“ How to exactly articulate better quality standards for fundamental theories of physics? Quantum gravity theories try to solve inconsistency between quantum mechanics and general relativity. I feel like this cuts right at the core of how to make AI generate actually creative useful novel ideas like our best scientists in the past! What is the equation of useful scientific novelty? I want digital Einstein, Neumann, Feynman, Godel, Hilbert, Ramajuan, Gauss, Perelman, Grotenderick, Turing, Tao, Witten, Pythagoras, Newton! Or analog, as it really doesn't matter which substrate, as long as it works! I want trillions of them in one datacenter collectively solving the equation of the universe, the equation of intelligence, all the math, billions of times faster than all of civilization combined so far! But what edits to the current AI architectures need to be done? What is the secret sauce of the brain? How to go beyond the secret sauce of the brain? What is the secret sauce of collective intelligence, what are all the environmental and genetic factors, that makes a biological or non-biological system invent something groundbreaking in science? Designing AGI system that can very deeply grok etc. classical mechanics, general relativity, quantum mechanics, standard model, loop quantum gravity, string theory, etc. and derive new physics that actually has a higher probability of being successful empirically, using something similar to whatever happened in Newton's, Einstein's and Schrodinger's brain when they came up with their models. AI system fully specialized in modelling nature across scales in different physics theories, using quantum/thermodynamic/deterministic theories on different scales, with some natural language interface on top of it. Maybe the answer is somewhere in NeuroAI and neurosymbolic AI or the free energy principle! https://x.com/skdh/status/1897153912315969773 [Catalyzing next-generation Artificial Intelligence through NeuroAI | Nature Communications](https://www.nature.com/articles/s41467-023-37180-x) " What is true in philosophy of mind? Physicalism? Ideliam? Panpsychism? Illusionism? Dualism? Monism? Mysterianism? I feel like big part of me personally became agnostic, as the space of all possible positions in philosophy of mind seems so large and kind of arbitrary what camp you pick. For a scientist, physicalism is useful. If you do meditation or psychedelics a lot, you'll gravitate towards idealism or mysterianism. There was a paper showing this correlation as well, but it may not be causation, but I suspect it is. And most normies in our culture usually think in cartesian dualism I feel like, maybe that's the current evolutionary baseline. From a physicalist perspective, I feel like all these positions in philosophy of mind have their own neural correlates that say how the brain constructs the model of self and other and of qualia. Dualists model a bigger boundary between inner experiential world and outer nonexperiental world, while dualists don't have a boundary and everything is one thing, either experience, no experience, or something third. Panpsychists label as experiental world everything. Illusionists label nothing. Open individualists have their model of inner self exploded to their model of the whole universe. Personally, I've experienced so many of these, that right now I'm like: Ok, all of them can feel true if you do the intellectualization or other activities that induce these states of mind, so reality maybe is incomprehensible instead and this is from scientific physicalist perspective all just useful programs for my ape brain. But it's useful to assume that this ape brain creates one conscious world simulation, like for example how Joscha assumes it, as that allows you to do all sorts of engineering of mental representations and of qualia, by internal engineering by conscious actions, or by external engineering by neurotechnology. So it feels like right now in my experiental world and in my intellectual world of mental representations, specifically in the philosophy of mind, there's currently a superposition of physicalism and mysterianism. Science and engineering me prefers physicalism with laws of physics with all it's emergent laws in our brain being our qualia, philosophical me prefers mysterianism swimming in the combinatorial explosion of possible philosophy mind positions, and experiential me experiences both at the same time. Co třeba skoro všechny vědy kde se nějak pracuje s daty kde teď právě probíhá AI revoluce? To se prostě k lidem tak nedostává > Ale já chci to AI ze sci-fi co pomáhá fyzice a biologii Ale to už se všechno do jistý míry děje! [https://youtu.be/XRL56YCfKtA?si=_SzJfJH60Zg1IOXC](https://youtu.be/XRL56YCfKtA?si=_SzJfJH60Zg1IOXC) [How can AI help physicists search for new particles? | CERN](https://home.cern/news/news/physics/how-can-ai-help-physicists-search-new-particles) [AI can now model and design the genetic code for all domains of life with Evo 2 | Arc Institute](https://arcinstitute.org/news/blog/evo2) https://fxtwitter.com/rohanpaul_ai/status/1880184389218496770 [FermiNet: Quantum physics and chemistry from first principles - Google