The biggest questions in science are: What is the equation of intelligence? What is the equation of the universe? classical generally relativistic statistical quantum field string loop theory Biotechnologie na nemoci mě mega fascinují! 😄 Z velký části proto hodně zbožňuju např AI for healthcare, jako protein folding, predikování fMRI dat apod., tím jak to zároveň rozlušťuje jak fungujou biologický organismy a jak je engineerovat když se v nich nějaký součástky nějak pokazí a jsou z toho nemoci a jak to opravit, nebo jak upravovat jejich funkce, přes biochemical drug engineering, genetic engineering, bioelectricity engineering, chirurgie, syntetický součástky, apod.. 😄 A myslím, že abychom pořádně pochopili co nejvíc v biologii, tak bude AI hodně potřeba (už tam hodně pomáhá teď), tím jak lidský mozky jsou dost limitovaný, jako tady popisuje Dario (týpek měl phd v biologii a teď v podstatě vede v konkrétním typu AI systémů). https://fxtwitter.com/tsarnick/status/1856076482772381970 Jedna z těch rovnic popisuje energii všech známých částic ve vesmíru. Vesmír se skládá z částic hmoty (šest kvarků a šest leptonů), částic přenášejících sílu (např fotony pro elektromagnetismus), a to vše řízeno čtyřmi základními silami: gravitací, elektromagnetismem, silnou jadernou silou a slabou jadernou silou. Gravitace ale dle obecný relativity zároveň způsobuje zakřivení časoprostoru, což nějak nemůžeme dát matematicky dohromady s ostatníma silama a fundamentálníma částicema v kvantovým světě. Enegie se dělí na kinetickou a potenciální energii. Mimo částice kinetická energie může být např letadlo co má nějakou rychlost, a potenciální může být např gravitace. Enegie se vždycky transformuje z jedný na druhou a nevytváří se z ničeho, jinak bychom měli perptetum mobile. :D Když dokážeš napsat veškerou potenciální a kinetickou energii systému, tak z toho můžeš odvodit rovnice toho jak se ten systém v čase mění, např to, že zrychlování objektu koresponduje se sílou co na ten objekt působí (Newtonovo druhý pravidlo), kde se vesmír snaží být co nejlínější a dělat co nejmíň akce, např že světlo z bodu a do bodu b jde co nejrychlejší nejkratší cestou. :D Computational irreducibility is all you need [https://youtu.be/xjH2B_sE_RQ?si=r9akdZ_xVrpHN_Oq](https://youtu.be/xjH2B_sE_RQ?si=r9akdZ_xVrpHN_Oq) We're just atoms that formed after hydrogen emerged a few hundred thousand years after the Big Bang, that were cooked into heavier elements inside stars, and that have been then vibing for nearly 14 billion years My understanding is, for example: Linear algebra is a great language for making concrete models that predict the physical world. You can cast linear algebra in set theoretical language or category theoretical language. You will get same kinds of predictions using different foundations, but you have to do slightly different thinking when doing proofs (category theory proves things by commuting diagrams). We're just atoms that formed after hydrogen emerged a few hundred thousand years after the Big Bang, that were cooked into heavier elements inside stars, and that have been then vibing for nearly 14 billion years Reduced beliefs understand psychedelics model ftw. Or sometimes it strengthens the valleys too, or alters them. Also I was thinking that since identity is a narrative made of emotionally salient experiences, that when people say that "psychedelics showed them the truth", that they essentially integrated the silent experience deeply into their identity. https://x.com/burny_tech/status/1857613174398202287?t=0H8aIrjN1JnAL-P0MnhRIw&s=19 Half life 3 will be controlled by thoughts [Gabe Newell, the Man Behind Steam, Is Working on a Brain-Computer Interface](https://futurism.com/neoscope/gaben-steam-bci-stealth) Cognitive behavioral therapy is good Konkrétně cognitive restructuring technika v ní Very well mind má super summary [Cognitive Behavioral Therapy (CBT): Tasks & Beliefs](https://www.verywellmind.com/what-is-cognitive-behavior-therapy-2795747) "faulty