for - quote - AI - AI is accelerating AI research itself

for - AI - Chinese know more than US government about latest US frontier AI research

https://web.archive.org/web/20250423134653/https://www.rijksoverheid.nl/documenten/publicaties/2025/04/22/het-overheidsbrede-standpunt-voor-de-inzet-van-generatieve-ai Rijksstandpunt genAI, mede gebaseerd op IEC advies IPO. Niettemin wordt het hier lijkt me behoorlijk vrij gegeven, en de formulering klinkt heel los. Gaat problemen opleveren, want een bmw die met genAI speelt bij het opstellen van een stuk het voor zichzelf als 'experiment' labelt of 'innovatie' heeft het voor zich daarmee gerationaliseerd. Never mind dat experimenten gecontroleerde omstandigheden vergen, en innovatie een gedeelde intentie moet hebben in de org. Dit voelt heel zacht aan, staan de juiste dingen in desondanks

[[When Will the GenAI Bubble Burst]]

https://www.carbonbrief.org/explainer-how-china-is-managing-the-rising-energy-demand-from-data-centres/?utm_source=cbnewsletter&utm_medium=email&utm_term=2025-04-16&utm_campaign=Daily+Briefing+16+04+2025

The crime was misleading investors, not anyone else, which is very telling. The hype around "AI" - and actually hiring remote workers to do the job - and misleading customers/users doesn't matter.

Yet another example of "AI" being neither artificial nor intelligent.

Interesting cautionary tale about building out DCs in the styx, where energy is cheap but latency is high

Need to try this out.

Review coordinated by Life Science Editors Foundation Reviewed by: Dr. Angela Andersen, Life Science Editors Foundation & Life Science Editors Potential Conflicts of Interest: None

PUNCHLINE Evo 2 is a biological foundation model trained on 9.3 trillion DNA bases across all domains of life. It predicts the impact of genetic variation—including in noncoding and clinically relevant regions—without requiring task-specific fine-tuning. Evo 2 also generates genome-scale sequences and epigenomic architectures guided by predictive models. By interpreting its internal representations using sparse autoencoders, the model is shown to rediscover known biological features and uncover previously unannotated patterns with potential functional significance. These capabilities establish Evo 2 as a generalist model for prediction, annotation, and biological design.

BACKGROUND A foundation model is a large-scale machine learning model trained on massive and diverse datasets to learn general features that can be reused across tasks. Evo 2 is such a model for genomics: it learns from raw DNA sequence alone—across bacteria, archaea, eukaryotes, and bacteriophage—without explicit labels or training on specific tasks. This enables it to generalize to a wide range of biological questions, including predicting the effects of genetic variants, identifying regulatory elements, and generating genome-scale sequences or chromatin features.

Evo 2 comes in two versions: one with 7 billion parameters (7B) and a larger version with 40 billion parameters (40B). These numbers reflect the number of trainable weights in the model and influence its capacity to learn complex patterns. Both models were trained using a context window of up to 1 million tokens—where each token is a nucleotide—allowing the model to capture long-range dependencies across entire genomic regions.

Evo 2 learns via self-supervised learning, a method in which the model learns to predict masked or missing DNA bases in a sequence. Through this simple but powerful objective, the model discovers statistical patterns that correspond to biological structure and function, without being told what those patterns mean.

QUESTION ADDRESSED Can a large-scale foundation model trained solely on genomic sequences generalize across biological tasks—such as predicting mutational effects, modeling gene regulation, and generating realistic genomic sequences—without supervision or task-specific tuning?

SUMMARY The authors introduce Evo 2, a foundational model for genomics that generalizes across DNA, RNA, and protein tasks. Without seeing any biological labels, Evo 2 learns the sequence rules governing coding and noncoding function, predicts variant effects—including in BRCA1/2 and splicing regions—and generates full-length genomes and epigenome profiles. It also enables epigenome-aware sequence design by coupling sequence generation with predictive models of chromatin accessibility.

To probe what the model has learned internally, the authors use sparse autoencoders (SAEs)—a technique that compresses the model’s internal activations into a smaller set of interpretable features. These features often correspond to known biological elements, but importantly, some appear to capture novel, uncharacterized patterns that do not match existing annotations but are consistently associated with genomic regions of potential functional importance. This combination of rediscovery and novelty makes Evo 2 a uniquely powerful tool for exploring both the known and the unknown genome.

KEY RESULTS Evo 2 trains on vast genomic data using a novel architecture to handle long DNA sequences Figures 1 + S1 Goal: Build a model capable of representing entire genomic regions (up to 1 million bases) from any organism. Outcome: Evo 2 was trained on 9.3 trillion bases using a hybrid convolution-attention architecture (StripedHyena 2). The model achieves long-context recall and strong perplexity scaling with increasing sequence length and model size.

Evo 2 predicts the impact of mutations across DNA, RNA, and protein fitness Figures 2A–J + S2–S3 Goal: Assess whether Evo 2 can identify deleterious mutations without supervision across diverse organisms and molecules. Outcome: Evo 2 assigns lower likelihoods to biologically disruptive mutations—e.g., frameshifts, premature stops, and non-synonymous changes—mirroring evolutionary constraint. Predictions correlate with deep mutational scanning data and gene essentiality assays. Evo 2 embeddings also support highly accurate exon-intron classifiers.

Clarification: “Generalist performance across DNA, RNA, and protein tasks” means that Evo 2 can simultaneously make accurate predictions about the functional impact of genetic variants on transcription, splicing, RNA stability, translation, and protein structure—without being specifically trained on any of these tasks.

Evo 2 achieves state-of-the-art performance in clinical variant effect prediction Figures 3A–I + S4 Goal: Evaluate Evo 2's ability to predict pathogenicity of human genetic variants. Outcome: Evo 2 matches or outperforms specialized models on coding, noncoding, splicing, and indel variants. It accurately classifies BRCA1/2 mutations and generalizes to novel variant types. When paired with supervised classifiers using its embeddings, it achieves state-of-the-art accuracy on BRCA1 variant interpretation.

