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 | 17d | Yes yes, first comes the collectivisation, then when that fails you try to walk it all back with a Perestroika, just before the whole thing… | Yes yes, first comes the collectivisation, then when that fails you try to walk it all back with a Perestroika, just before the whole… | ↗Not briefed |
hn hn-ai | 17d | Ask HN: Credit card and loyalty analytics, where to start? | I inherited responsibilities for creating credit card and loyalty analytics reporting/dashboarding for a firm that processes roug… | ↗Not briefed |
bluesky bluesky-ai | 17d | “We've developed an autonomous vulnerability discovery, exploitation, and patching pipeline. In one run with our cyber model, it validated … | “We've developed an autonomous vulnerability discovery, exploitation, and patching pipeline. In one run with our cyber model, it valid… | ↗Not briefed |
bluesky bluesky-ai | 17d | I suspect they’ll continue to do so as long as they’re using things like social media, 4-chan, LLM products, (some) video games, etc, & mai… | I suspect they’ll continue to do so as long as they’re using things like social media, 4-chan, LLM products, (some) video games, etc,… | ↗Not briefed |
arxiv arxiv-ai | 17d | Adversarial Multiplicity: Cherry-Picked Models and the Limits of Disclosure | Predictive multiplicity -- where similarly accurate models disagree on individual predictions -- can be leveraged to optimize secondar… | ↗Not briefed |
bluesky bluesky-ai | 17d | Matt Scherer of the Open Markets Institute argues the SEC should block Anthropic’s IPO until the company provides “two consecutive quarters… | Matt Scherer of the Open Markets Institute argues the SEC should block Anthropic’s IPO until the company provides “two consecutive qua… | ↗Not briefed |
hn hn-ai | 17d | OpenAI spent $1.2M in API cost to fix 948 Linux vulnerabilities | — | ↗Not briefed |
bluesky bluesky-ai | 17d | classic vox where the headline is just irrelevant ragebait not related to the article, really. its just talking about specialized neural ne… | classic vox where the headline is just irrelevant ragebait not related to the article, really. its just talking about specialized neur… | ↗Not briefed |
bluesky bluesky-ai | 17d | all the other companies have many faces and do many things (meta has 1.7k github repos...) but Anthropic is entirely an LLM seller. So imo … | all the other companies have many faces and do many things (meta has 1.7k github repos...) but Anthropic is entirely an LLM seller. So… | ↗Not briefed |
rss techmeme | 17d | Berlin-based Pliant raised an $80M Series C led by Seaya Nova and Speedinvest, as it opens its B2B payments infrastructure to AI agents (Cristian Dina/The Next Web)🔧 ChatGPT | Cristian Dina / The Next Web: Berlin-based Pliant raised an $80M Series C led by Seaya Nova and Speedinvest, as it opens its B2B payme… | ↗Not briefed |
arxiv arxiv-ai | 17d | When Scores Conflict with Preferences: Pre-Alignment Diagnosis and Adaptive Correction of Constraint Drift in DPO | Direct Preference Optimization (DPO) is increasingly used in settings where pairwise preferences coexist with pointwise criteria such… | ↗Not briefed |
arxiv arxiv-ai | 17d | EnvOpt: Automated Optimization for Agent Environments | An LLM agent's cost and success depend on its model, its harness, and its environment. In addition to the widely studied directions of… | ↗Not briefed |
bluesky bluesky-ai | 17d | Buying dirty Colossus compute (that also hugely supports SpaceX) knocked them off the “one good AI company” pedestal for me, probably perma… | Buying dirty Colossus compute (that also hugely supports SpaceX) knocked them off the “one good AI company” pedestal for me, probably… | ↗Not briefed |
bluesky bluesky-ai | 17d | Very emotional night, with speeches from families and folks directly impacted by this anti-trans proposition. Will upload speeches when my … | Very emotional night, with speeches from families and folks directly impacted by this anti-trans proposition. Will upload speeches whe… | ↗Not briefed |
