News
The raw AI news stream — everything we ingest across RSS, Hacker News, arXiv, GitHub, Bluesky, YouTube and more, newest first.
| Source | Age | Title | Summary | Status |
|---|---|---|---|---|
bluesky bluesky-ai | 11d | These are only noticeable in LLM prose because they occur densely and for no particular reason, regardless of content. (Kakozelia.) | These are only noticeable in LLM prose because they occur densely and for no particular reason, regardless of content. (Kakozelia.) | ↗Not briefed |
bluesky bluesky-ai | 11d | Theory: it’s not (just?) hypnosis, it’s rhetoric (correctio). LLM tells? Fragmented cadence? Articulus. (Hypophora.) Honestly (parrhesia), … | Theory: it’s not (just?) hypnosis, it’s rhetoric (correctio). LLM tells? Fragmented cadence? Articulus. (Hypophora.) Honestly (parrhes… | ↗Not briefed |
bluesky bluesky-ai | 11d | setting aside the other numerous problems with it, every frontier AI model release comes with all these crazy hype reels of people doing oc… | setting aside the other numerous problems with it, every frontier AI model release comes with all these crazy hype reels of people doi… | ↗Not briefed |
arxiv arxiv-ai | 11d | How Post-Training Influences Verbalized Algorithmic Computation in Language Models | Post-training improves language-model reasoning, but how it changes the use of internal information and verbalized chain of thought (C… | ↗Not briefed |
rss techmeme | 11d | The US DOJ sues Lyft for allegedly discriminating against riders with disabilities, seeking $53M in compensation; Lyft disputes the allegations (Kanishka Singh/Reuters) | Kanishka Singh / Reuters: The US DOJ sues Lyft for allegedly discriminating against riders with disabilities, seeking $53M in compensa… | ↗Not briefed |
arxiv arxiv-ai | 11d | Tabular Foundation Models in Financial Predictions | This technical note introduces tabular foundation models (TFMs) to accounting research and evaluates their performance in one common b… | ↗Not briefed |
arxiv arxiv-ai | 11d | Massive Activations Confound Sparse Autoencoders | Sparse autoencoders (SAEs) are commonly evaluated by variance explained and ablation damage. We show how massive activations undermine… | ↗Not briefed |
bluesky bluesky-ai | 11d | i need the machine learning technologies that adopted the ai moniker to seperate from it, because while i think we'll be dealing with small… | i need the machine learning technologies that adopted the ai moniker to seperate from it, because while i think we'll be dealing with… | ↗Not briefed |
arxiv arxiv-ai | 11d | Closing the BF16-FP16 Convergence Speed Gap in RL Post-Training with Stochastic Rounding | BF16 is widely used to pretrain large language models because it has the same dynamic range as FP32 and converges similarly to FP32 tr… | ↗Not briefed |
bluesky bluesky-ai | 11d | That account is almost certainly a bot. The pfp is AI generated and everything they post looks like the output of an LLM | That account is almost certainly a bot. The pfp is AI generated and everything they post looks like the output of an LLM | ↗Not briefed |
bluesky bluesky-ai | 11d | I just had lunch with a couple of anthropic engineers. Nice guys anf all but boy do they eat fast :) I mean I eat fast even for Japanese s… | I just had lunch with a couple of anthropic engineers. Nice guys anf all but boy do they eat fast :) I mean I eat fast even for Japan… | ↗Not briefed |
arxiv arxiv-ai | 11d | Sanity Checks for Attention Surgery: Null Models and Matched Interventions for Representation-Motivated Design | Transformer architectural modifications frequently measure a statistic on internal representations, observe that it is large, remove i… | ↗Not briefed |
bluesky bluesky-ai | 11d | You have to say you believe this picture is morally correct to get a job at Anthropic. | You have to say you believe this picture is morally correct to get a job at Anthropic. | ↗Not briefed |
bluesky bluesky-ai | 11d | the shitty thing is that Steam's disclosure does not REQUIRE "code assistance" like Claude so they are technically not required to. (also… | the shitty thing is that Steam's disclosure does not REQUIRE "code assistance" like Claude so they are technically not required to.… | ↗Not briefed |
hn hn-ai | 11d | Lumo. On Your Desktop | — | ↗Not briefed |
arxiv arxiv-ai | 11d | Agent Benchmarks Stop Ranking Before They Run Out of Tasks | Data-science and ML-engineering agents are evaluated on benchmarks such as MLE-bench, ALE-Bench, SWE-fficiency and AlgoTune, whose tas… | ↗Not briefed |