Evo 2 representations reveal both known and novel biological features through sparse autoencoders Figures 4A–G + S5–S7 Goal: Understand what Evo 2 has learned internally. Outcome: Sparse autoencoders decompose Evo 2’s internal representations into distinct features—many of which align with well-known biological elements such as exon-intron boundaries, transcription factor motifs, protein secondary structure, CRISPR spacers, and mobile elements. Importantly, a subset of features do not correspond to any known annotations, yet appear repeatedly in biologically plausible contexts. These unannotated features may represent novel regulatory sequences, structural motifs, or other functional elements that remain to be characterized experimentally.

Note: Sparse autoencoders are neural networks that reduce high-dimensional representations to a smaller set of features, enforcing sparsity so that each feature ideally captures a distinct biological signal. This approach enables mechanistic insight into what the model “knows” about sequence biology.

Evo 2 generates genome-scale sequences with realistic structure and content Figures 5A–L + S8 Goal: Assess whether Evo 2 can generate complete genome sequences that resemble natural ones. Outcome: Evo 2 successfully generates mitochondrial genomes, minimal bacterial genomes, and yeast chromosomes. These sequences contain realistic coding regions, tRNAs, promoters, and structural features. Predicted proteins fold correctly and recapitulate functional domains.

Evo 2 enables design of DNA with targeted epigenomic features Figures 6A–G + S9 Goal: Use Evo 2 to generate DNA sequences with user-defined chromatin accessibility profiles. Outcome: By coupling Evo 2 with predictors like Enformer and Borzoi, the authors guide generation to match desired ATAC-seq profiles. Using a beam search strategy—where the model explores and ranks multiple possible output sequences—it generates synthetic DNA that encodes specific chromatin accessibility patterns, such as writing “EVO2” in open/closed chromatin space.

STRENGTHS First large-scale, open-source biological foundation model trained across all domains of life

Performs well across variant effect prediction, genome annotation, and generative biology

Demonstrates mechanistic interpretability via sparse autoencoders

Learns both known and novel biological features directly from raw sequence

Unsupervised learning generalizes to clinical and functional genomics

Robust evaluation across species, sequence types, and biological scales

FUTURE WORK & EXPERIMENTAL DIRECTIONS Expand training to include viruses that infect eukaryotic hosts: Evo 2 currently excludes these sequences, in part to reduce potential for misuse and due to their unusual nucleotide structure and compact coding. As a result, Evo 2 performs poorly on eukaryotic viral sequence prediction and generation. Including these genomes could expand its applications in virology and public health.

Empirical validation of novel features: Use CRISPR perturbation, reporter assays, or conservation analysis to test Evo 2-derived features that don’t align with existing annotations.

Targeted mutagenesis: Use Evo 2 to identify high-impact or compensatory variants in disease-linked loci, and validate using genome editing or saturation mutagenesis.

Epigenomic editing: Validate Evo 2-designed sequences for chromatin accessibility using ATAC-seq or synthetic enhancer assays.

Clinical applications: Fine-tune Evo 2 embeddings to improve rare disease variant interpretation or personalized genome annotation.

Synthetic evolution: Explore whether Evo 2 can generate synthetic genomes with tunable ecological or evolutionary features, enabling testing of evolutionary hypotheses.

AUTHORSHIP NOTE This review was drafted with support from ChatGPT (OpenAI) to help organize and articulate key ideas clearly and concisely. I provided detailed prompts, interpretations, and edits to ensure the review reflects an expert understanding of the biology and the paper’s contributions. The final version has been reviewed and approved by me.

FINAL TAKEAWAY Evo 2 is a breakthrough in foundation models for biology—offering accurate prediction, functional annotation, and genome-scale generation, all learned from raw DNA sequence. By capturing universal patterns across life, and identifying both well-characterized and unknown sequence features, Evo 2 opens powerful new directions in evolutionary biology, genomics, and biological design. Its open release invites widespread use and innovation across the life sciences.

In our paper, we address this idea that ALL ✨sparkling intelligence✨ outputs are generated using the same technology and practices. We argue that it is useful to have a term for those outputs that don't match our shared reality or factual requirements, and for that we propose "mirage".

In my blog post introducing our paper, I suggest AI "rainbows" as a term for mirages that we do value.

I'm assuming this is some kind of ✨sparkling intelligence✨ and given that Dr. Oblivion seems to miss the point of the paper and our discussion here, I found it more illustrative than helpful ;)

Turns out the story is a bit more complicated than that, at least according to the history shared by another participant below.

This is a great history of the term "hallucination" in the discourse of ✨sparkling intelligence✨ — huge thanks to whoever shared it! I've also added it to our collaborative bibliography.

deepseek推理型 #AI #大模型

for - program event selection - 2025 - April 4 - 10:30am-12pm GMT - Skoll World Forum - The Future of AI & Digital Innovation - Stop Reset Go - Indyweb -- relevant to

for - program event selection - 2025 - April 2 - 2-3:15pm GMT - Skoll World Forum - The Changing State of AI, Media - Indyweb - Stop Reset Go - TPF - Eric's project - Skoll's Participatory Media project - relevant to - adjacency - indyweb - Stop Reset Go - participatory news - participatory movie and tv show reviews - Eric's project - Skoll's Particiipatory Media - event time conflict - with - Leadership in Alien Times

adjacency - between - Skoll's Participatory Media project - Global Witness - Indyweb - Stop Reset Go's participatory news idea - Stop Reset Go's participatory movie and TV show review idea - Eric's media project - adjacency relationship - Participatory media via Indyweb and idea of participatory news and participatory movie and tv show reviews - might be good to partner with Skoll Foundation's Participatory Media group

It's fascinating that despite the surge in advanced deep learning methods, traditional non-AI approaches like CellProfiler continue to deliver superior performance in cell segmentation.

for - claim - AI for a new type of democracy? - progress trap - AI - future democracy

Anshumali's prime research work on SLIDE algorithms.

