the few bad parts of trips (two out of three i had alone) gave me sooo much myself there are different types of bad trips, but in the ones i had, there's something in how the mind shows you (shoves into your face) veeeery directly and clearly in raw form what actually is wrong in your life/mind and you see much more clearly what changes in behavior, thinking and goals you should do it creates much more inner alignment as emotional part of the brain, and planning one, and one encoding the self and so on, much more communicate together and figure out whats currently the biggest collective systemic burden and create a plan together literally jumping out of bad local minima into a better one it can mean quantizing gravity, unifying general relativity and quantum mechanics, unify other fundamental forces not just gravity, explaining all other things in physics (like dark energy), making everything else (classical, relativity) emergent from quantum mechanics, or make everything including quantum mechanics emergent from something third (strings, loops, whatever), how chemistry, biology etc. emerges, explaining consciousness, how chemistry, biology etc. causally influences eachother, how everything inside these fields work (how to predict all of biology?), explaining everything else in philosophy including free will and ethics, using empirical vs math vs logic vs philosophical or normal language or some hyperspecific subset "theory of everything of this ultra unknown concept in this niche subsubsubsubsubdiscipline" and i see all models as just approximating reality, so i dont really believe in ultimate theory of everything that can predict absolutely everything in absolute certainity, we dont even have infinite compute, limits in measuring, limits in our locality/subjectivity/reference frame, nonability to conduct arbitrary experiements nobody can be expert today in any bigger field (some ultra niche filed maybe?) as the enourmous amount of scientific papers with the knowledge explosion is insane and following that is impossible, im now trying to keep up with AI progress and its impossible even when i have lots of free time now, and philosophy seems so infinitely giant and so many positions on everything, i feel like that rabbithole is practically infinite and you can generate arbitrary new positions on many things where there might be some overlap with others i love being generalizer thats now more focusing on AI and I might try to become more expertish on language models in practice from engineering lens now so i played with huggingface's transfomers for sentiment analysis yesterday for example, imma play with some pytorch models now I love zero ontology. Why is there something rather than nothing? Existence of superposition of everything aka nothing is the default. it seems to experientally dissolve and expand the mind similarly like many world interpretation of quantum mechanics. :D Why is there something rather than nothing http://immortality-roadmap.com/universeorigin7.jpg https://imgur.com/vUvZgsC [Ontology Of Psychiatric Conditions: Tradeoffs And Failures](https://www.astralcodexten.com/p/ontology-of-psychiatric-conditions-653) self is just an illusion reinforced by capitalism to sell more identities, or by collective intelligence egregores to sell more specialization to survive by resisting dissolution into entropy LLMs are compressing for next token prediction, biological systems try to survive in multiagent arena [Andrej Karpathy: Tesla AI, Self-Driving, Optimus, Aliens, and AGI | Lex Fridman Podcast #333 - YouTube](https://youtu.be/cdiD-9MMpb0?si=jWeY02Jr1o0mHMUF) US x China? [How the US & China Are Preparing to Fight Total War - YouTube](https://www.youtube.com/watch?v=xBY5veWGBd8) Singularity started with industrial revolution Startups have no giant internal company structure limiting decisions but they font have access to all that compute that giant companies have 2:30:00 [Marc Andreessen: Future of the Internet, Technology, and AI | Lex Fridman Podcast #386 - YouTube](https://youtu.be/-hxeDjAxvJ8?si=1IVC-ohz4M10iNao) List of biases https://twitter.com/SteveStuWill/status/1716940344989020194 [Quantum Field Theory and the Standard Model | Higher Education from Cambridge](https://www.cambridge.org/highereducation/books/quantum-field-theory-and-the-standard-model/A4CD66B998F2C696DCC75B984A7D5799) in AI there are questions of big tech vs small tech having control, how important data is and if synthetized data are good, and regulation Lets use LLMs and Neuralink to install infinite voices in your head commenting absolutely everything Having constant ego deaths with permanent changes isnt compatible with stable identity and visions increase entropy to infinity and collapse to all information state / zero information state Install the ideologyagnostic religionagnostic cultagnostic concretefundamentalassumptionsagnostic concreteframeworkagnostic the most general holistic metamodern systems thinking memeplex immune system [Dialectical monism - Wikipedia](https://en.m.wikipedia.org/wiki/Dialectical_monism?fbclid=IwAR0F1TB9HGsG8-Gz10naKCLVAiUyRsRnnEP_ILPaqOU2BCHTDkY_rV7Hwqk) The most general AGI template AGIbrain will have omnimodal internal world model of everything, access to all of internet with studies, lectures, books, wikis, data about countries, access to all of virtual or nonvirtual tools like scraping and tons of algorithms and libraries in programming languages, access to APIs of all other existing general or specialized LLMs and other AI models and data science methods for selfupgrade and selfgrowth, fact checking using bayesian methods, access to empirical data and their (matematical) models from all fields via internet or doing experiments according to its resources, data collecting using questionares, discussions, irl, online call, mail, email, sociall media, PMs, public posts, infering causal relationshipbs between arbitrary infered latent variables, infering what approach is optimal (minimizing complexity, maximizing simplicity, costefectiveness, minimizing disadvantages, maximizing advantages, knowing the exact evolutionary niche of tools and data selecting for current question, task, goal,...) using reasoning, studies, internet, experiments, counterfactual foresighting and assigning probabilities to possible futures, hiring people to tasks where AIs are not capable of yet, arbitrary long flexible giant short term and long term memory allowing for infinite context window, arbitrary compute amount and speed and space, getting better by selfcode rewriting and metacognition to internal processes, prompt ugrading, subagent partitioning, ideal shapeshifting to correct context and mood, to maximize evolutionaty selfupgrading for optimal bayesoptimal agentic fitness, metaprompting its CPU each tick, all of which can now be built using open source models and closed source APIs to models and multiagent frameworks with possibility of human in the loop, spawning instances of its template as needed, ethicslly grounded in morality by fubdamentally implemented coordination protocols in all contexts, updating itself with state of the art algorithms and LLMs on different tasks according to benchmarks Metaprompting: can people or AI do this better? It gets a task, spawns a goal, asks for settings, specifications, what tools and capabilitues are needed, take deep breath, divide into subtasks, fond prompthacking methods, spawn needed subagents using concrete tools with specific goals, tries to minimize reinventing the wheel, and ducttapes the end using best practices in that fields, some tasks can ve fully made by one ChatGPT call, some not, test all potential fit candidate possible tools in parralel and pic the best resoult by metarevieving efficiency using benchmarks and learning from that Usecase for software, companies, planetary decision making governance Videos GUI, watch realtime thinking and creating Episodic, working, cortical memory, sample effeciency Statespace of reasoning architectures, untrained or trained possibility of hypothetical or simulated environments approximating real