thoughts are identified, challenged, and replaced with more objective, realistic thoughts" Takovej debugging a opravování softwaru mozku 😄 Internal family systems therapy se mi líbí víc [IFS Therapy: Techniques, Benefits, Considerations](https://www.verywellmind.com/what-is-ifs-therapy-internal-family-systems-therapy-5195336) Ale když jde člověk k psychoterapeutovi tak nejčastěji jsou trénovaní na cognitive behavioral therapy Pak je ještě relativně častá psychodynamická terapie <https://www.verywellmind.com/psychodynamic-therapy-definition-uses-effectiveness-5094933> A pak humanistic therapy <https://www.verywellhealth.com/humanistic-therapy-5216335> V český psychedelický společnosti jsem potkal therapeuta co praktikuje gestalt terapii <https://www.verywellmind.com/what-is-gestalt-therapy-4584583> Nejvíc dělám deconstructing meditation https://deconstructingyourself.com/dissolve-all-thought-in-space.html a na druhým místě love and kindness meditation [How to Practice Loving Kindness Meditation](https://www.verywellmind.com/how-to-practice-loving-kindness-meditation-3144786) <https://deconstructingyourself.com/mother-buddha-love.html> Ale dělám to mnohem míň jak kdysi, a dost na svým mental health vidím jak moc mi to chybí, tak se ten habit, hodinu meditace denně, snažím teď restartovat Je zajímavý jak každýmu funguje trochu něco jinýho, podobně jak každýmu funguje trochu jiný antidepresivní látky What is God's system prompt? Make sure action is stationary All possible perspectives fall into the ontological realness category depending on their degree of predictivity = predictive realism If you are learning physics and machine learning in parallel then you'll more and more feel like AI is a subfield of physics. The math used in both of these fields is too similar! dost lidí bere různý supplementy i když to má celkem weak evidence jako formu gamblingu co je alespon trochu evidence based fakt miluju 100% confident claims bez zdrojů ale obecně miluju 100% confident claims i se zdrojama, protoze věda neni tak jednoduchá a ja nerikam ze to 100% funguje, i kdyby to bylo s nejakym miniaturnim statistickym benefitem, jestlis to tak interpretoval ale to ze dle ruznych dat by to mohlo fungovat (kde nekteri biohackeri u nekterych latek sazi i na extrapolaci ze studies s unhealthy animals, když nejsou studie, ktera vubec nemusí platit) (a v substances vědě i biliarda studii muze byt broken, viz recent SSRIs metametaanalyza tvorici chaos) V klasický ekonomii vidím spíš věci jako klasickou teorii her, a econophysics vidím jako mnohem víc physics based, jako např aplikovanou statistickou fyziku Kde econophysics mi příjde jako mnohem víc niche podobor Hmm, jsi si jistá že mluvíme o tom samým? Na různých místech vidím, že o klasický ekonomii a ekonofyzice mluví jako o dvou odlišných oborech, co mají určitý překryv, např v: [Physics, economics, and econophysics: what's the nexus? | The Financial Express](https://thefinancialexpress.com.bd/views/views/physics-economics-and-econophysics-whats-the-nexus) "The term econophysics is a composite concept – a hybrid discipline stitched by combining physics and economics – first introduced by a theoretical physicist Eugene Stanley in 1995 at a conference on Dynamics of Complex Systems held in Kolkata, India. Physicists’ interests in applying physics models to economic problems grew out of the notion that several traditional economic theories such as rational expectations, the efficient market hypothesis, and the view of equilibrium are too axiomatic and formal to analyse complex systems such as financial markets. They argue that prevailing theories “dealt with homogeneous agents and equilibrium, while many of the more interesting phenomena in financial markets fundamentally depended on heterogeneous agents and far-from-equilibrium situations.” Econophysics hypothesises that the economy should be treated as though behaving like a collection of electrons in atoms or a cluster of water molecules that