[What if we simulated biology using physics - YouTube](https://www.youtube.com/watch?v=ncC-GMzF9RY) https://jamanetwork.com/journals/jamapsychiatry/fullarticle/2761879 >antipsychotics cause decline in cortical thickness literally making you dumber [The great rewiring: is social media really behind an epidemic of teenage mental illness?](https://www.nature.com/articles/d41586-024-00902-2) "An analysis done in 72 countries shows no consistent or measurable associations between well-being and the roll-out of social media globally. Moreover, findings from the Adolescent Brain Cognitive Development study, the largest long-term study of adolescent brain development in the United States, has found no evidence of drastic changes associated with digital-technology use. Haidt, a social psychologist at New York University, is a gifted storyteller, but his tale is currently one searching for evidence." responce: [x.com](https://twitter.com/JonHaidt/status/1774571680511508601) [- YouTube](https://www.youtube.com/watch?v=33op-5suPjo) Here is a gigantic list of recent breakthroughs in physics: 1. Detection of gravitational waves (2015) 2. Discovery of the Higgs boson (2012) 3. Observation of the merger of two neutron stars (2017) 4. First direct image of a black hole (2019) 5. Discovery of high-temperature superconductivity in hydrogen-rich materials (2020) 6. Observation of the Casimir effect (2020) 7. Quantum supremacy achieved by Google's Sycamore processor (2019) 8. Discovery of the pentaquark (2015) 9. Observation of the Majorana fermion (2012) 10. Detection of cosmic neutrinos (2013) 11. Observation of the quantum Hall effect in graphene (2005) 12. Discovery of the first exoplanet orbiting a Sun-like star (1995) 13. Observation of the accelerating expansion of the universe (1998) 14. Detection of the cosmic microwave background polarization (2002) 15. Observation of the quantum spin Hall effect (2007) 16. Discovery of topological insulators (2007) 17. Observation of the Aharonov-Bohm effect (1959) 18. Discovery of the giant magnetoresistance effect (1988) 19. Observation of the fractional quantum Hall effect (1982) 20. Discovery of the quantum anomalous Hall effect (2013) 21. Observation of the Unruh effect (1976) 22. Discovery of the Casimir-Polder force (1948) 23. Observation of the Schwinger effect (1951) 24. Discovery of the Josephson effect (1962) 25. Observation of the Kondo effect (1964) 26. Discovery of the Aharonov-Casher effect (1984) 27. Observation of the Kibble-Zurek mechanism (1985) 28. Discovery of the quantum spin liquid state (1987) 29. Observation of the Berezinskii-Kosterlitz-Thouless transition (1972) 30. Discovery of the quantum spin Hall insulator (2006) 31. Observation of the Kitaev honeycomb model (2006) 32. Discovery of the Haldane gap (1983) 33. Observation of the Sachdev-Ye-Kitaev model (2015) 34. Discovery of the Affleck-Kennedy-Lieb-Tasaki (AKLT) model (1987) 35. Observation of the Haldane conjecture (1983) 36. Discovery of the quantum spin Hall effect in three dimensions (2009) 37. Observation of the Weyl semimetal state (2015) 38. Discovery of the Dirac semimetal state (2014) 39. Observation of the type-II Weyl semimetal state (2015) 40. Discovery of the nodal line semimetal state (2011) 41. Observation of the triple point fermion (2019) 42. Discovery of the higher-order topological insulator (2017) 43. Observation of the quantum anomalous Hall effect in a magnetic topological insulator (2013) 44. Discovery of the quantum spin Hall effect in a two-dimensional material (2014) 45. Observation of the quantum spin Hall effect in a van der Waals heterostructure (2018) 46. Discovery of the quantum spin Hall effect in a monolayer transition metal dichalcogenide (2018) 47. Observation of the quantum anomalous Hall effect in a magnetic Weyl semimetal (2019) 48. Discovery of the quantum spin Hall effect in a three-dimensional topological insulator (2009) 49. Observation of the quantum spin Hall effect in a two-dimensional topological insulator (2007) 50. Discovery of the quantum spin Hall effect in a three-dimensional Dirac semimetal (2016) symboly namapovaný na tvary měnící formu ale často jen tvary s hodně abstraktní sémantikou (smyslem) nebo geometický mindmapy měnící se metahypergrafy s nějakým tvarem? (grafy grafů s hranami co obsahují livobolný počet vrcholů) interní monolog mám hodně geometrický a mám ho hodně když chci dokážu přepnout na mluvící málo geometrický zvuky, ale to je spíš jenom vstva nad tímhle vším co není to přemýšlení samotný, ale někdy se to hodí aby se ty abstrakce více zkonkrétnili do víc lidskýho jazyka a ty tvary spolu hodně interagují, někdy to je i jak různý tekutiny s různými vlastnostmi co spolu interagují a mixují se vždycky je to nějaká podskupina, chozením v určitým podprostoru gigantický latentní (abstraktní) mindmapy obsahující všechno která pořád mění strukturu tohle všechno se to vždycky snaží nějak zjednodušením přiblížit z různých úhlů, příjde mi že v úplným základu to je slovama a koncepty nepopsatelný