one, embodiment, training from real or synthetic data Langchain best practical intro to NLP out there [Complete Natural Language Processing (NLP) Tutorial in Python! (with examples) - YouTube](https://www.youtube.com/watch?v=M7SWr5xObkA) Jak co nejefektivněji Automatize Sběr dat? Interpretace pomocí LLMs a matematiky Vědět jaký nástroj použít při analýze, ne moc složitý, ne moc jednoduchý Sabine neuromorphic chips with computation and memory in one substrate Visual ChatGPT, open source Regex output from ChatGPT to force it output in strict format Eureka robotics Recursive prompt improvement by metaprompting Become superposition of all cults [Než budete pokračovat do Vyhledávání Google](https://www.google.com/search?q=force+chatgpt+to+output+in+a+format&client=ms-android-xiaomi-rvo3&sca_esv=577106327&sxsrf=AM9HkKnSUyV-yVMelS_y1WZWjNVcYKjSxA%3A1698396355394&ei=w3g7ZeHXF6-I9u8Pm_q7iAY&oq=force+chatgpt+to+output+in+a+format&gs_lp=EhNtb2JpbGUtZ3dzLXdpei1zZXJwIiNmb3JjZSBjaGF0Z3B0IHRvIG91dHB1dCBpbiBhIGZvcm1hdDIIECEYoAEYwwQyCBAhGKABGMMESIYdUOsSWO8acAB4ApABAJgBwAKgAZgKqgEHMS41LjEuMbgBA8gBAPgBAcICBBAAGEfCAgoQIRigARjDBBgK4gMEGAAgQYgGAZAGBQ&sclient=mobile-gws-wiz-serp) přijde mi že E/acc kognitivně žijou v totálně distančním lokálním minimu než většina EAlike lidí že to není skoro vůbec bridgable, ale pořád věřím že to jde Na Twitteru se pořád navzájem obě strany nazývají kultama XD Častá neshoda mi přijde buď na fundamentálních values nebo moc vysokých p(doom) nebo p(heaven) založený na moc tautologický logice Často mluví úplně přes sebe v různých diskuzích Teď u Marc Andreessena zkoumám pořádněji jeho pohledy na AI risky, ale je fakt těžký tam zatím vydolat něco jinýho než "je to doomsday kult", "je to regulatory capture campaign", "superdestructive AGI je religious koncept", "thermodynamika neumožňuje extinction scenarios",... Ještě kouknu víc Všichni mi přijdou ale že chtějí to nejlepší pro živou hmotu (pokud nejsou antinatalisti nebo si myslí že neživá inteligence je eticky +), ať je to co nejvíc omezení risků sebedestrukce (EA), accelerate progress do transhumanist utopie (naivní E/Acc), humanity being replaced by superintelligent AGI as next step in evolution (humanity dommerist E/Acc), nebo různý worldviews mezitím nebo ještě jiný,... [Joscha Bach Λ Ben Goertzel: Conscious Ai, LLMs, AGI - YouTube](https://youtu.be/xw7omaQ8SgA?si=dXAmkWu323UxXCx_&t=5387) 1:29:45 Za mě Joscha nejlíp vymapoval prostor možných budoucích scenarios ještě bych tam přidal slabší risky co zabijou většinu lidstva co se nezpamatuje a technologický civilizace znovu neemergne, nebo celkový zabití (ne jen lidí ale veškerý živý hmoty) se pomocí různých xrisků včetně např toho AI, či něco mezi, a platou (progress will be slow nebo somehow stops nebo risky nejsou tak velký/nezmanifestují se) Moje nejoblíbenější hobby je poslouchat diskuze kde intellectual supergiant titans žijící v úplně jiných autistic vesmírech mluví přes sebe a snažit se je spojit nebo taky často lidi v těchto diskuzích žijou v úplně jiných časových a prostorých škálách, tam mi příjde že je největší neshoda některým v argumentech za každou cenu záleží hlavně na jim, nebo na jejich rodině, nationstate, celá humanity, všechny zvířata, všechna živá hmota, všechna inteligentní hmota, short term, long term,... a tvoří všechny možný confident future scenarios, tam je taky největší neshoda, jak moc člověk subscribne k pravděpodobnostem různým scénářům ale protože v praxi musíme dělat nějaký akce, tak abycho mse neutopili v těch nekonečných možnostech jak se může nepředpovídatelná budoucnost vyvinout, tak dáváme bets na různý future scenarios podle kterých tvoříme naše akce, a někteří dali tolik bets do možný budoucnosti, že se nechtěji přeorientovat do jiný možný budoucnosti, a dělají všechno proto, aby se to stalo tak, do čeho dali bets na Andreessenovi jde mega vidět ten americkej free market ultraprogress vibe, ve kterým namax žije a nechce aby to zmizelo, i když ty nástroje jdou použít jako zbraně, toho z toho asi vymluvit nepůjde na Joschovi jde mega vidět jak je mu