DeepMind](https://deepmind.google/discover/blog/ferminet-quantum-physics-and-chemistry-from-first-principles/) [Google DeepMind and Isomorphic Labs introduce AlphaFold 3 AI model](https://blog.google/technology/ai/google-deepmind-isomorphic-alphafold-3-ai-model/) [https://www.youtube.com/watch?v=P_fHJIYENdI](https://www.youtube.com/watch?v=P_fHJIYENdI) https://x.com/tsarnick/status/1882238097594916872?t=6NrtaVrBFZRtiOYpJU8XoQ&s=19 ADfullHDing the space of all knowledge What is the fundamental structure of reality? generalized generalizing The human morphology is just a tiny point in the space of all possible configurations of physical systems AI x physics is endless rabbit hole: You can study AI using methods from physics, you can study physics using AI models, you can try to make AI systems model physics as accurately as possible through physics biases, you can design better AI architectures using physics, many AI architectures are applied physics, etc. Welcome to another episode of omnidisciplinary exploration of intelligence " Are we quantum computers? ❤️ If the brain eventually turns out to use the principles of quantum mechanics to compute and process information more efficiently, which is an as of now still unconfirmed hypothesis with insufficiently strong empirical evidence, then we are all quantum computers. ❤️ Or there are some confirmed quantum processes in biology in general. ❤️ But at least all of us, and all other physical systems, are at the fundamental level the fundamental particles of the universe that make up matter and forces, which operate on the principles of quantum mechanics, and other physics that we know, and that we don't know yet, so we're still quantum anyway. ❤️ Or any physical system can be seen through quantum information theory. ❤️ " universal darwninism across all scales of organization [Universal Darwinism - Wikipedia](https://en.wikipedia.org/wiki/Universal_Darwinism) Different kind of proofs/truths: philosophers: it feels good mathematicians: its derived from these axioms under these rewriting rules scientists: all (or most) (or in just some context) empirical data supports this model, so its true (or partially true) (or true just in this context) until its not engineers: boom! it fookin works, i got Doom to run on this alien computer that uses biology-inspired bioelectric selforganizing system running adaptive hiearchical bayesian belief mechanics! i want all the transhumanist upgrades, specifically the ones that increase intelligence i believe relatively soon there will get exponentially cheap and be everywhere [Significantly Enhancing Adult Intelligence With Gene Editing May Be Possible — LessWrong](https://www.lesswrong.com/posts/JEhW3HDMKzekDShva/significantly-enhancing-adult-intelligence-with-gene-editing) for example the price of sequencing genome is falling extremely, which leads to democratization it was 100 000 000 dollars 25 years ago, now its less than 1 000 dollars more generally playing with human genes isnt science fiction anymore, its already a reality, for example we have FDA approved gene theraphy for sickle cell diesease https://www.genengnews.com/topics/genome-editing/fda-approves-the-first-crispr-therapy-for-sickle-cell-disease/ and since gene related technologies are falling exponentially, i believe it will ge tsuper cheap and abundant soon similar to how first other kinds of cures were expensive at first one particular application of intelligence enhancing technology I see is it helping us solve the mysteries of the universe in physics having tons of Einsteins or SuperEinstens focused on solving the equation of the universe is something i dream about often both augmented humans and AIs i think we would get so much more progress so much more faster there thanks to that "well... let's look at the world realistically. if we, as a species, survive long enough, we will do genetical alterations, first out of necessity (imagine how desired can some mutations be in the climate-devastated world, like tolerance for heat, better water retention systems...), then for fun and profit. that's how people are. and it's not necessarily a bad thing; many good stuff for the meta-species can be discovered this way." that's also true, and to survive in the universe more generally thanks to advanced technology and these potential upgrades to our source code engineered by evolution so far, since evolutionary natural selection currently works less on us, and the sun in our solar system might likely swallow our Earth when it will deplete its hydrogen fuel and turn into a red giant 😄 (for example deep learning is already helping in CERN) hezkej článek z CERNu [Learning by machines, for machines: Artificial Intelligence in the world's largest particle detector | ATLAS Experiment at CERN](https://atlas.cern/Updates/Feature/Machine-Learning) s trénovaním neuronek na datasetu kolem detekce Higgs bosonu kterej by měl být z CERNu jsem si hrál 😄 and i think we could do so much more there with augmented humans and better AIs Já se chci někdy vrátit do krásnýho perfektního abstraktního světa teorie kategorií to je prostě svět kde člověk dělá matematiku nad matematikou samotnou, to je tak strašně nádherný svět empirická aplikovaná matematika je zase strašně messy bordel plná hacků, ale ta je