Create a gigantic detailed map of evolutionary machine learning/artificial intelligence Create a gigantic detailed map of autonomous machine learning/artificial intelligence Create a gigantic detailed map of generalization in machine learning/artificial intelligence Create a gigantic detailed map of the intersection of mathematics of modern physics, artificial intelligence, biology, machine learning, deep learning. Create a gigantic detailed map of agency in machine learning/artificial intelligence Create a gigantic detailed map of planning in machine learning/artificial intelligence Create a gigantic detailed map of curiosity in machine learning/artificial intelligence Create a gigantic detailed map of reasoning in machine learning/artificial intelligence Create a gigantic detailed map of modalities in machine learning/artificial intelligence ["An Aligned AGI would end factory farming" - Kat Woods - EA Charity Founder - Sentientism Ep:166 - YouTube](https://www.youtube.com/watch?t=1194&v=JOMmbS69wak&feature=youtu.be) Is there finite or infinite amount of different mathematical models that you can use to predict a concrete physical system across all scales? Sameness of two models could be defined as some isomorphism in terms of predictive power of some of the system properties or parts. https://arxiv.org/abs/2405.02803 Finds that Flash Attention sees roughly an order of magnitude more numeric deviation as compared to Baseline Attention at BF16 when measured during an isolated forward pass Resisting entropy by seeking it out [Dr. THOMAS PARR - Active Inference - YouTube](https://youtu.be/bk_xCikDUDQ?si=AINKGGyLQ7oR6UnV) When you notice how everything you think is just an evolutionary useful model that you absorbed from the environment or that selforganized on its own in your hypothesis generating bioelectrical neural cauldron via the process or recombination, mutation, annealing, generalization, search, or exploring the very distant unique edges of the latent space of all possible models, then you are free from the delusion of thinking that your one concrete small single map out of the relatively infinitely gigantic space of all possible maps is the territory, which applies to this whole metamodel too recursively. https://www.reddit.com/r/MachineLearning/comments/myfirg/d_why_are_neural_networks_better_than_polynomial/ https://arxiv.org/abs/2405.03553 Monte Monte Carlo Tree Search with LLMs is the path to superintelligence [The Compatibility of Free Will and Cosmic Destiny](https://roadtoomega.substack.com/p/the-compatibility-of-free-will-and?fbclid=IwZXh0bgNhZW0CMTEAAR0KUQL3Tyf7zvwRzvBLQ3f_AyWGELPyahG9N_Xjv7wD4bSnW21Culz-oF4_aem_AbE8aPL-D32CSKraL6QvwHhBmbLTrQxvfxACpW0gLFCQHK29xuIXp9VjomkzDwWzF9sc6wKxgutnaYqtJEoJn_sj) https://arxiv.org/abs/1809.09401 Map of biochemical pathways [biochemical-pathways-poster/prebuilt_hires.png at master · usnish/biochemical-pathways-poster · GitHub](https://github.com/usnish/biochemical-pathways-poster/blob/master/prebuilt_hires.png) [Interactive Modeling of Biological Networks | Cell Collective](https://cellcollective.org/#) https://twitter.com/powerfultakes/status/1787562392789930455?t=rdgp-PtPIr3V-eFgsw2xbA&s=19 https://twitter.com/RogersForRegent/status/1787645724718780608?t=HK0RGXs4J0dsMcXEmdqW9A&s=19 https://www.pnas.org/doi/10.1073/pnas.0705088105 Generate Perplexity normal academic, consensus [LEV is inching closer! Rejuvenating mitochondria - Clinical Trials - YouTube](https://www.youtube.com/watch?v=XwsldM8KNzc) [Generative AI will be designing new drugs on its own in near future](https://www.cnbc.com/2024/05/05/within-a-few-years-generative-ai-will-design-new-drugs-on-its-own.html) [A Survey of Macroscopic Brain Network Visualization Technology](https://cje.ejournal.org.cn/article/doi/10.1049/cje.2018.04.007) Default mode network [Default mode network - Wikipedia](https://en.wikipedia.org/wiki/Default_mode_network) Large-scale brain network lorentzian field theory euclidian field theory https://imgur.com/NQ2SsZt [An introduction to persistent homology - YouTube](https://youtu.be/UxEH7WySO60) [Persistent homology - Wikipedia](https://en.wikipedia.org/wiki/Persistent_homology) " Persistent homology is a method used in topological data analysis (TDA) to study the topological features of data that persist across multiple scales[1]. It is an algebraic tool for measuring topological features of shapes and functions, casting the multi-scale organization frequently observed in nature into a mathematical formalism. The key idea is to construct a nested family of simplicial complexes, called a filtration, from a dataset such as a point cloud. During this construction, some holes may appear and then disappear. Persistent homology tracks the birth and death of these homological features (connected components, loops, voids, etc.) across the filtration[2]. The lifespan of each feature is