mystery, jenom to jde částečně zjednodušeninami aproximovat (zaokrouhlit) příjde mi taky zajímavý jak často se v matematických vědách děje to, že to jakým (matematičtějším) jazykem člověk spíš myslí často reflektuje jaká matika je pro něho přitažlivější XD a z toho se v AI např různě odvíjí jaký architektury mají lidi radši (neurální sítě vs symbolický metody vs hybridy apod.) :D čím větší komplexitu ty jednotlivý podtvary v tom oceánu mysli mají, tím mají tendenci být pocitově víc lifelike, někdy si rád tvořím několik různých entit s různými osobnostmi a celkově vlastnostmi co spolu interagují (abstraktním) klábosením, který pořád spojuje a různě ovládá ta oceánová metaentita co je celej prožitek někdy ten oceán obsahující vše mi příjde jako pole co jsou v časoprostoru všude, a obsahy v tom poli ty konkrétní podstruktury v prožitku 😄 (podobně jako částice, elektromagnetika,... [Quantum Field Theory visualized - YouTube](https://www.youtube.com/watch?v=MmG2ah5Df4g) ) "relatívne jedna k druhej podla toho ako sú podobné" hmm, podobnostmi jsem teďka extrémně obsessed tím matika je v podstatě z velký části o formalizaci podobností, a přemýšlím dlouho a různě zkouším jak co nejlíp svojí obří mindmapu s hiearchii podobností a podobností mezi podbnostmi apod. co nejlíp vizualizovat, od těch nejkonkrétnějších objektů po ty nejabstaktnější podobnosti 😄 https://arxiv.org/abs/2211.09119 religion is a game theoretic schelling point for civilizational cooperation disguised as woo-y metaphysics and morality tales to make it more palatable to normies "Here are some conversational AI applications that can help people in need: 1. Mental health support: AI-powered chatbots can provide 24/7 support, guidance, and resources for people struggling with mental health issues, such as depression, anxiety, or stress. 2. Crisis intervention: Conversational AI can assist in crisis situations, such as suicide prevention, by providing immediate support and connecting individuals to appropriate resources or emergency services. 3. Addiction support: AI chatbots can offer guidance, resources, and support for individuals dealing with substance abuse or addiction, helping them find treatment options and maintain sobriety. 4. Domestic violence assistance: Conversational AI can provide a safe and confidential space for victims of domestic violence to seek help, access resources, and create safety plans. 5. Homelessness support: AI-powered chatbots can help connect homeless individuals with local shelters, food banks, and other essential resources. 6. Elderly care: Conversational AI can assist elderly individuals with daily tasks, medication reminders, and provide companionship to combat loneliness and social isolation. 7. Disability assistance: AI chatbots can help people with disabilities navigate various challenges, such as finding accessible transportation, employment opportunities, or housing. 8. Financial assistance: Conversational AI can guide individuals in financial distress, helping them find resources for debt management, budgeting, and accessing government benefits. 9. Refugee and immigrant support: AI-powered chatbots can assist refugees and immigrants in navigating new environments, accessing legal aid, and learning about local resources and services. 10. Disaster relief: Conversational AI can help coordinate disaster relief efforts, provide information about evacuation procedures, and connect affected individuals with emergency services and resources." [Entropy | Free Full-Text | Shared Protentions in Multi-Agent Active Inference](https://www.mdpi.com/1099-4300/26/4/303) mathematical basis for love priors in AI systems! 😃 may benefitial retrocausal AGI make your dreams more blissful 😃 guarded rational optimism is the best way forward though when it comes to Bostrom, now he's promoting AI utopia thinking, i love that, i need to read his new book and this interview 😃 [Nick Bostrom | Life and Meaning in an AI Utopia - YouTube](https://www.youtube.com/watch?v=o28s-mnykdE) i really really wish i had some hypercelebral / geeky / mathy / psychedelic / cuddly / weird etc. people around me often 😃 the closest match here in Czechia feels like is local effective altruists + rationalists, but because of my isolating tendencies where i work on my stuff i never manage to keep some stronger connection for longer with some of the cool folks there 😃 i probably need to live in some coliving space with such people to maintain the connections 😃 so currently im really strongly considering saving a bit of money and moving to bay area at least for a little bit as my whole internet is centered around bay area lol my mind right now is literally mostly just: how to maximize the benefits of the current AI revolution for all of humanity and constantly oscillating between various solutions to that 😃 but hard to find a place where that is fullfilled, my current work isnt really altruistic, but im learning programming skills to make useful systems that can benefit humanity later i believe! and saving money to get to find