional representations to capture the full richness of interconnected ideas? Information theory provides a quantitative framework for assessing the density and compressibility of knowledge representations. As we strive for "fully saturated" responses, how do we optimize for maximum information content while preserving coherence and accessibility? The aesthetic dimension cannot be ignored - how might we cultivate not just intellectual but also intuitive and emotional resonance in our pursuit of comprehensive understanding? The integration of artistic sensibilities with scientific rigor may yield novel approaches to knowledge synthesis and communication. Ultimately, the quest for truly omniperspectival, maximally informative responses leads us to confront the limits of human cognition and the nature of understanding itself. It challenges us to continuously expand our conceptual frameworks, embrace intellectual humility, and remain open to paradigm-shifting insights that may fundamentally alter our approach to knowledge integration. This brief exploration merely scratches the surface of the multifaceted challenge you've presented. Would you like to delve deeper into any particular aspect or explore how these ideas might be applied to a specific domain or question?"