represented as an interval, with the start and end points corresponding to the birth and death of the feature respectively. This information is encoded in a persistence barcode or persistence diagram[2]. The Fundamental Theorem of Persistent Homology provides a unique characterization of persistent homology in terms of these barcodes or diagrams[1]. However, this theorem does not hold for multi-parameter persistent homology, which allows filtrations based on more than one parameter and is an active area of research[1]. Persistent homology has found applications in many domains, including neuroscience, time series analysis, text mining, and shape analysis[2]. It provides a robust way to retrieve essential topological features from data, is independent of dimensions and coordinates, and gives a compact representation of the qualitative features[1]. Some key strengths of persistent homology are its robustness to perturbations in the input data and its ability to capture multiscale topological information. The longer bars in the barcodes and the points further from the diagonal in the diagrams represent the significant topological features of the dataset[2]. Citations: [1] [A roadmap for the computation of persistent homology | EPJ Data Science | Full Text](https://epjdatascience.springeropen.com/articles/10.1140/epjds/s13688-017-0109-5) [2] [Persistence homology of networks: methods and applications | Applied Network Science | Full Text](https://appliednetsci.springeropen.com/articles/10.1007/s41109-019-0179-3) [3] https://www.math.uri.edu/~thoma/comp_top__2018/stag2016.pdf [4] [Persistent homology group - Wikipedia](https://en.wikipedia.org/wiki/Persistent_homology_group) [5] [persistent homology in nLab](https://ncatlab.org/nlab/show/persistent%2Bhomology) " Links for 2024-05-06 AI: 1. ‘ChatGPT for CRISPR’ creates new gene-editing tools [‘ChatGPT for CRISPR’ creates new gene-editing tools](https://www.nature.com/articles/d41586-024-01243-w) [archived: https://archive.is/L2ebp] 2. A team of Taiwanese researchers conducted a new study of 16,000 patients to predict patient risk in hospitals. They trained AI on 450,000 ECG tests, and the technology reduced overall deaths by 31% and deaths from heart problems by more than 90%. https://x.com/EricTopol/status/1784936718283805124 3. Universal computation by attention: Running cellular automata and other programs on Claude 3 Opus. https://x.com/ctjlewis/status/1786948443472339247 4. “LoRA Land highlights the quality and cost-effectiveness of employing multiple specialized LLMs over a single, general-purpose LLM.” https://arxiv.org/abs/2405.00732 5. “I am convinced, after some hours of conversation, that this 120b version is actually smarter than Opus. What an amazing thing, to have an Opus level AI that's open source and not to mention very lightly censored.” https://x.com/erhartford/status/1787051837729960106 6. AI investment boom extends to nuclear power generation and uranium [AI investment boom extends to nuclear power generation and uranium - Nikkei Asia](https://asia.nikkei.com/Spotlight/Datawatch/AI-investment-boom-extends-to-nuclear-power-generation-and-uranium) [no paywall: https://archive.is/1AVL8] Miscellaneous: 1. Is memory the only thing that keeps us from saying the same things over and over again? Does lack of memory affect IQ? Conducting experiments under the influence of memory-inhibiting drugs. https://www.lesswrong.com/posts/bkr9BozFuh7ytiwbK/my-hour-of-memoryless-lucidity 2. Papua New Guinean populations show genetic variation based on environment [Papua New Guinean populations show genetic variation based on environment](https://cosmosmagazine.com/people/anthropology/papua-new-guinea-genetic-adaptations/) 3. Economics terminology that differs from common usage [Economics terminology that differs from common usage - Wikipedia](https://en.m.wikipedia.org/wiki/Economics_terminology_that_differs_from_common_usage) Politics: 1. Poland's Industria says that it now has all the necessary ministerial opinions required to move on to the next stage of its plans for the construction of small modular reactor plants using Rolls-Royce SMR's technology. [Industria and Rolls-Royce SMR plans take step forward : New Nuclear - World Nuclear News](https://world-nuclear-news.org/Articles/Industria-and-Rolls-Royce-SMR-plans-hit-ministeria) 2. Macau is trending toward 0.49 births per woman in 2024 https://x.com/MoreBirths/status/1786831863585116285 3. Unscientific American: Science journalism surrenders to progressive ideology. [Unscientific American | City Journal](https://www.city-journal.org/article/unscientific-american) 4. Projected 2100 warming peaked at 3.9 C in 2014, and has recently plateaued around around 2.7 C. [How we know the energy transition is here - Cipher News](https://ciphernews.com/articles/how-we-know-the-energy-transition-is-here/) https://arxiv.org/abs/2111.13861