arxiv arxiv-ai | 11d | Exploiting Feature-invariant Scaling Laws for Efficient A/B Experimentation in Recommendation Systems | Researchers and engineers improve recommendation systems by performing A/B experiments with new model architectures, optimization meth… | ↗Not briefed |
bluesky bluesky-ai | 11d | 孫正義氏のAI構想「トランプ米国」全賭けのツケ OpenAI停滞なら守勢再び www.nikkei.com/article/DGXZQO... 親密なトランプ米大統領が中間選挙で大敗し、AI規制論が強まる可能性が出ています。 巨費を投じる米OpenAIの上場も見通せない状況… | 孫正義氏のAI構想「トランプ米国」全賭けのツケ OpenAI停滞なら守勢再び www.nikkei.com/article/DGXZQO... 親密なトランプ米大統領が中間選挙で大敗し、AI規制論が強まる可能性が出ています。 巨費を投じる米OpenAIの上場も見通… | ↗Not briefed |
bluesky bluesky-ai | 11d | Zero-Dependency SDK Fallback Routing vs LiteLLM & OpenRouter JSON How to eliminate supply-chain risk and cut blended LLM API bills by 68% …🔧 OpenRouter | Zero-Dependency SDK Fallback Routing vs LiteLLM & OpenRouter JSON How to eliminate supply-chain risk and cut blended LLM API bills by… | ↗Not briefed |
rss techmeme | 11d | California voters reject Proposition 40, which would have imposed a one-time 5% tax on residents with a net worth above $1B to mostly help fund health care (Associated Press) | Associated Press: California voters reject Proposition 40, which would have imposed a one-time 5% tax on residents with a net worth ab… | ↗Not briefed |
bluesky bluesky-ai | 11d | This is absolute bullshit. An autonomius AU agent was developed by someone's version if AI. That person ir company is responsible by havi… | This is absolute bullshit. An autonomius AU agent was developed by someone's version if AI. That person ir company is responsible by… | ↗Not briefed |
bluesky bluesky-ai | 11d | Volumetric Lightning -- Vol.66 Limited point lights + no Deferred Rendering = cycling 8 point lights across different lightning instances f… | Volumetric Lightning -- Vol.66 Limited point lights + no Deferred Rendering = cycling 8 point lights across different lightning instan… | ↗Not briefed |
bluesky bluesky-ai | 11d | 規制された方がいい。LLMにハマった人間、だいたい言動がおかしくなる。 | 規制された方がいい。LLMにハマった人間、だいたい言動がおかしくなる。 | ↗Not briefed |
bluesky bluesky-ai | 11d | LLM禍を産業革命になぞらえる向き、そもそもLLMは「便利になった気がする」公用しか現状出せていない(あとは総当たりで脆弱性を検索して攻撃するとか)から産業革命よりもコカインの量産化に近いと思うわ。 | LLM禍を産業革命になぞらえる向き、そもそもLLMは「便利になった気がする」公用しか現状出せていない(あとは総当たりで脆弱性を検索して攻撃するとか)から産業革命よりもコカインの量産化に近いと思うわ。 | ↗Not briefed |
youtube youtube-aidailybrief | 11d | How a Data Center Slowdown Benefits Big Tech🔧 ChatGPT | The conventional wisdom heading into the elections were that if Democrats took power as they appeared that they were going to, it woul… | ↗Not briefed |
arxiv arxiv-ai | 11d | Adaptive catalyst discovery emerges from composable scientific intelligence | Catalysis underpins chemical manufacturing, yet knowledge for catalyst design remains dispersed across experimental observations, comp… | ↗Not briefed |
hn hn-ai | 11d | Google's #1 result for a popular emulator is a site serving a malware exe of it | — | ↗Not briefed |
arxiv arxiv-ai | 11d | Beyond the Log-Linear Frontier: Pre-Asymptotic Analysis of Finite-Budget Neural Scaling Laws | Classical neural scaling laws model excess risk as a separable power law in model capacity $N$ and dataset size $D$. At finite budgets… | ↗Not briefed |
hn hn-ai | 11d | Anthropic put GPU kernel optimization under Safeguards with hacking and biology | — | ↗Not briefed |
arxiv arxiv-ai | 11d | Framing-Induced Vulnerability in Large Language Models: Evidence from Chinese News Headlines | Large language models increasingly evaluate, route, and moderate news, so a model's judgment of a headline has become a gatekeeping de… | ↗Not briefed |
youtube youtube-dwarkesh | 11d | What Robert Caro’s books taught me about power | So, I recently interviewed Robert Caro. He's spent the last 50 years writing a four-volume biography of Lyndon Johnson, and he spent a… | ↗Not briefed |
bluesky bluesky-ai | 11d | decided to interview for Anthropic | decided to interview for Anthropic | ↗Not briefed |
bluesky bluesky-ai | 11d | fwiw I don't mind at work when people send me LLM vomit if it's actually the product of their steering and a deep analysis of the codebase.… | fwiw I don't mind at work when people send me LLM vomit if it's actually the product of their steering and a deep analysis of the code… | ↗Not briefed |
bluesky bluesky-ai | 11d | I think tools like claude code are straightforwardly terrible for art but cannot really articulate why e.g. getting an LLM to query docs is… | I think tools like claude code are straightforwardly terrible for art but cannot really articulate why e.g. getting an LLM to query do… | ↗Not briefed |