interact with each other. That attracted statistical physicists to study the statistical properties of complex economic systems consisting of many economic units such as firms, individuals, families, households etc. The physics models used in econophysics include Kinetic exchange models of markets, percolation models, chaotic models developed to study cardiac arrest, earthquakes to name a few. For example, the multi-agent Kinetic exchange models developed from the statistical physics of energy distribution are used to explain the robust and universal features of income/wealth distributions – one that economists have been struggling to demystify for over a century. The consideration of economic agents as particles allows econophysicists to treat markets as complex macroscopic systems with internal structure that is microscopic in character. For a layman, this simply means that each of the four markets (that is market for goods, money, labour, and bond) is macroscopic and agents in each market are microscopic having an internal character. In this internal microscopic structure, economic particles are assumed to interact among themselves to generate systemic macroscopic properties. Econophysics research papers have been published mostly in Journals dedicated to physics and statistical mechanics. These papers are yet to intrude into foremost economics journals. Additionally, mainstream economists are yet to be enthused with econophysics and only a handful of economists such as Mauro Gallegati (Italian), Steve Keen (Australian), Paul Ormerod (British), and Alan Kirman (French) have shown both enthusiasm and disappointment with econophysics. However, Nobel laureate and founder of experimental economics Vernon Smith (won 2002 economics Nobel Prize) has used “econophysics to model sociability via implementation of ideas in Humanomics”. So far, some successful impacts of econophysics are realised in applied areas of quantitative finance primarily because the scope and goals are quite different from those of economic theories. " kde koukám že jeden transition mezi ekonomikou a ekonofyzikou je homogeneous agents and equilibrium -> heterogeneous agents and far-from-equilibrium jak to slovo definujou asi může být confusing :3b1b_pi_shrug: [Econophysics - Wikipedia](https://en.wikipedia.org/wiki/Econophysics) Na tý wikině to taky tak odlišujou: "Econophysics was started in the mid-1990s by several physicists working in the subfield of statistical mechanics. Unsatisfied with the traditional explanations and approaches of economists – which usually prioritized simplified approaches for the sake of soluble theoretical models over agreement with empirical data – they applied tools and methods from physics, first to try to match financial data sets, and then to explain more general economic phenomena. It became apparent that traditional methods of analysis were insufficient – standard economic methods dealt with homogeneous agents and equilibrium, while many of the more interesting phenomena in financial markets fundamentally depended on heterogeneous agents and far-from-equilibrium situations." Ale nedivil bych se, kdyby za ty roky ty modely založený na heterogeneous agents and far-from-equilibrium situations už byly zakomponovaný do klasickýho curriculum Jako nějakej introductory předmět, kde si člověk pak může econophysics vybrat jako specializaci dál. Pokud to tak je, tak to může být podobný tomu, jak deep learning v praxi rozbil některý věci v klasický learning theory, a začal se mnohem víc vyučovat. 