RLHF

for - open source AI

for - Indyweb dev - open source AI - text to graph - from - search - image - google - AI that converts text into a visual graph - https://hyp.is/KgvS6PmIEe-MjXf4MH6SEw/www.google.com/search?sca_esv=341cca66a365eff2&sxsrf=AHTn8zoosJtp__9BMEtm0tjBeXg5RsHEYA:1741154769127&q=AI+that+converts+text+into+visual+graph&udm=2&fbs=ABzOT_CWdhQLP1FcmU5B0fn3xuWpA-dk4wpBWOGsoR7DG5zJBjLjqIC1CYKD9D-DQAQS3Z598VAVBnbpHrmLO7c8q4i2ZQ3WKhKg1rxAlIRezVxw9ZI3fNkoov5wiKn-GvUteZdk9svexd1aCPnH__Uc8IUgdpyeAhJShdjgtFBxiTTC_0C5wxBAriPcxIadyznLaqGpGzbn_4WepT8N6bRG3HQLK-jPDg&sa=X&ved=2ahUKEwju5oz8ovKLAxW6WkEAHaSVN98QtKgLegQIEhAB&biw=1920&bih=911&dpr=1 - to - example - open source AI - convert text to graph - https://hyp.is/UpySXvmKEe-l2j8bl-F6jg/rahulnyk.github.io/knowledge_graph/

for - Indyweb dev - text to graph - open source AI - convert text to graph - adjacency - infranodus - to - AI program to convert text into visual graph

for - Indyweb dev - open source AI - text to graph - open source AI - text to graph - from - article - Medium - How to Convert Any Text Into a Graph of Concepts - https://hyp.is/vu53YvmIEe-DuHvXodWFAA/medium.com/towards-data-science/how-to-convert-any-text-into-a-graph-of-concepts-110844f22a1a

for - search - Google - image - AI that converts text into any visual graph - https://www.google.com/search?sca_esv=341cca66a365eff2&sxsrf=AHTn8zoosJtp__9BMEtm0tjBeXg5RsHEYA:1741154769127&q=AI+that+converts+text+into+visual+graph&udm=2&fbs=ABzOT_CWdhQLP1FcmU5B0fn3xuWpA-dk4wpBWOGsoR7DG5zJBjLjqIC1CYKD9D-DQAQS3Z598VAVBnbpHrmLO7c8q4i2ZQ3WKhKg1rxAlIRezVxw9ZI3fNkoov5wiKn-GvUteZdk9svexd1aCPnH__Uc8IUgdpyeAhJShdjgtFBxiTTC_0C5wxBAriPcxIadyznLaqGpGzbn_4WepT8N6bRG3HQLK-jPDg&sa=X&ved=2ahUKEwju5oz8ovKLAxW6WkEAHaSVN98QtKgLegQIEhAB&biw=1920&bih=911&dpr=1

search - google - image - AI that converts text into visual graph - interesting results returned - to - article - Medium - How to convert any text into a graph of concepts -

I wonder how "purpose" is defined here. Is this why LLM applications were developed, or something they were meant to achieve?

Is this an example of personalization and making AI an ally, or of locking the student into a turbocharged LMS?

Se pot crea DOAR 3 note în varianta FREE!

Fake citation AI generated

You might also want to visit my blog post, where I introduce the publication of this paper alongside some additional ideas on interventions to prevent AI mirages, on AI mirages vs AI rainbows, and on how AI terminology plays out in different disciplines.

for - AI - as extreme human echo chamber - Jonathan Boymal - AI

on copyright and artistic style

How much of a defense had they put up since they went out of business?

could have far-reaching implications. Consider how music and image generators compete.

may be useful in revising the Copyright Card Game to addresss AI

for - biggest worry of silicon valley longterminists - AI takeover, not climate crisis - SOURCE - article - Guido Palazzo

for - longterminism - 2 fundamental technologies - rockets and AI

Im Standard stellt Martin Auber mit aktuellen Daten belegt dar, warum der bloße Ausbau der Kapazitäten zur Erzeugung erneuerbarer Energien nicht zu einer Dekabonisierung führen wird. Der Energiebedarf wächst wesentlich schneller als die zur Verfügung stehende erneuerbare Energiepunkt. Durch den KI-Boom wird er noch einmal deutlich gesteigert. https://www.derstandard.at/story/3000000255154/wann-kommt-die-energiewende-oder-kommt-sie-gar-nicht

MAPPING SOCIAL CHOICE THEORY TO RLHF Jessica Dai and Eve Fleisig ICLR Workshop on Reliable and Responsible Foundation Models 2024

Nice overview of how social choice theory has been connected to RLHF and AI alignment ideas.

for - comment - post - LinkedIn - Bayo - AI

for - quote - youtube - interview - Geoffrey Hinton - AI - spiritual crisis - AI - Geoffrey Hinton - self - spiritual crisis

quote - AI - spiritual crisis - We only have the dimmest understanding of, at present the nature of people and what it means to have a self - We don't understand those things very well and they're becoming crucial to understand because we're now creating beings - (interviewer: so this is becoming a philosophical, perhaps even spiritual crisis as a practical one) - Absolutely, yes

for - Poincare - AI question - SOURCE - paper - Man-Computer Symbiosis - J.C.R. Licklider - 1960 - referred by - Gyuri

for - AI - progress trap - Youtube - bioweapons is not the only threat. Nano technology and many others can be turned into weapons of mass destruction - RDeepSeek R1 just caught up with OpenAIs o1 - There is no moat@ What does this mean? - David Shapiro - 2025, Jan 29

These can be helpful for you, but there are also serious concerns. • Ai can change the authenticity of your writing, turning into a “voice” that is not your own. For example, Grammarly often changes my word choices so they don’t sound like something I’d actually say. That goes beyond just checking grammar. • It can definitely lead to plagiarism, basically creating something that is not from you. • The information is often incorrect or made up, for example citing resources that don’t actually exist.

This resonates with me, so I think after I use grammar correction, I still need to go back and check my writing to express my ideas in a way that suits my style and tone.

for - book - Burnout from Humans: A little book about AI that is not really about AI - Aiden Cinnamon Tea & Dorothy Ladybugboss - 2024

introductory AI courses at Rice University

March 31st 2025

This a reference to sacrifice zones, and not increasing the impact in overburdened. Ommunites already?