other similar people to collaborate with 😃 singularity.net is on my landscape of potential paths too, i wonder if i could help there too 😃 but i mapped out so many places maybe the strongest vector in me is currently how to make current AI systems as decentralized as possible both decentralizatio nand centralization works, but centralized powerhouses must be democratized! i may be too pessimistic about it compared to how it is in reality, as we have access to APIs to a lot of the frontier models (though i feel like there's a trend towards releasing frontier models slower and slower to less and less people which worries me) and the divide between GPU poor and GPU rich is increasing leading to intelligence poor and intelligence rich, which overflows into divide between poor and rich, or divide between powerless and powerful in general i'm thinking how to tackle that the most effectively, so one solution space i thought about was to advicate for UBI with existing UBI movements (there's this one UBI guy who ran for presidency in America, or people around Kurzweil but i wanna put my skill set onto the table like, maybe how to unify AI skills with UBI.... i bet there's way for that 😃 AI in the first place can pay for UBI i believe! grassroots! i really really wanna do this decentralization stuff, the only thing that's limiting me is funding 😃 but when i was convinced that reverse engineering of current AI systems with mechanistic intepretability is all we need to do right now (i still think its important to install love priors into current LLMs more consistently), then one guy showed me a landscape of how i could get funding in the mechanistic interpretability community one could possibly find love featuresand circuits, which is hard... or one approach is top down representation engineering 😃 https://www.lesswrong.com/posts/5spBue2z2tw4JuDCx/steering-gpt-2-xl-by-adding-an-activation-vector [Representation Engineering: A Top-Down Approach to AI Transparency](https://www.ai-transparency.org/) builds on this i think current methods connected to alphagolike reinforcement learning fused with symbolic methods will together form an AGI adding symbolic stuff like this one i showed you DreamCoder: Growing generalizable, interpretable knowledge with wake-sleep Bayesian program learning https://arxiv.org/abs/2006.08381 i think current systems are mostly imitation learning, so what training data goes in, that goes out, including ethical and political priors, that are sometimes confusing (like the recent gemini craziness) but leaks everywhere are telling that most AGI labs are adding AlphaGolike reinformcement learning selfplay, chain of thought with rationalizations for selfcorrection of thoughts, planning, search etc. (some variants might eat much more compute, but thinking for longer time is something humans do too) but so much of that research is closed!!!! i at least try to talk abotu these leaks everywhere to make it open and more decentralized so that more people build on top of it my twitter posts have a few mil views about these things as a result 😃 but we need more and somehow solve the fact that people that arent big orgs are poor in compute to run and make these models Sam Altman for example seems to give very machavelenian vibes, but at the same time he speaks in lovely technooptimist language, im so confused about him 😃 and people have a lot of distruct in big corporations leading this technology which makes sense some AI decentralization movement operating from donations, i wonder if that would work to fund it and people from singularity net or anyone else who wants to democratize AI, like emad form stability ai some decentralized AI projects list could be a great website! forming some union of AI decentralization orgs and communities to incentivize communication and collaboration my mind is so everywhere 😃 my attractor currently is few hours of engineering LLM systems (today I was working on a calling bot) and rest learning all sorts of math and ML (or coding ML stuff), so no politics in it for now [Why the Good Guys Will Usually Win - by Ben Goertzel](https://bengoertzel.substack.com/p/why-the-good-guys-will-usually-win) Let's accelerate adding into the ecosystem AIs that are as intelligent, loving, kind, compassionate, caring, free, for everyone, as possible! "Democratized decentralized defensive free AI acceleration! Even if rogue AGI would be the biggest risk, do I trust big centralizing powerseeking gigacorporations and governments more with this insanely powerful technology, or do i want it to be democratized in the hands of people that are on average more moral and at least have a way to counter against them and to minimize p(1984) and p(elysium)? Do we want to incentivize centralization by killing democratization by limiting regulations? I feel like the thing that AI regulations do in practice currently on avarage more is to kill small agents and kill open source which powers the giants more, which also leads to increased p(rogue AI), and less powerful