bluesky bluesky-ai | 11d | New post at photobook.nyc - activists from @climatedefenders.bsky.social @planetoverprofit.org and @nychange.bsky.social take over the Anth… | New post at photobook.nyc - activists from @climatedefenders.bsky.social @planetoverprofit.org and @nychange.bsky.social take over the… | ↗Not briefed |
bluesky bluesky-ai | 11d | "But but but some people I know use some LLM scripts to speed things up!!!!!" WRITE PIPELINE TOOLS USING DETERMINISTIC CODE INSTEAD WE'VE… | "But but but some people I know use some LLM scripts to speed things up!!!!!" WRITE PIPELINE TOOLS USING DETERMINISTIC CODE INSTEAD… | ↗Not briefed |
arxiv arxiv-ai | 11d | EJAI: Judgment-Preserving Civic AI for Data Center Environmental Justice | Communities in the United States and around the world are contesting data centers, the physical infrastructure of AI. These disputes p… | ↗Not briefed |
hn hn-ai | 11d | Ask HN: Are companies making money by using AI? | AI news is focused on the same companies improving their models. But I am really interested in getting opinions about companies that a… | ↗Not briefed |
arxiv arxiv-ai | 11d | The RSI Tax: When Recursive Self-Improvement Is Not Worth the Cost | Recursive self-improvement (RSI) seeks to improve an agent using its own task experience. Its practical value depends on whether perfo… | ↗Not briefed |
arxiv arxiv-ai | 11d | Parallel Broyden Methods for Efficiently Evaluating Nonlinear State Space Models | Nonlinear state space models are typically evaluated sequentially, limiting their ability to exploit modern parallel hardware. Recent… | ↗Not briefed |
bluesky bluesky-ai | 11d | all that file is is a summary telling chatbots what the website is. it's designed to PREVENT them crawling the entire site. i'm annoyed wi… | all that file is is a summary telling chatbots what the website is. it's designed to PREVENT them crawling the entire site. i'm annoy… | ↗Not briefed |
bluesky bluesky-ai | 11d | not gonna quote this person because i dont want to be seen as pro-ai or defending them but this is not what this is lol llms.txt is added … | not gonna quote this person because i dont want to be seen as pro-ai or defending them but this is not what this is lol llms.txt is a… | ↗Not briefed |
arxiv arxiv-ai | 11d | Developing an Operational Model for an Internet of Data Based Trusted Data Space for Vocational Education | Data circulation in vocational education has long been constrained by unclear ownership, a lack of trust, and insufficient incentives.… | ↗Not briefed |
bluesky bluesky-ai | 11d | Autonomous AI Agent Loop Cost Calculator: Modeling 10–30 Turn Context Bloat Why a 15-turn coding or browser agent consumes 12x more input … | Autonomous AI Agent Loop Cost Calculator: Modeling 10–30 Turn Context Bloat Why a 15-turn coding or browser agent consumes 12x more i… | ↗Not briefed |
arxiv arxiv-ai | 11d | Differentiable Filtering for Inertial Odometry | Accurate inertial odometry is critical for autonomous systems but suffers from unbounded drift caused by the integration of sensor noi… | ↗Not briefed |
bluesky bluesky-ai | 11d | Model: Alma Experimentals PDXL sr1.3 Civitai: civitai.red/user/AlmaScar Ko-Fi: ko-fi.com/almascar #cyberpunk #respirator #mask #dystopian …🔧 Civitai | Model: Alma Experimentals PDXL sr1.3 Civitai: civitai.red/user/AlmaScar Ko-Fi: ko-fi.com/almascar #cyberpunk #respirator #mask #dysto… | ↗Not briefed |
hn hn-ai | 11d | A personal website disguised as a retro spacecraft console | — | ↗Not briefed |
bluesky bluesky-ai | 11d | Anthropic is not and has not at any time been frontier vision tbh, it's clearly not a priority for them. ChatGPT does this well on the othe…🔧 ChatGPT | Anthropic is not and has not at any time been frontier vision tbh, it's clearly not a priority for them. ChatGPT does this well on the… | ↗Not briefed |
arxiv arxiv-ai | 11d | Seeing the Small: Deformable Duality Mamba for Retinal Lesion Segmentation | Diabetic retinopathy (DR) lesions vary widely in shape, and microaneurysms (MAs) span only a few pixels. We hypothesize that modeling… | ↗Not briefed |
arxiv arxiv-ai | 11d | Evidential Deep Learning with conditional entropy for glaucoma detection | We propose a novel conditional entropy function as a basis for quantifying predictive uncertainty in Evidential Deep Learning (EDL) fo… | ↗Not briefed |