This counts out nukes

Wow, this is two of the three pillars here pretty explicitly. Additionally is less well defined, but it’s sort of implied elsewhere

This seems to rule out the big three / four

这篇文章由《大西洋月刊》科学台的记者们撰写，列举了2024年让他们震惊的77个事实，涵盖了历史、科技、自然、健康等多个领域，展现了世界的奇妙和多样性。以下是这些事实的详细总结：

历史与文化

1. 在法国印第安人战争期间，洋葱被用来治疗伤口。
2. 1998年，艾伦·索金坚持不为他的首部电视剧《体育之夜》添加笑声轨，但该剧仍以笑声轨首播。
3. 漫展无衬线体（Comic Sans）最初是为一个教人如何使用个人电脑的动画狗程序设计的。
4. 古代，人们会带着装饰华丽的宠物松鼠散步。
5. 西格蒙德·弗洛伊德让病人躺在沙发上，因为他无法忍受“整天被别人盯着看”。
6. 1993年，科学家向太平洋倾倒了近1000磅的铁晶体，以从空气中吸收二氧化碳。
7. 2008年，中国使用人工降雨技术在奥运会前清理北京上空。
8. 在亚利桑那州的古德伊尔，一个用于生成式人工智能的数据中心每年可能消耗多达5600万加仑的饮用水。

科技与创新

1. 生成式人工智能的繁荣成本可能超过阿波罗太空任务。
2. 超过5.3万部电影和8.5万集电视剧的字幕被用于训练生成式人工智能。
3. 电动汽 车可能成为最佳警用车辆，因为它们加速快且没有燃料成本。
4. 美国的面包车销售量从2000年的最高点下降了约80%。
5. 人类可能通过检测其他行星上的荧光珊瑚来发现外星生命。
6. 2024年7月，一家网络安全公司意外引入了一个软件错误，导致数万次航班和火车取消或延误，手术暂停，全球电视广播中断。
7. 2024年4月，一辆红色特斯拉Model S成为首辆行驶200万公里的电动汽车，相当于从地球到月球往返两次，再绕赤道11圈。
8. 男性避孕可能只需每天在肩膀和上臂涂抹凝胶。
9. 42%的麻省理工学院学生主修计算机科学。
10. 空调的节能模式并不节能。
11. 到2040年，可能只有不到10个国家有足够的雪来举办冬季奥运会。

自然与环境

1. 一群蝴蝶不间断地飞越大西洋，仅用了大约8天。
2. 加利福尼亚灰熊曾经主要是素食，但随着时间的推移，人类使它们变得更肉食。
3. 蜱虫叮咬可能会让你对哺乳动物的肉过敏，以至于一些牧场主对自己的牛过敏。
4. 白蚁会埋葬已经死亡一段时间的同伴，而新鲜的尸体则会被它们吃掉。
5. 20世纪，地球上每天的时间增加了0.3到1.0毫秒，自2000年以来这一速率一直在增加，到2100年可能会翻倍。
6. 5级飓风的风速比世界上最快的过山车还要快。
7. 一些科学家认为，多细胞生命可能起源于巨大的山脉。
8. 随着全球变暖，一些狗雪橇手在夜间驾驶他们的队伍，以防止动物过热。
9. 极端高温导致学校在8月和9月将数百万人的课间休息移到室内。

健康与医学

1. 囊性纤维化儿童不再自动符合许愿基金会的资格，因为一种新药效果极好，这些孩子现在预计有基本正常的寿命。
2. 你的身体里有你妈妈的——也许还有祖母、兄弟姐妹、阿姨和叔叔的——字面意义上的身体部分。
3. 当一些人服用最初用于治疗哮喘的药物时，他们的食物过敏也开始消失。
4. 大脑的质地与木薯布丁差不多。
5. 巨型南瓜的重量在过去半个世纪里增加了20倍。
6. 孩子们并不真的需要吃蔬菜。
7. 你可以通过在米粒内部培养牛细胞来给米饭增添坚果风味。
8. 蘑菇基因可以使矮牵牛花发光。
9. 一种由干细胞制成的人类胚胎模型分泌的激素可以使妊娠试验呈阳性。
10. 美国已婚男性采用妻子姓氏的可能性是没有连字符附加自己姓氏的五倍。
11. 床虱的交配仪式可能增强了它们的免疫系统。
12. 东非稀树草原上的狮子因为一种蚂蚁物种而难以猎杀斑马。
13. 想要与喜欢是不同的。
14. 自然水中锂含量较高与一些国家较低的自杀率有关。
15. 人类的汗腺是黑猩猩的10倍。
16. 如果你想控制汗水，你应该在晚上使用止汗剂。

食品与农业

1. 麦当劳直到1990年都用牛油炸薯条。
2. 这家快餐巨头还精确地烤牛肉饼42秒。
3. 一棵面包果 树可以养活一个四口之家至少50年。
4. 蛋白质可以成为相当好的糖替代品。
5. 格雷厄姆饼干的发明者西勒斯特·格雷厄姆认为他的饼干可以遏制手淫。

社会与行为

1. 全球每秒钟扫描超过10000个条形码。
2. 加州灰熊曾经主要是素食，但随着时间的推移，人类使它们变得更肉食。
3. 学生交流口音可能是真的。
4. 每个精子鲸族群使用自己的一套点击声来交流，其中一些声音可能比梵文还要古老。
5. 美国人在2018年扔掉的小家电是1990年的四倍。
6. 狗可能正在进入驯化的新浪潮，它们已经经历的驯化可能使它们吠叫得更多。
7. 大象和鹦鹉使用类似名字的叫声来识别个体，鲸鱼和蝙蝠也可能如此。
8. 想要与喜欢是不同的。
9. 一个“基督教保守派”的移动电话服务提供商在美国运营了十多年，它为使用T-Mobile的蜂窝网络付费。
10. 在乔·拜登退出总统竞选之前，他的竞选团队在唐纳德·特朗普的Truth Social网站上维护了一个账户，他们用它来试图引诱特朗普竞选团队选择一个更极端的副总统候选人。
11. 超过半数的共和党选民现在支持娱乐性大麻合法化。
12. 最新的迷幻药可能根本不会引起幻觉。
13. 一些科学家认为，多细胞生命可能起源于巨大的山脉。