zase fascinující jinak, protože fyzikální svět se těžce popisuje jinak 😄 Racionalismus sám o sobě nejobecněji je o maximalizaci empirický pravdy přes mentální modely a vědeckou metodu Efektivní altruismus sám o sobě nejobecněji je použit vědeckou metodu na co nejefektivnější dělání dobra A pak na to existuje bilion různých konkrétních implementaci ovlivněný spousty různýma biasama Plus je do toho ještě hodně často propojený transhumanismus a futurologie (forecasting možných budoucností) A velký častý focus na AI tady listuje ty dobrý věci např [In Continued Defense Of Effective Altruism](https://www.astralcodexten.com/p/in-continued-defense-of-effective) co chápu tak Harry Potter and the methods of rationality se snaží vysvětlit různý koncepty z vědecký metody, kognitivní vědy a filozofii v jazyce fanficu, což mi příjde dobrý jako idea, ale nevím jak přesně to je duělaný mám rád hodně Eliezerových (autorových) myšlenek, ale dost nesouhlasím s jeho fatalistickým extrémním AI doomerismem třeba kousky multiagent models of mind jsou cool [Multiagent Models of Mind - LessWrong](https://www.lesswrong.com/s/ZbmRyDN8TCpBTZSip) tenhle žánr, dát vědu do příběhů, mě dost fascinuje a inspiruje, třeba včera jsem přemýšlel jak že vezmu moje oblíbený matiky a každou část zasadím do magickýho příběhu zkoumatele potkávající magický starý kouzelníky plný moudrosti a různý mimozemský stvoření mluvící mimozemštinou co se musí luštit 😄 ⁠ nebo chci udělat byl jednou jeden život ale pro fyziku fundamentálních částic vesmíru 😄 ⁠ plus ještě od nich často čtu různý technický analýzy AI technologií a jejich reverse engineering [An Extremely Opinionated Annotated List of My Favourite Mechanistic Interpretability Papers v2 — LessWrong](https://www.lesswrong.com/posts/NfFST5Mio7BCAQHPA/an-extremely-opinionated-annotated-list-of-my-favourite) racionalitu bych taky mohl zkusit 😄 třeba one by one vylistovat všechny kognitivní biasy v příběhu 😄 a matematickou teorii pravděpodobnosti a statistiky pro vědeckou metodu 😄 nebýt týhle mý divergentní ungovernable feral knowledge seeking impulze a potřeby impulzivně tvořit abych se cítil stimulovanej, tak bych si např neudělal vlastní wikipedii co se snaží přes odkazy obsáhnout věškerý technický vědění lidstva s focusem na AI, inteligenci, fyziku, matiku, vědomí, futurologii, filozifii, transhumanismu, racionalismu, altruismu, atd. 😄 [Exocortex - Burny](https://burnyverse.com/Exocortex) a doslova se teď moji neurodivergenci snažím matematicky formalizovat a implementovat do AI, lol... 😄 i love it ještě se osobně snažím do efektivního altruismu víc prosazovat AI pro zdravotnictví, protože věřím, že kdyby se mnohem víc investovalo do technologiíí typu AlphaFold co zjišťuje jak fungují proteiny, různý metody na hledání a tvoření léků, různý diagnostický metody, další metody jak zjistit jak funguje biologie a mozek, atd., tak množství nemocí co nás ohrožuje by padalo dolů mnohem rychleji než teď pak je tam ještě velká skupina co řeší extreme global poverty s nějakýma výslekama [https://www.youtube.com/watch?v=2DUlYQTrsOs](https://www.youtube.com/watch?v=2DUlYQTrsOs) myšlení na budounost je dle mě naprosto super, miluju futurologii, ale ideální je to propojit se současností, a zároveň nedojít k výsledku, že např cesta k lepší budoucnosti je způsobit šílený utrpení teď nějakým způsobem lidi co dělají na AlphaFold co strašně pomáhá zjistit jak fungují proteiny např často daydreamují o budoucnosti bez nemocí díky advanced budoucí AI na tohle využití pro všechny [https://www.youtube.com/watch?v=P_fHJIYENdI](https://www.youtube.com/watch?v=P_fHJIYENdI) teď je tam např ještě velká skupina co chce v podstatě udělat surveilence dystopii grafických karet a AI systémů obecně, se kterou silně nesouhlasím, což mě dost odpuzuje singleton superintelligent ASI: dávám tomu malou nenulovou pravděpodobnost, můj nejpravděpodobnější scénář je hodně velká diverzita hodne diverzních typů AI systémů co jsou různě autonomní a různě využívaný lidmi, a různý typy cyborgů se rozhodnou tím různě rozšířit svoje schopnosti ještě víc propojeně, a do toho se teď ještě vyvíjí syntetická biologie, což se všechno mixuje dohromady, apod. připadá mi že výzkum a praxe tímhle směrem dost jde, diverzita různých typů AI systémů pro různý účely je obrovská už teď můj největší dream jsou teď různý AIs specialized for inventing and doing all sorts of science, trying to understand the universe, beyond limited human intuition, since our minds are currently constrained by our evolutionary programming in a lot of contexts [https://www.youtube.com/watch?v=T_2ZoMNzqHQ](https://www.youtube.com/watch?v=T_2ZoMNzqHQ) Adding more diverse intelligences into the mix acts more as a multiplier of our capabilities and as an upgrade in places where our brain's architecture made by evolution is too limited and constrained. That seems to have been the case so far, each type of intelligence excelling in different ways, that are even stronger together. And if it will lead to for example breakthroughs in physics or curing diseases faster, then I think that's amazing. Oh damn mahajana buddhism sutra 😄 Mám rád nějaký ideje z Heart sutry z druhý ruky od mýho oblíbenýho učitele meditace 😄 Form is emptiness, emptiness is