😄 Jako: [Double descent - Wikipedia](https://en.wikipedia.org/wiki/Double_descent) "perceived contradiction between the conventional wisdom that too many parameters in the model result in a significant overfitting error and the empirical observations in the 2010s that some modern machine learning models tend to perform better with larger models". >Zajimalo by me, jak by potom vypadala ekonomie, co neni econophysics Tady máš odpověď no 😄 Já mám obecněji tendenci většině aplikovaný matematice na fyzikální systémy říkat fyzika, takže chápu tvou intuici. Ale vypadá to, že pro ekonofyziky je ekonofyzika dost konkrétní podmnožina (heterogeneous agents and far-from-equilibrium situations). Is there a physical system that doesn't have entropy? <https://en.wikipedia.org/wiki/Entropy_(classical_thermodynamics)> (edit: in the sense of not being able to assign some entropy value to it) Ale je pravda afaik že black holes jsou unikátní fyzikální systémy v tom, že mají maximální entropii given their size, má to implikace pro kvantovou gravitaci 😄 (proportional to the surface area of its event horizon) (Bekenstein-Hawking entropy formula <http://www.scholarpedia.org/article/Bekenstein-Hawking_entropy> <https://en.wikipedia.org/wiki/Black_hole_thermodynamics> ) K entropii se vracím pořád, je všude. Např v machine learningu je teď asi nejpoužívanější metoda minimalizace cross entropie. 😄 <https://en.wikipedia.org/wiki/Cross-entropy> A život je možná emergentní důsledek nonequilibrium thermodynamiky [Nonequilibrium Statistical Thermodynamics of becoming (and remaining) Life-Like](<[https://youtu.be/10cVVHKCRWw>](https://youtu.be/10cVVHKCRWw>) Hmm! <:HyperThink:991915571593609276> Heat death of the universe is similar to a black hole, because in both cases, the entropy of the physical systems is maximum! <:FractalThink:991803016049066014> <https://en.wikipedia.org/wiki/Heat_death_of_the_universe> Úplně ekvivalent nejsou no, ale obojí jsou logaritmický funkce na kvantifikaci disorder/randomness/uncertainty s kombinatorickou náturou (counting microstates vs possible outcomes of a random variable) :smile: <https://en.wikipedia.org/wiki/Entropy> <https://en.wikipedia.org/wiki/Entropy_(classical_thermodynamics)> <https://en.wikipedia.org/wiki/Entropy_%28information_theory%29> [Entropy in thermodynamics and information theory - Wikipedia](https://en.wikipedia.org/wiki/Entropy_in_thermodynamics_and_information_theory) A Landauer's principle tyhle dvě formy entropie propojuje! (erasure of one bit of information in a physical system is associated with a minimum dissipation of heat into the environment) <https://en.wikipedia.org/wiki/Landauer%27s_principle> Hmm, pokud se "doesn't have entropy" interpretuje jako entropy = 0 místo entropy = undefined, tak technicky teoreticky perfektní čistý perfektně ordered krystal s teplotou na absolutní nule má teoreticky 0 entropie, ale to je nerealistická idealizace co neexistuje no 😄 Nikde není absolutní nula, nebo nic není úplně izolovaný systém, nebo quantum uncertainty to rozbíjí <https://www.reddit.com/r/Physics/comments/6zn04v/is_it_possible_for_a_system_to_reach_0_entropy/> <https://chem.libretexts.org/Bookshelves/Physical_and_Theoretical_Chemistry_Textbook_Maps/DeVoes_Thermodynamics_and_Chemistry/06:_The_Third_Law_and_Cryogenics/6.01:_The_Zero_of_Entropy> The space of all possible lifeforms is so vast https://x.com/burny_tech/status/1860585759834529899 [https://www.youtube.com/watch?v=6w5xr8BYV8M](https://www.youtube.com/watch?v=6w5xr8BYV8M) "do non-human animals make predictions and minimize prediction error too" According to some, even molecules and atoms and particles do it, but not really sure about the usefulness of that I'm gonna stay at max the cell level probably, or maybe viruses Hmm [Determinants of Virus Variation, Evolution, and Host Adaptation - PMC](https://pmc.ncbi.nlm.nih.gov/articles/PMC9501407/?utm_source=perplexity) Not sure if you could cast it in this framework [Bacteria opt for the best price-to-quality ratio to predict the future - AMOLF](https://amolf.nl/news/bacteria-opt-for-the-best-price-to-quality-ratio-to-predict-the-future?utm_source=perplexity) [New study shows how cells sense their microenvironment | Karolinska Institutet](https://news.ki.se/new-study-shows-how-cells-sense-their-microenvironment?utm_source=perplexity) Chris Fields likes to talk about cells predicting [https://youtu.be/qA6oMSXhljk?si=-BesIiRb_ZcDBqBv](https://youtu.be/qA6oMSXhljk?si=-BesIiRb_ZcDBqBv) Or