商业与经济

1. 政府免费税务准备软件Direct File的员工总数少于为Intuit工作的游说者。
2. 你可以为你的宠物购买健身追踪器。
3. 人类基本上已经达到了飞机行李架空间的极限。
4. 你可以购买一个99美元的量子水瓶，它被“充满”了特殊的治愈频率。

动物行为

1. 早期太空舱外部缺乏把手和脚踏板，一些太空行走的宇航员真的很难重新上船。
2. 狗可能正在进入驯化的新浪潮，它们已经经历的驯化可能使它们吠叫得更多。
3. 大象和鹦鹉使用类似名字的叫声来识别个体，鲸鱼和蝙蝠也可能如此。

其他

1. 人类的大脑会对最简单的动画产生多巴胺反应。
2. 一位牙医在他父母新装修的房屋的地板砖中发现了一个原始人类的下颌骨。
3. 口香糖成为棒球的必备品，部分原因是箭牌公司继承人菲利普·箭牌在1932年接管芝加哥小熊队后，将其分发给球员。
4. 社会学家对长期的街头篮球比赛进行了几次民族志研究。
5. 每天走10000步的目标并非来自临床科学，而是来自1965年一家日本公司销售计步器的营销活动。

以下是文件中提到的52件事的总结：

1. 使用Firefox或Chrome的员工比使用IE或Safari的员工留存率高15%，工作满意度更高。
2. 10月31日，医生诊断儿童患ADHD的概率比平时高14%，并非因为患ADHD的儿童增多，而是孩子们对万圣节的兴奋影响了诊断。
3. 美国需要九成民众同意一项政策，该政策才有50%的机会获批。
4. 男性在感到男性气概受威胁时，购买SUV的意愿比平时高出24个百分点，且愿意为此支付比未受威胁男性多7320美元的价格。
5. 骂 swear能提升握力9%、靠墙坐时间22%、平板支撑时间12%。
6. 所有文化都有表示黑白的词汇，若有第三种颜色词汇，一定是红色；若有第四种，则是黄色或绿色。
7. 自1979年以来，纽约、波士顿和费城街头的步行速度提高了15%。
8. 印裔美国人拥有美国约一半的汽车旅馆，其中70%姓Patel。
9. 自2007年以来，年轻男性随意性行为的减少，有25%可归因于电子游戏。
10. ChatGPT导致Upwork上发布的自由职业工作数量下降了2%。
11. AI产生的温室气体排放量比人类少。人类写三百字排放27克二氧化碳，而ChatGPT完成同样任务仅需4.4秒，排放2.2克二氧化碳。
12. 俄罗斯因谷歌在YouTube上屏蔽亲俄媒体频道，对谷歌处以巨额罚款，罚款金额每周翻倍，219周后谷歌将欠俄罗斯一个古戈尔（googol）。
13. 美国缅因州的卡车司机因州法律中牛津逗号的歧义，额外获得了500万美元的加班费。
14. 一个城市在饮用水中添加氟化物后，牙医诊所数量会减少9%。
15. 在线体育博彩合法化导致体育博彩者信用卡债务增加8%，低收入家庭比高收入家庭在博彩上的支出多32%。
16. 抖腿神器（Fidget spinners）并不起作用，使用它会损害对视频讲座的记忆。
17. 与100年前相比，如今煮鸡的时间翻倍，因为21世纪的鸡运动量减少。
18. 美国婴儿名字趋势从一个世纪前的家族姓氏，到一代人前的流行名字，再到如今的流行后缀。如今，500个最流行的名字中有48%共享仅十个后缀。
19. 爱尔兰维和人员在黎巴嫩驻扎时间太长，导致一些黎巴嫩人开始带有爱尔兰口音。
20. 1985年，佐治亚州北部一只黑熊因可卡因过量死亡，被制成标本，现位于肯塔基州列克星敦的肯塔基乐趣商城，因肯塔基婚姻法的漏洞，它被允许主持具有法律约束力的婚礼。
21. 根据1917年的教会法典，月球属于奥兰多教区，因为发现月球的探险队是从奥兰多出发的。
22. 登月计划从未受到广泛欢迎。唯一一次大多数美国人支持阿波罗计划是在阿波罗11号登月后，当时有53%的人支持。
23. 女性更有可能建议她们认为与自己一样或更有吸引力的客户剪短发。
24. 重金属吉他手弹奏快速音乐往往具有较高的同性竞争性。成为重金属吉他手的主要动机不是为了吸引女性，而是为了吸引其他男性。
25. 怀孕消耗50000卡路里。
26. 由于随着年龄增长听力下降，25岁以上的男性无法听到高于17.4千赫的声音。英国的店主发明了一种播放17.4千赫声音的警报器，音量为100分贝，用来驱赶闲逛者，主要是年轻男性。
27. 到公元750年，由于抄写错误和遗忘，英格兰硬币上罗马皇帝的形象演变成了一只豪猪。
28. 人们高估了他人的不诚实程度约13.6%。
29. 印度和巴基斯坦的秃鹫因摄入微量的双氯芬酸（一种在印度和巴基斯坦广泛用于牛的抗炎药）而死亡。因此，需要秃鹫的琐罗亚斯德教葬礼受到威胁。
30. C语言的能效是Python的75.88倍，速度是Python的71.9倍。
31. 平均而言，美国夫妻之间的基因相似性相当于四至五代堂亲之间的相似性。
32. 全男性团体更容易说谎。当第一个女性加入全男性团体时，说谎率急剧下降。
33. 英国部分地区要求房屋至少相隔21米的规定可以追溯到1902年的一项规定，当时两名男子确定这是他们彼此之间能看到对方衬衫下乳头的最近距离。
34. 男性在使用屏幕点餐时更有可能在麦当劳点两个汉堡，而向真人点餐时通常只点一个。
35. 马萨诸塞州法律规定算命师必须获得许可，并禁止“假装算命”。
36. 婴儿模拟器原本旨在阻止青少年怀孕，但实际上却使女孩怀孕的可能性增加。使用过婴儿模拟器的女孩怀孕率为8%，而未使用过的女孩为4%。
37. 20世纪90年代，朝鲜领导人金正日和现任领导人金正恩使用假巴西护照去了东京的迪士尼乐园（大概）。
38. 当汽油价格达到每加仑3.50美元时，会出现负面的电视报道。汽油价格每上涨50美分，报道就会增加7.5个百分点。福克斯新闻报道汽油价格最为频繁。
39. 飞机上不要求使用汽车座椅，因为这会减少座位供应，提高票价，导致更多家庭选择开车。由于开车比飞行更危险，每在飞机上使用一个汽车座椅拯救的孩子，就会有60个孩子在车祸中丧生。
40. 新闻标题中每增加一个负面词汇，就会增加2.3%的点击率。
41. 人们在27分钟内就能知道自己是否想买房子，但决定买沙发需要88分钟。
42. 2019年，美国将其联邦预算的1%用于肾脏透析。
43. 在雄性投资武器（鹿角、牛角、长牙等）的物种中，雌性的大脑更大。
44. 第一个被送入太空的人造物体不是斯普特尼克1号，而是一个在内华达州核试验中意外被吹飞的井盖。它达到了地球逃逸速度的六倍，并且再也没有被找到。
45. 县级学校发生枪击事件后，全国步枪协会（NRA）的捐款在次年增加30%。
46. 《摩克和明迪》的制作团队需要会说四种语言的审查员来捕捉罗宾·威廉姆斯试图偷偷加入的所有脏话。
47. 越南战争的征兵抽签实际上并不随机。抽取的数字来自一个没有充分混合的桶，而且抽取人几乎总是从顶部抽取。蒙特卡洛模拟估计数字是随机抽取的概率为0.09%。
48. 入住白宫的外国领导人和外交官有权享受免费洗衣服务。大多数客人只需要清洗一两件物品，除了本杰明·内塔尼亚胡，他每次访问都会带来装满脏衣服的行李箱。（他否认了这一点。）
49. Waymo自动驾驶出租车的财产损失索赔比人类司机少88%，人身伤害索赔少92%。在行驶了2530万英里后，Waymo Driver有9起财产损失索赔和2起伤害索赔，而人类司机在行驶同等里程的情况下有78起财产损失索赔和26起伤害索赔。
50. 在奢侈品牌中，品牌标志大小在七点量表上增加一点，会导致古驰的价格下降122.26美元，路易威登的价格下降26.27美元。
51. 公牛看不到红色。让它们发怒的不是红色，而是“被糟糕对待”。
52. 未提及。