arxiv arxiv-ai | 17d | Test-Time Self-Play: Scaling without a Verifier | Test-time scaling methods spend additional compute at deployment to improve a large language model's performance on a hard target prob… | ↗Not briefed |
bluesky bluesky-ai | 17d | Mistral Small 3.1 24B Instruct vs Gemini 2.0 Flash: Which LLM API actually protects your SaaS gross margin at 10,000 DAU? Here is the live… | Mistral Small 3.1 24B Instruct vs Gemini 2.0 Flash: Which LLM API actually protects your SaaS gross margin at 10,000 DAU? Here is the… | ↗Not briefed |
bluesky bluesky-ai | 17d | Meet gozinaki, our on-device LLM inference module powered by llama.cpp! 🤖💬 Drop it into your game loop to generate dynamic NPC dialogues … | Meet gozinaki, our on-device LLM inference module powered by llama.cpp! 🤖💬 Drop it into your game loop to generate dynamic NPC dialo… | ↗Not briefed |
hn hn-ai | 17d | Transcribe Video to Text – Transcribe Video to Text Online Free | — | ↗Not briefed |
arxiv arxiv-ai | 17d | Asymptotic-Preserving Tensor Neural Networks for 3D High-Field Kinetic Equations with Applications to Uncertainty Quantification🔧 FLUX | Kinetic equations in the high-field regime describe particle transport dominated by electric-field acceleration and collisional relaxa… | ↗Not briefed |
hn hn-ai | 17d | Get answers from what you save | — | ↗Not briefed |
arxiv arxiv-ai | 17d | Acoustic Anomaly Detection for Early Identification of Drill Jamming | In the context of Industry 4.0, predictive maintenance is essential to minimize unplanned downtime and ensure operational safety. In m… | ↗Not briefed |
rss techmeme | 17d | OpenAI details a pipeline that used an upcoming cyber model to validate and patch 948 local privilege escalation vulnerabilities in Linux for $1.2M in API costs (OpenAI) | OpenAI: OpenAI details a pipeline that used an upcoming cyber model to validate and patch 948 local privilege escalation vulnerabiliti… | ↗Not briefed |
bluesky bluesky-ai | 17d | It’s essay grading time. I wish students wouldn’t make me read LLM generated bullshit. I much prefer papers with weird grammar and misspell… | It’s essay grading time. I wish students wouldn’t make me read LLM generated bullshit. I much prefer papers with weird grammar and mis… | ↗Not briefed |
bluesky bluesky-ai | 17d | PHP2SHELLの問題はLLMの問題ではなく、PHPのようなインターネットの基礎インフラを世界中の企業がオープンソースのボランティアであるPHPFoundationにただ乗りしていたという、テック大手だけでなくインターネットに関わるすべての企業の「オープンソースの善意にただ乗… | PHP2SHELLの問題はLLMの問題ではなく、PHPのようなインターネットの基礎インフラを世界中の企業がオープンソースのボランティアであるPHPFoundationにただ乗りしていたという、テック大手だけでなくインターネットに関わるすべての企業の「オープンソースの善… | ↗Not briefed |
bluesky bluesky-ai | 17d | Back to Rob Sand. I just learned that he was a model in the early 2000s on the runaway in Paris and Milan. And his opponents made AI photos… | Back to Rob Sand. I just learned that he was a model in the early 2000s on the runaway in Paris and Milan. And his opponents made AI p… | ↗Not briefed |
hn hn-ai | 17d | Show HN: Edgepad – Turn an Android phone into a trackpad for your laptop | — | ↗Not briefed |
arxiv arxiv-ai | 17d | From What Is Used to What Is Missing: Rethinking Dynamic Rank in LoRA | Parameter-efficient fine-tuning with LoRA is governed by rank, which controls parameter budget and locally accessible task updates. Ex… | ↗Not briefed |
arxiv arxiv-ai | 17d | AtlasULP: Domain-aware Universal Link Prediction via Relation Atlas | Link prediction (LP) is a widely applied task in graph learning. Conventional LP approaches typically follow a dataset-specific paradi… | ↗Not briefed |
hn hn-ai | 17d | Asteroid defense hits a lump in the road | — | ↗Not briefed |
bluesky bluesky-ai | 17d | used llm 4 interview prep for a biggun and yeah now I totally get the overwhelming ‘you’re the best’ it gives you… ngl it’s good for impo… | used llm 4 interview prep for a biggun and yeah now I totally get the overwhelming ‘you’re the best’ it gives you… ngl it’s good for… | ↗Not briefed |