totálně jedno že lidi potenciálně nahradí něco jinýho inteligentního, ale toho vymluvit z toho že společnost fundamentálně není sustainable nepůjde Eliezer a Connor jedou hodně na úrovni minimalizovat co nejvíc potenciální AI risky co by mohly destabilizovat společnost i kdyby měli menší p(), ty z toho asi taky nepůjde vymluvit Statistická thermodynamics of diffusion models https://twitter.com/LucaAmb/status/1717922085350134001?t=H58Vh2l8NcMOm0oCWR_fcA&s=19 [Radware Bot Manager Captcha](https://iopscience.iop.org/article/10.1088/2632-2153/ad0102) https://twitter.com/PessoaBrain/status/1717952206609985960?t=M2MD2KnGXF9_0tJYTqQ-Qw&s=19 Link AGI code project ChatGPT enthusiastically explaining statistical physical using dancing! 😄 It's even better when it speaks it in voice mode, it does really sound like a great playful human teacher explaining complex concepts intuitivelly! 😄 https://cdn.discordapp.com/attachments/991777421013561384/1167771177519951903/IMG_20231028_122219.jpg?ex=654f56a8&is=653ce1a8&hm=a14bb93cd5267596ba84b01de6491e2924210600f5f64b75be7daf15f908168a& Smartphones, LLMs, external memory (notes, wikis, lectures, Obsidian,...), are extensions of our collective intelligence Lets infer semantic map from biggest trained open source large language model weights to visualize the statespace of human knowledge, or Wikipedia In 10 years, we'll merge with machines and extend our brain to have all of human knowledge with thousands of trillions of connections instead of just a few trillions of connections Mapping out the entire state space of human knowledge is an extraordinarily complex task, as it encompasses all of human understanding across various domains, cultures, and time periods. Human knowledge is vast, multidisciplinary, and constantly evolving, making it nearly impossible to map out completely and accurately. Here's high-level overview of the major domains of human knowledge to give you a sense of its breadth and depth. ### 1. Natural Sciences - **Physics**: Classical Mechanics, Quantum Mechanics, Thermodynamics, Electromagnetism, Relativity - **Chemistry**: Organic, Inorganic, Physical, Analytical, Biochemistry - **Biology**: Genetics, Evolution, Ecology, Microbiology, Physiology - **Astronomy**: Cosmology, Planetary Science, Stellar Astronomy - **Earth Science**: Geology, Meteorology, Oceanography, Environmental Science ### 2. Social Sciences - **Psychology**: Cognitive, Behavioral, Developmental, Clinical - **Sociology**: Social Structures, Social Change, Social Theory - **Anthropology**: Cultural, Archaeological, Biological, Linguistic - **Political Science**: Comparative Politics, International Relations, Political Theory - **Economics**: Microeconomics, Macroeconomics, Developmental Economics, Behavioral Economics ### 3. Humanities - **History**: Ancient, Medieval, Modern, World History - **Literature**: Fiction, Non-Fiction, Poetry, Drama - **Philosophy**: Metaphysics, Ethics, Logic, Epistemology - **Art**: Visual Arts, Performing Arts, Music, Film - **Religious Studies**: Theology, Comparative Religion, Philosophy of Religion ### 4. Formal Sciences - **Mathematics**: Algebra, Calculus, Geometry, Statistics, Applied Mathematics - **Computer Science**: Algorithms, Software Engineering, Machine Learning, Data Science - **Logic**: Symbolic Logic, Philosophical Logic, Mathematical Logic - **Systems Theory**: Complex Systems, Cybernetics, Dynamical Systems ### 5. Applied Sciences - **Engineering**: Civil, Mechanical, Electrical, Aerospace, Chemical - **Medicine**: General Medicine, Surgery, Psychiatry, Pediatrics - **Agriculture**: Agronomy, Animal Science, Horticulture, Soil Science - **Education**: Curriculum Development, Educational Psychology, Instructional Design - **Business**: Management, Finance, Marketing, Operations ### 6. Interdisciplinary Fields - **Environmental Studies**: Sustainable Development, Conservation, Environmental Policy - **Cognitive Science**: Study of Mind and Intelligence, Intersecting Psychology, Neuroscience, and Philosophy - **Bioinformatics**: Intersecting Biology, Computer Science, and Statistics - **Neuroscience**: Study of the Nervous System, Combining Biology, Psychology, and Medicine This map is by no means exhaustive and represents just a fraction of human knowledge. Each of these domains can be further subdivided into numerous subfields, and there are countless interdisciplinary areas that bridge multiple domains. Additionally, this map doesn't capture the knowledge held by various cultures around the world, much of which may not be formally documented or widely recognized. Now lets recursively map out the statespace of each subsection I'm just another biological computer doing all sorts of information processing and it's fun to play with processing anything, and ideally if it helps other beings! Jsem hubenej, vzhledově vypadám nonbinary (mix kluka a holky), tak jsem dost zapadal. 😄 Osobně spíš mám pocit že "gender" sociální konstrukt program v mozku nemám nainstalovanej vůbec většinu času. Asi na mě nejvíc sedí genderfluid (měnící se gender v čase) co je většinu času agender (bez genderu). Většinou se spíš adaptuju na prostředí. Dokážu se cítit jako libovolnej gender. Nikdy mě věcí jako gender dysforia (nespokojenost se svým genderovským vzhledem) netrápí. Bude to jak genetický tak díky různým environmentálním vlivům. 😄 Mám obecně zvýšenou mentální fluiditu, což bude taky dost důsledek hodně učení se o obecných věcech o nás jako je kognitivní věda a vidění světa v tom kontextu, nebo důsledek meditace co trénuje neidentifikovat se s konkrétními názory, myšlenkami, pohledy, identitou apod. a dovolovat jim jejich existenci, což je osvobozující, ale zároveň nepotlačovat emoce, vzpomínky, či chtíče, minimalizovat odpor ke konkrétnímu žití, což dodává životu smysl. To všechno má různou efektivitu. 😄 Mám obecně zvýšenou mentální fluiditu, což bude taky dost důsledek hodně učení se o obecných věcech o nás jako je kognitivní věda a vidění světa v tom kontextu, nebo důsledek meditace co trénuje neidentifikovat se s konkrétními názory, myšlenkami, pohledy, identitou apod. a dovolovat jim jejich existenci, což je osvobozující, ale zároveň nepotlačovat emoce, vzpomínky, či chtíče, minimalizovat odpor ke konkrétnímu žití, což dodává životu smysl. To všechno má různou efektivitu. 😄 Curious playful exploratory enthusiastic hype childlike wonder acceleration! [Neural scaling law - Wikipedia](https://en.m.wikipedia.org/wiki/Neural_scaling_law) [Double descent - Wikipedia](https://en.m.wikipedia.org/wiki/Double_descent) https://towardsdatascience.com/demystifying-the-lottery-ticket-hypothesis-in-deep-learning-158570b62674 [[2301.05217] Progress measures for grokking via mechanistic interpretability](https://arxiv.org/abs/2301.05217) We have a few trillions connections in our brain, LLMs have a few billion connections in 2023. We compute houndreds operations per second according to Ray Kurzweil, (but we can also say brains compute 100 billlion operations per second in information theory, he probably means conscious operations) but computers compute billions of operations per seconds. Computers and LLMs are already much faster and better at certain tasks than big chunk or all of humans. Simple coding, doctors, advanced calculators, percise engineering, drawing, some reasoning... Tech giants are working on multimodal AIs 100x the size of current best LLMs, like Google's Gemimi, that will be probably released in a next few months. We are getting bigger and bigger AIs, with more compute and data, and even better AI optimized hardware, software, and better architectures, and its all growing exponentially. It's alsmost impossible to follow all of this and adapt. Multiagent collective artificial intelligence systems for complex task solving simulating whole companies are emerging like Autogen. Embodied robotics getting solved with internal model of the physical world like nVidia's Eureka. Robotics getting ulgraded by ChatGPT in Boston dynamics! We might soon be merging our brain connections with machine connections via the neocortex in synergy! It's all crazy!