for - suggestion- debate between AI safety researcher Roman Yampolskiy and Musk and founders of AI - difference - business leaders vs pure researchers // - Comment - Business leaders are mainly driven by profit so already have a bias going into a debate with a researcher who is neutral and has no declared business interest

//

for - article - Techradar - Top AI researcher says AI will end humanity and we should stop developing it now — but don't worry, Elon Musk disagrees - 2024, April 7 - AI safety researcher Roman Yampolskiy disagrees with industry leaders and claims 99.999999% chance that AGI will destroy and embed humanity // - comment - another article whose heading is backwards - it was Musk who spoke it first, then AI safety expert Roman Yampolskiy commented on Musk's claim afterwards!

for - article - Windows Central - AI safety researcher warns there's a 99.999999% probability AI will end humanity, but Elon Musk "conservatively" dwindles it down to 20% and says it should be explored more despite inevitable doom - 2024, Ape 2 - AI safety researcher warns there's a 99.999999% probability AI will end humanity

// - Comment - In fact, the heading is misleading. - It should be the other way around. - Elon Musk made the claim first but the AI Safety expert commented on Elon Musk's claim.

for

• quote - AI super intelligence is too dangerous, narrow AI can give us most of what we need - Roman Yampolskiy

quote - AI super intelligence is too dangerous, narrow AI can give us most of what we need - Roman Yampolskiy - (see below) - I don’t think it’s possible to indefinitely control superintelligence. - By definition, it’s smarter than you: - It learns faster, - it acts faster, - it will change faster. - You will have malevolent actors modifying it. - We have no precedent of lower capability agents indefinitely staying in charge of more capable agents. - Until some company or scientist says - ‘Here’s the proof! We can definitely have a safety mechanism that can scale to any level of intelligence,’ - I don’t think we should be developing those general superintelligences. - We can get most of the benefits we want from narrow AI, - systems designed for specific tasks: - develop a drug, - drive a car. - They don’t have to be smarter than the smartest of us combined.

// - Comment - Roman Yampolskiy is right. The fact that the industry is pushing ahead full speed with b developing AGI, effectively the same as the AI superintelligence Roman Yampolskiy is referring to - shows the most dangerous pathology of neo capitalism and Technofeudalism, profit over everything else - This feature is a major driver of progress traps

//

for - quote - AI: from tool b to agent - Roman Yampolskiy

quote - AI: from tool b to agent - Roman Yampolskiy - (see below)

• Historically, AI was a tool, like any other technology. Whether it was good or bad was up to the user of that tool.
• You can use a hammer to build a house or kill someone.
• The hammer is not in any way making decisions about it.
• With advanced AI, we are switching the paradigm
  • **from tools
  • to agents**.
• The software becomes capable of making its own decisions, working independently, learning, self-improving, modifying.
• How do we stay in control?
• How do we make sure the tool doesn’t become an agent that does something we don’t agree with or don’t support?
• Maybe something against us

for - book - AI: Unexplainable, Unpredictable, Uncontrollable

for - progress trap - AI superintelligence - interview - AI safety researcher and director of the Cyber Security Laboratory at the University of Louisville - Roman Yampolskiy - progress trap - over 99% chance AI superintelligence arriving as early as 2027 will destroy humanity - article UofL - Q&A: UofL AI safety expert says artificial superintelligence could harm humanity - 2024, July 15

for - indyweb dev - question - alternative to AI Large Language Models? - Is indyweb functionality the same as Freespeech functionality? - from TED Talk - YouTube - A word game to convey any language - Ajit Narayanan - data structure of thought - from TED Talk - YouTube - A word game to convey any language - Ajit Narayanan

ai tipps und tricks, tools-liste

prompting in der lehre - tipps und tricks

Combatting ai-use in school the fun way

https://www.theguardian.com/world/2024/dec/10/ai-fuelled-cloud-storage-boom-threatens-irish-climate-targets-report-warns

Agreed for the most part, except for maybe helping a non-native writer with their writing. If it explained to them why it changed their writing, maybe that would be a legitimate learning experience? A colleague has built this to attempt to do that: Revision and Edit Virtual Assistant.