arxiv arxiv-ai | 17d | Selective Retrieval Routing for Multi-Violation Knowledge Graph Repair | Knowledge graphs often contain missing or inconsistent facts that undermine the applications built on them. Schema validation detects… | ↗Not briefed |
hn hn-ai | 17d | How to Tailor Your Resume for Every Job Application | — | ↗Not briefed |
hn hn-ai | 17d | Kimi K3.1 Pro writes fake OpenAI ciphertext | — | ↗Not briefed |
bluesky bluesky-ai | 17d | One final remark on all of this. I perceive that people are a bit spooked that the LLM can write neural network weights by hand as opposed … | One final remark on all of this. I perceive that people are a bit spooked that the LLM can write neural network weights by hand as opp… | ↗Not briefed |
bluesky bluesky-ai | 17d | OpenAI, ready to IPO in 2027, now counts majority of sales from enterprises, Brockman says reut.rs/4iFfSW8 | OpenAI, ready to IPO in 2027, now counts majority of sales from enterprises, Brockman says reut.rs/4iFfSW8 | ↗Not briefed |
bluesky bluesky-ai | 17d | OpenAI is basically Digger but for restaurant food pictures | OpenAI is basically Digger but for restaurant food pictures | ↗Not briefed |
arxiv arxiv-ai | 17d | Last-Iterate Convergence of Stochastic Gradient Methods for Unbounded Problems | Last-iterate convergence guarantees for stochastic gradient methods have attracted significant attention as practical implementations… | ↗Not briefed |
arxiv arxiv-ai | 17d | Permanent and Transient Representations for Continual Reinforcement Learning | Continual Reinforcement Learning (CRL) agents struggle to adapt to new situations while retaining prior knowledge, leading to a stabil… | ↗Not briefed |
hn hn-ai | 17d | I built an alternative to Conductor for Claude and Codex | — | ↗Not briefed |
rss techmeme | 17d | AppLovin reports Q3 revenue up 55% YoY to $2.18B, above $2.07B est., and forecasts Q4 revenue above estimates; APP surges 20%+ after hours (William Gavin/MarketWatch) | William Gavin / MarketWatch: AppLovin reports Q3 revenue up 55% YoY to $2.18B, above $2.07B est., and forecasts Q4 revenue above estim… | ↗Not briefed |
bluesky bluesky-ai | 17d | What is he talking about? | What is he talking about? | ↗Not briefed |
arxiv arxiv-ai | 17d | Rethinking Data Selection: The Importance of Coverage over Difficulty in LLM Fine-Tuning | Selecting high-quality training data can reduce computation cost for LLM fine-tuning. Prior work on data selection developed a variety… | ↗Not briefed |
arxiv arxiv-ai | 17d | WeightLoop: Weight-Space Programming with Experimental Feedback | Weight-space programming (WSP) shows that language models can reason about and rewrite neural network weights; recursive self-improvem… | ↗Not briefed |
bluesky bluesky-ai | 17d | OpenAI is discovering that there's money to be made by playing both sides of the cybersecurity game. - Sell AI tools that can be used to h… | OpenAI is discovering that there's money to be made by playing both sides of the cybersecurity game. - Sell AI tools that can be used… | ↗Not briefed |
bluesky bluesky-ai | 17d | The comment section is gold github.com/openai/codex... | The comment section is gold github.com/openai/codex... | ↗Not briefed |
rss openai-news | 17d | Setting a new standard for responsible AI infrastructure | OpenAI introduces its National Community Compact and Compute Campus Policy Blueprint to protect local resources, create jobs, and guid… | ↗Not briefed |
bluesky bluesky-ai | 17d | AHAHHAHAHAA github.com/openai/codex... | AHAHHAHAHAA github.com/openai/codex... | ↗Not briefed |
bluesky bluesky-ai | 17d | wasn't that how anthropic always was tho they were always weird | wasn't that how anthropic always was tho they were always weird | ↗Not briefed |
bluesky bluesky-ai | 17d | 1000 LPEs! openai.com/index/a-thou... | 1000 LPEs! openai.com/index/a-thou... | ↗Not briefed |
hn hn-ai | 17d | Show HN: Autorun-audit – find what runs when you open, CD into or install a repo | — | ↗Not briefed |