Joscha Bach and Michael Levin [https://youtu.be/kgMFnfB5E_A?si=q8q13bWsjnM4I5j1](https://youtu.be/kgMFnfB5E_A?si=q8q13bWsjnM4I5j1) Josha: Every neuron, like every cell, is an individual reinforcement learning agent that tries to survive by cooperating with its environment. The main difference between neurons and other cells is this ability to communicate over long distances rapidly. Other cells typically only communicate with adjacent cells. A neuron is essentially a "telegraph cell" that can send messages over very long distances quickly within an organism. The neuron's ability to process information quickly over long distances allows for the development of a model of the world that can be updated at higher rates than the rest of the cellular system. To merge it with QRI: these long range connections help the holistic field computing <:FractalThink:991803016049066014> ochočili jsme fyziku přírody přes inženýrský krotitelský kouzla využívající a ohýbající zákony přírody k našim cílům We have tamed nature's physics through engineering wizardry using and bending the laws of nature to serve our goals reality's insane complexity is infinitely nuanced https://en.wikipedia.org/wiki/Vault_(organelle) vyskytuje se téměř u všech možných organismů, kromě populárních modelových organismů, a nikdo neví, k čemu ten prevít slouží i had that on my feed bunch of times recently https://fxtwitter.com/Rainmaker1973/status/1868342613088129462 Někdy je úplně mindboggling co biologie dokáže evolucí vyčarovat Myslím že se do mainstreamu eventuálně víc vrátí evoluční algoritmy, dneska se používají někdy, např na neural architecture search (automatický hledání konfigurace architektur) Hmm tady někdo zkombinoval gradient descent a evoluci [[1810.06773] Evolutionary Stochastic Gradient Descent for Optimization of Deep Neural Networks](https://arxiv.org/abs/1810.06773) nanotechnology isnt magic, all life runs on it [Imgur: The magic of the Internet](https://imgur.com/Q4lLEsP) [https://youtu.be/7Hk9jct2ozY?si=upbAa8Tya1DseTe7](https://youtu.be/7Hk9jct2ozY?si=upbAa8Tya1DseTe7) poslední dobou mě ještě začaly fascinovat error correcting codes v biologii :smile: [Grid cells generate an analog error-correcting code for singularly precise neural computation](<https://www.nature.com/articles/nn.2901>) [Sci-Hub | Grid cells generate an analog error-correcting code for singularly precise neural computation. Nature Neuroscience, 14(10), 1330–1337 | 10.1038/nn.2901](https://sci-hub.se/https://www.nature.com/articles/nn.2901) error correcing codes jsou potřeba fakt úplně všude... jeden ze vzorů co se objevuje skoro úplně všude ať v systémech co jsme vytvořili, nebo v systémech co vytvořila evoluce, nebo obojí Líbí se mi Veritasium commentary nad tímhle, plus jak ukazuje ještě jiný animace. Hlavně ještě Kinesin walking protein je asi můj oblíbenej. 😄 [https://www.youtube.com/watch?v=X_tYrnv_o6A](https://www.youtube.com/watch?v=X_tYrnv_o6A) (i když tenhle to trochu kritizuje, že měl dát větší emphasis na chaotic complex unpredictable fluid self-organising agential stochastic polyfunkcionalitu částí [https://www.youtube.com/watch?v=jPhvic-eqbc](https://www.youtube.com/watch?v=jPhvic-eqbc) ) Nebo ještě nedávno jsem zkoumal tenhle rotující molecular motor 😄 [https://youtu.be/VPSm9gJkPxU?si=hiAX6BNPWtjiXGQF&t=586](https://youtu.be/VPSm9gJkPxU?si=hiAX6BNPWtjiXGQF&t=586) tady je matika za biologickým motorem [https://www.youtube.com/watch?v=elRbbdJUiTU](https://www.youtube.com/watch?v=elRbbdJUiTU) logický hradla in action! 😄 eventually genes will be transcended by transhumanism and poshumanism, all being part of universal darwinism Why do these self-organizing organic molecules work at all Why is there so many incredibly unreasonably effective branches of math and how do they all connect Fundamental equation of intelligence and of the universe is all you need All futuristic technologies are a decade away for decades until they aren't