form. obecně mi příjde, že dost budhistický filozofie schválně rozbíjí klasickou logiku, aby ten jazyk právě transcendnula 😄 "neither true nor not true" tohle asi taky nechápu úplně přesně tím že to mám přeložený přes druhýho člověka, ale příjde mi, že v moderním vědeckým jazyce k tomuto existuje paralela, kde mozek tvoří extrémně zjednodušenou simulaci externího světa, co je v jistým smyslu úplně mimo fyzikální realitu (emptiness) a to stačí na přežití, ale zároveň je to hodně napojený na realitu v tom, že to je ta jediná realita ve který žijeme (form) a nic jinýho jako realitu neznáme [https://www.youtube.com/watch?v=wo_e0EvEZn8](https://www.youtube.com/watch?v=wo_e0EvEZn8) ale je často strasně sranda si s tím herním enginem co mozek generuje různě hrát 😄 např různě rozbíjet nebo modifikovat simulovaný čistý svět, simulovaný objekty, simulovaný příbehy, simulovanou identitu, simulovaný abstraktní koncepty, simulovaný časoprostor, apod. 😄 ale je na houby když se mozek přesvědčí v depresivní přesvědčení o sobě, ale tam je těch faktorů víc, jak hardwarový tak softwarový je zajímavý jak mozek většinou v podstatě předpovídá nejpravěpodobnější a nejužitečnější realitu pro přežití organismu (co může a nemusí být až moc odpojená od fyzický reality), což se ale u různých mental problémů právě rozbije do buďto ani trochu užitečný reality, do reality co vůbec přežití nepomáhá a nepředpovídá důležitý věci v naší koncenzusový realitě dostatečně... 😄 From modern cognitive science language I think you can approximate parts of Buddhist "emptiness is form, form is emptiness" as the brain creating an extremely simplified simulation of the external physical world, which is in a sense completely out of touch with actual physical reality (emptiness), but it's enough to survive in our evolutionary environment, but at the same time it is very much connected to reality in that it is the only reality we live in (form) and we know nothing else as reality, and it still predicts the reality well enough for our organism to survive. It's often a lot of fun to play with the generative game engine, that the brain generates, in different ways, e.g. different ways to break or modify the simulated raw sensory data world, simulated objects, simulated stories, simulated identity, simulated abstract concepts, simulated spacetime, etc. :D but it sucks when the brain gets into reinforcing maladaptive beliefs itself, but there are more factors to that, both hardware and software. It's interesting how the brain basically predicts the most probable and useful reality for the organism's survival (which may or may not be too disconnected from physical reality), but which for various mental problems just breaks down into either not even a slightly useful reality, a reality that doesn't help survival at all and doesn't predict the important things in our consensus reality enough. :D But you have so much freedom to engineer our inner world simulation when you are selfaware and agentic enough! Will AI in the future come up with theories of fundamental physics that predict empirical data better than our theories but that are incomprehensible to human intuition? Is brain quantum? If so, is it necessary for its intelligence? And in terms of quantum computation being used for better information processing in the brain, afaik, it's an as of now still unconfirmed hypothesis with insufficiently strong empirical evidence, but there are some quantum processes in biology in general unrelated to cognition. But we're still fundamentally to some approximation Standard model stuff which operates under quantum field theory. Still, afaik, we are struggling to find quantum phenomena like superposition and entanglement in the biological neural computations for cognition, because it's a really noisy and warm environment, where quantum computations have trouble surviving, so classical models of information processing are relatively winning more so far in terms of predicting cognition, but we still don't know so much about what equations is the brain using. My ideal scifi would be about benevolent superintelligence that cures all diseases, makes all beings happy, figures out how biology, fundamental physics, consciousness, intelligence, etc. works by countless scientific breakthroughs, understands all math, understands everything in philosophy, creates post-scarcity abundance for all, creates infinitely fascinating complex art, and in the process grows infinitely more and more in intelligence and creativity, maximizes morphological freedom, and does no harm Benevolent superintelligence explosion when are we building galaxy supercluster sized particle collider to probe quantum spacetime foam Advanced AI will be needed to overcome human limitations in the search for theory of everything I see theory of everything as a path, as gradual refining of our predictive models of the world, that will never get to 100%, and AI will help us get closer to it beyond our human limitations All models are wrong but some approximate the practically infinitely nuanced complexity of reality for empirical predictions better than others. I want machine scientists developing theories and experiments about the universe that transcend human limitations