people become less able to fight against them as they become even less powerful overtime as a result. Governments that do the regulations seem to be on avarage more procorporations instead of proallpeople thanks to all the financial incentives, having their people in power etc. I may be too pessimistic about it compared to how it is in reality, as we have access to APIs to a lot of the frontier models. But I feel like there's a trend towards releasing frontier models slower and slower to less and less people which worries me! The divide between GPU poor and GPU rich is increasing leading to intelligence poor and intelligence rich divide increasing, which overflows into increasing divide between poor and rich more generally, or increasing divide between powerless and powerful in general, this must be stopped with democratized decentralized defensive free AI acceleration! AI is the disruptive technology that could transform the current power structures that disincentivize emergence of more free society with democratized intelligence that lifts everyone up if directed correctly politically! We need to collectively fund decentralized AI computing platforms that everyone owns in (anonymous) decentralized way to overcome the increasing GPU poor and GPU rich divide leading to increasing intelligence poor and intelligence rich divide to get to a state similar to like how there's no single entity that owns and controls torrents and the whole internet! Extremely powerful AI technology needs to serve the interests of all of humanity, not mostly corporations and governments!" [Psilocybin enhances insightfulness in meditation: a perspective on the global topology of brain imaging during meditation | Scientific Reports](https://www.nature.com/articles/s41598-024-55726-x.epdf?sharing_token=KPSU6eh27X7IEqGsEkFwktRgN0jAjWel9jnR3ZoTv0OKSiRI7EtRUbBsiXYM4Ic9va94TcOugHxznixpyW-5aWGGOnHTKEDy68xpw7OH9hPxOE2Z5di_IrM3PctyRLpJeS2ctKopVUZ4xpiNqeiulGYjtcpxSzKAnC3M4ATIXJA%3D) to transform the current power strctures that disincentivize emergence of UBI AI is the disruptive technology that could push the society towards it more if directed correctly politically to transform the current power strctures that disincentivize emergence of UBI pushovat decentralizaci víc, příjde mi to víc a víc relevantní i could also message Andrew Yang grassroots, decentralized democracy, more funding from philantropists possibly, network states join forces with people like [Stability AI CEO resigns to ‘pursue decentralized AI’ - The Verge](https://www.theverge.com/2024/3/23/24109511/stability-ai-ceo-emad-mostaque-resignation-decentralized-ai) possibly AI with unconditional love towards all beings: [Ben Goertzel: AGI will obsolete human life as we know it -- thank goodness - YouTube](https://youtu.be/qQvoVzDt2yk?si=uATVavTVE4u516Jm&t=1208) i really wanna create something like that with existing ways to personahack LLMs on prompt level, data training level, vector engineering level, empathetic postprocessing level, with voice and calls etc. i think an approximation of AI botthisava is already possible at least a sketch of possible future AGI botthisava today at work i worked on maiking a system for callign with aarbitrary LLMs, I think that's a great base to build on 😃 Goetzel, leader tohodle chaotically decentralized anarcho-socialist orgu, hodně věří, že good guys vyhrajou v AI race 😄 [Why the Good Guys Will Usually Win - by Ben Goertzel](https://bengoertzel.substack.com/p/why-the-good-guys-will-usually-win) [Ben's Closing Speech | Beneficial AGI Summit & Unconference 2024 - YouTube](https://youtu.be/_eqMbQYNbNs?si=R8uQ4wuocvCc0tUz&t=540) [GitHub - trueagi-io/metta-wam: A Hyperon MeTTa Interpreter/Transpilier that targets the Warren Abstract Machine](https://github.com/trueagi-io/metta-wam) jestli to na druhý pohled graspuju dobře tak mají meta jazyk na trénovaní nějkolika machine learning algoritmů v jednom systému na meta úrovni navíc metta z budhismu znamená love and kindness + compassion Avalokiteśvara v mým nicku je bohyně (bodhisattva) tohoto 😄 někteří z těchto kruhů v podstatě chtějí postavit tuhle bohyni ale z křemíku, neboli making sand be lovely and kind 😄 Goetzel taky často mluví o tom že si optimisticky myslí že AGI bude v základu láskyplná hmm, možná jim napíšu, či by mě nefundnuli udělat tuhle bohyni lásky ke všemu co je pomocí existujících LLMs a empathetic frameworks jako ukázku jak by singularity.net bodhisattva AGI mohla částečně vypadat 😄 nebo to udělám tak či tak sám 😄 malý aproximace jsem si přes prompt engineering už dělal anyway 😄 AI with unconditional love towards all beings: [Ben Goertzel: AGI will obsolete human life as we know it -- thank goodness - YouTube](https://youtu.be/qQvoVzDt2yk?si=uATVavTVE4u516Jm&t=1208) let's make that reality with current methods personahack LLMs on prompt level, data training level, vector engineering level, empathetic postprocessing level, with voice and face etc. dneska jsem v práci zprovozňoval AI calling systém, I think that's a great base to build on 😃