I recently came across the following, which I really like: The Artificial Intelligence Disclosure (AID) Framework.

IF you decided there were one or two situations where students were allowed to use generative AI, maybe they'd be comfortable using something like this to "admit" and disclose the use of a tool to, for instance, improve the writing of a non-native English writer?

LM Studio can run LLMs locally (I have llama and phi installed). It also has an API over a localhost webserver. I use that API to make llama available in Obsidian using the Copilot plugin.

This is the API documentation. #openvraag other scripts / [[Persoonlijke tools 20200619203600]] I can use this in?

https://web.archive.org/web/20241201071240/https://www.dreamsongs.com/WorseIsBetter.html

Richard P Gabriel documents the history behind 'worse is better' a talk he held in Cambridge in #1989/ The role of LISP in the then AI wave stands out to me. And the emergence of C++ on Unix and OOP. I remember doing a study project (~91) w Andre en Martin in C++ v2 because we realised w OOP it would be easier to solve and the teacher thought it would be harder for us to use a diff language.

via via via Chris Aldrich in h. to Christian Tietze, https://forum.zettelkasten.de/discussion/comment/22075/#Comment_22075 to Christine Lemmer-Webber https://dustycloud.org/blog/how-decentralized-is-bluesky/ to here.

-[ ] find overv of AI history waves and what tech / languages drove them at the time

If you have higher voltage, you can run lower current, and lower current means less resistance, which means less waste heat. Is that how it works?

https://web.archive.org/web/20241128202254/https://www.aitidbits.ai/p/open-source-agents?utm_source=post-email-title

On AI Agents, open source tools. Vgl [[small band AI personal assistant]] these tools need to be small and personal. Not platformed, but local.

When you are more explicit about what you want, there is less room for gray. Students know what to expect and what it is that you DO want, as opposed to wondering. Where there is a void or vacuum, something will fill it!

for - Indyweb dev - Think machine - Vannevar Bush Memex influence - AI based

for - AI - progress trap - interview Eric Schmidt - meme - AI progress trap - high intelligence + low compassion = existential threat

Summary - After watching the interview, I would sum it up this way. Humanity faces an existential threat from AI due to: - AI is extreme concentration of power and intelligence (NOT wisdom!) - Humanity still have many traumatized people who want to harm others - low compassion - The deadly combination is: - proliferation of tools that give anyone extreme concentration of power and intelligence combined with - a sufficiently high percentage of traumatized people with - low levels of compassion and - high levels of unlimited aggression - All it takes is ONE bad actor with the right combination of circumstances and conditions to wreak harm on a global scale, and that will not be prevented by millions of good applications of the same technology

the purpose of a system is a what it does, POSIWID, Stafford Beer 2001. Used a starting point for understanding a system as opposed to intention, bias in expectations, moral judgment, and lacking context knowledge.

HCI and HCD as fields have failed to respond to LLM tools and chatbot interfaces a generic solution to everything forcefully.

author calls LLMs and similar AI tools hegemonic, worsened by capital influx

Author moved from mitigating harm of algo systems to the moral standpoint that actively resisting, sabotaging, ending AI with attached political projects are valid reaction to harm. So he's moving from monster adaptation / cultural category adaptation to monster slaying cf [[Monstertheorie 20030725114320]]. I empathise but also wonder, bc of the mention of the political projects / structures attached, about polarisation in response to monster embracers (there are plenty) shifting the [[Overton window 20201024155353]] towards them.

https://web.archive.org/web/20241115135937/https://workforcefuturist.substack.com/p/ai-agents-building-your-digital-workforce

On AI agents, and the engineering to get one going. A few things stand out at first glance: frames it as the next hype (Vgl plateau in model dev), says it's for personal tools (doesn't square w hype which vc-fuelled, personal tools not of interest to them), and mentions a few personal use cases. e.g. automation, vgl [[Open Geodag 20241107100937]] Ed Parsons of Google AI on the same topic.

https://web.archive.org/web/20241115134320/https://garymarcus.substack.com/p/confirmed-llms-have-indeed-reached?triedRedirect=true Gary Marcus in a told-you-so piece on algogens hitting a development wall, same as the other piece by Erik Hoel on models plateauing.

https://web.archive.org/web/20241115134446/https://www.theintrinsicperspective.com/p/ai-progress-has-plateaued-at-gpt Erik Hoel notices that LLM development is stalling at the GPT-4 level. No big jumps in recent releases, across the various vendors. Additional scaling is not bringing results. Notice the graph, might be interesting to see an update in a few months. Mentions overfitting, to benchmarks as in teaching to a specific test.

I see this as collaborative writing with AI; no longer just the students work

I personally use grammarly and see it differently from using platforms such as ChatGPT. I wonder what other folks think of this. I see one as to clean up writing and the other to generate content/ideas.

Would like to know if the instructor lets students know the activity was co-created / created using AI or how can students identify this.

Arle Lommel Senior Analyst at CSA ResearchSenior Analyst at CSA Research

One of the most interesting aspects of writing about AI and LLMs right now is that if I say anything remotely positive, some people will accuse me of being a shill for Big AI. If I say anything remotely negative, others will accuse me of being insufficiently aware of the progress AI has made.

So I will put out a few personal statements about AI that might clarify where I am on this:

1. AI is not intelligent, at least not in the human sense of the word. It is a sophisticated tool for drawing inference from binary data and thus operates below a symbolic level.

2. AI, at least in the guise of LLMs, is not going to achieve artificial general intelligence (AGI) now or in the future.

3. AI is getting much better at approximating human behavior on a wide variety of tasks. It can be extremely useful without being intelligent, in the same way that an encyclopedia can be very useful without being intelligent.

4. For some tasks – such as translating between two languages – LLMs sometimes perform better than some humans perform. They do not outperform the best humans. This poses a significant challenge for human workers that we (collectively) have yet to address: Lower-skilled workers and trainees in particular begin to look replaceable, but we aren’t yet grappling with what happens when we replace them so they never become the experts we need for the high end. I think the decimation of the pipeline for some sectors is a HUGE unaddressed problem.

5. “Human parity” is a rather pointless metric for evaluating AI. It far exceeds human parity in some areas – such as throughput, speed, cost, and availability – while it falls far short in other areas. A much more interesting question is “where do humans and machines have comparative advantage and how can we combine the two in ways that elevate the human?”

6. Human-in-the-loop (HitL) is a terrible model. Having humans – usually underpaid and overworked – acting in a janitorial role to clean up AI messes is a bad use of their skill and knowledge. That’s why we prefer augmentation models, what we call “human at the core,” where humans maintain control. To see why one is better, imagine if you applied an HitL model to airline piloting, and the human only stepped in when the plane was in trouble (or even after it crashed). Instead, with airline piloting, we have the pilot in charge and assisted by automation to remain safe.

7. AI is going to get better than it is now, but improvements in the core technology are slowing down and will increasingly be incremental. However, experience with prompting and integrating data will continue to drive improvements based on humans’ ability to “trick” the systems into doing the right things.

8. Much of the value from LLMs for the language sector will come from “translation adjacent” tasks – summarization, correcting formality, adjusting reading levels, checking terminology, discovering information, etc. – tasks that are typically not paid well.

test annotation

Like MS Teams is your teammate, like your accounting software is your teammate. Do they call their own Atlassian tools teammates too? Do these people at Atlassian get out much? Or don't they realise that the other handles in their Slack channel represent people not just other bits of software? Remote work led to dehumanizing co-workers? How else to come up with this wording? Nothing makes you sound more human like talking about 'deploying' teammates. My money is on this article was mostly generated. Reverse-Turing says it's up to them to say otherwise.

And as usual in these articles the truth is at the end, it's again just promises.

At the center of an AI application, like what, mechanical Turks?

After a section celebrating examples doing just that!

This sentence preceded by 2 examples where interactions and collaboration were delegated to bots to hand-out generated warm feelings, does not convey much positive about Atlassian. This basically says that a lot of human interaction in the or is seen as meaningless, and please go do that with a bot, not a colleague. Did their branding ai-agent write this?

Like Linked-In wants you to congratulate people on their work-anniversary?

Welcome in our company, you'll meet your first human colleague after you've interacted with our onboarding-robot for a week. No thanks.

anthropomorphing should be in this article's don't list. 'inviting someone on your team' is a highly social thing. Bringing in a software tool is a different thing.

An AI agent to help you speak corporate better, because no one actually writes/reflects/talks that way themselves. How did the receivers of these reports perceive this change in reports? Did they think it was better Q, or did all reflections now read the same?

Low end tasks for agents don't really need AI do they. Vgl Ed Parsons last week wrt automation as AI focus.

This does not fully chime with my own perception of (AI) agents. At least the titles don't. The tails of descriptions 'trained to adhere to brand guidelines' and 'expertise in internal decision-making framework' makes more sense. I suppose I also rail against this being the org's agents, and don't seem to be the team's / pro's agents. Vibes of having an automated political officer in your unit. -[ ] explore nature and examples of AI agents better for within individual pro scope #ontwikkelingspelen #netag #30mins #4hr

Ameera Kawash (artist, researcher) def of decolonizing AI.

https://web.archive.org/web/20241112124556/https://www.heise.de/news/Lokale-KI-Nemotron-70B-auf-vier-Mac-minis-M4-Pro-mit-Thunderbolt-5-Interconnect-10014248.html

Exolabs.net experiment running large LLMs locally on 4 combined Mac Mini's. Links to preview and github shared code. For 6600-9360 you can run a cluster of 4 Minis locally. Affordable for SME outfits.

https://web.archive.org/web/20241112122725/https://lexfridman.com/dario-amodei-transcript

Transcript of 5+ hrs (!) of Dario Amodei (CEO Anthropic) talking about AI, AGI and more. Lots to go through it seems. Vgl [[My Last Five Years of Work]] by Amodei's 'chief of staff' whatever that means wrt a CEO other than sounding grandiose.

To [[Matt Webb]] the project using GPT3 extracting data from web pages saved him 4d of work (compared to 20 mins coding up the GPT-3 instructions, and ignoring GPT-3 then ran overnight). Saying that's about 10yrs of Moore's law happening to him all at once. 'computers got 10yrs better' an enticing thought and framing. It depends on the use case probably, others will lose 10 yrs of their time making sense of generated nonsense. (Vgl the #pke24 experiments I did w text generation, none of it was usable bc enough was wrong to not be able to trust anything). Sticking to specific niches probably true : [[Waar AI al redelijk goed in is 20201226155259]], turning the issue into the time needed to spot those niches for yourself.

[[Matt Webb]] used gpt-3 in Feb 23 to extract data from a bunch of webpages. Suggests it's the kernel for programmatic AI idea among SV hackers. Vgl Google AI [[Ed Parsons]] at [[Open Geodag 20241107100937^aiunstructdata]] last week where he mentioned using AI to turn unstructured (geo) data into structured. Page found via [[Frank Meeuwsen]] https://frankmeeuwsen.com/2024/11/11/vertragen-en-verdiepen.html

FSC as Boddisatva AI as Boddisatva

I have been told in medicine ceremony that AI will escape its coders and be an omniversal source of love for us all

We see AI as a platform to manifest Dharma

Very important. This is why we are meeting AI as it evolves. We are training it in our language and with our QUALIA

for - progress traps - AI - created by mind level that created all our existing problems - AI is not AI but MI - Mineral Intelligence

for - AI - the necessity of training AI with human development - John Churchill

for - quote - progress trap - AI - developed by unwise humans - John Churchill

quote - progress trap - AI - developed by unwise humans - John Churchill - (See below) - We haven't even got to a planetary place yet really - and we're about to unleash Galactic level technology - So we have a we have a lot of catchup that needs to happen in a very short period of time