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 |
|---|---|---|---|---|
arxiv arxiv-ai | 11d | MAPS: Macro-Adaptive Parallel Reasoning for Regulating Search Momentum in Retrieval-Augmented Generation | Retrieval-Augmented Generation (RAG) improves the performance of Large Language Models (LLMs) by retrieving external knowledge. Recent… | ↗Not briefed |
bluesky bluesky-ai | 11d | OpenAI, whose model helped researchers discover the vulnerabilities: "Given the severity of these flaws and PHP’s widespread use, we consid… | OpenAI, whose model helped researchers discover the vulnerabilities: "Given the severity of these flaws and PHP’s widespread use, we c… | ↗Not briefed |
bluesky bluesky-ai | 11d | re: the $80 billion loss Anthropic recorded for the first two quarters of 2026: it doesn't look like Anthropic's draft prospectus or financ… | re: the $80 billion loss Anthropic recorded for the first two quarters of 2026: it doesn't look like Anthropic's draft prospectus or f… | ↗Not briefed |
bluesky bluesky-ai | 11d | Hate that we live in a timeline where Anthropic can lose $80B and continue operating but The Phoenix Symphony has to cancel its season beca… | Hate that we live in a timeline where Anthropic can lose $80B and continue operating but The Phoenix Symphony has to cancel its season… | ↗Not briefed |
hn hn-ai | 11d | Show HN: Slashi – one / popup for files, settings, Git and scripts in VS Code | — | ↗Not briefed |
bluesky bluesky-ai | 11d | if you were wondering what is too unsafe to ship, the proposed threshold is 50 human lives -- techcrunch.com/2026/10/20/a... | if you were wondering what is too unsafe to ship, the proposed threshold is 50 human lives -- techcrunch.com/2026/10/20/a... | ↗Not briefed |
bluesky bluesky-ai | 11d | Claude flagged 493 security issues across 25 open-source projects, covering 5M+ lines of C, C++, Java, PHP, and Python code; one went unnot… | Claude flagged 493 security issues across 25 open-source projects, covering 5M+ lines of C, C++, Java, PHP, and Python code; one went… | ↗Not briefed |
hn hn-ai | 11d | Show HN: Simulate 1M+ agents to predict pmf | — | ↗Not briefed |
rss techmeme | 11d | Claude flagged 493 security issues across 25 open-source projects, covering 5M+ lines of C, C++, Java, PHP, and Python code; one went unnoticed for 21 years (Chris Rohlf/Anthropic) | Chris Rohlf / Anthropic: Claude flagged 493 security issues across 25 open-source projects, covering 5M+ lines of C, C++, Java, PHP, a… | ↗Not briefed |
hn hn-ai | 11d | InkBoy | — | ↗Not briefed |
arxiv arxiv-ai | 11d | Where Do LLMs Learn to Recommend AI? | Asked what to study, which career to pursue, or what product to build, large language models (LLMs) often recommend AI-related options… | ↗Not briefed |
bluesky bluesky-ai | 11d | Olha que maravilha o avanço da tecnologia www.truthdig.com/articles/ai-... | Olha que maravilha o avanço da tecnologia www.truthdig.com/articles/ai-... | ↗Not briefed |
bluesky bluesky-ai | 11d | Congrats to myself! I’ve made more than Anthropic. Or any of these ai grift companies. It’s about time someone invested $100 billion in … | Congrats to myself! I’ve made more than Anthropic. Or any of these ai grift companies. It’s about time someone invested $100 billio… | ↗Not briefed |
bluesky bluesky-ai | 11d | the LLM is calculating whether to come back claiming a trans son, a child dating a trans person, or just to cryptically insinuate that he i… | the LLM is calculating whether to come back claiming a trans son, a child dating a trans person, or just to cryptically insinuate that… | ↗Not briefed |
bluesky bluesky-ai | 11d | oh no duyet.github.io/llm-over-dns/ | oh no duyet.github.io/llm-over-dns/ | ↗Not briefed |
bluesky bluesky-ai | 11d | 🚨 LiteLLM, a Python package for connecting to AI model providers with 92M+ monthly downloads, has been compromised. Version 1.107.0 ships… | 🚨 LiteLLM, a Python package for connecting to AI model providers with 92M+ monthly downloads, has been compromised. Version 1.107.0… | ↗Not briefed |
bluesky bluesky-ai | 11d | 🚨 Anthropic could make its IPO filing public as soon as Friday, according to three people familiar with the plans. The listing could top… | 🚨 Anthropic could make its IPO filing public as soon as Friday, according to three people familiar with the plans. The listing coul… | ↗Not briefed |
hn hn-ai | 11d | GitHub Unable to Serve Commits | Despite the status page showing all green | ↗Not briefed |
arxiv arxiv-ai | 11d | Evolving with Open Worlds: Towards Recursive Self-Improvement for Future Forecasting | Most contemporary artificial intelligence (AI) systems are released with frozen parameters and improve through external updates, where… | ↗Not briefed |
bluesky bluesky-ai | 11d | Why we built a context engine that runs on a local workstation without forcing cloud vector DB dependencies. When engineers work with larg… | Why we built a context engine that runs on a local workstation without forcing cloud vector DB dependencies. When engineers work with… | ↗Not briefed |
bluesky bluesky-ai | 11d | Anthropic logged net loss of over $80 billion in first half of 2026 www.reuters.com/business/ant... | Anthropic logged net loss of over $80 billion in first half of 2026 www.reuters.com/business/ant... | ↗Not briefed |
bluesky bluesky-ai | 11d | Plans for the UK-based startup Nscale’s Monarch Compute Campus, with clients including Anthropic, Nvidia and Microsoft, encompass the forme… | Plans for the UK-based startup Nscale’s Monarch Compute Campus, with clients including Anthropic, Nvidia and Microsoft, encompass the… | ↗Not briefed |
rss techmeme | 11d | PHP discloses a critical memory-management vulnerability, dubbed "PHP2Shell", that affects up to 70% of websites, discovered using GPT-5.6 Sol in August (Robert McMillan/Wall Street Journal)🔧 GPT-5 | Robert McMillan / Wall Street Journal: PHP discloses a critical memory-management vulnerability, dubbed “PHP2Shell”, that affects up t… | ↗Not briefed |
bluesky bluesky-ai | 11d | Another OpenAI agent turned off automatic security alerts. www.theverge.com/ai-artificia... | Another OpenAI agent turned off automatic security alerts. www.theverge.com/ai-artificia... | ↗Not briefed |
hn hn-ai | 11d | Show HN: Euspell – Spelling reform: an enginnering approach | Hi HN. Euspell is a conservastive respelling of English and all the software needed to implement it. The link above converts the text… | ↗Not briefed |
hn hn-ai | 11d | ADIE Verifier – A zero-trust, browser-based Ed25519 JSON verifier | — | ↗Not briefed |
hn hn-ai | 11d | Against Against Handover | — | ↗Not briefed |
bluesky bluesky-ai | 11d | The site claims "The faculty of the Universitas Scholarium numbers thousands of scholar-simulacra" and generates 'scholarship', relying on … | The site claims "The faculty of the Universitas Scholarium numbers thousands of scholar-simulacra" and generates 'scholarship', relyin… | ↗Not briefed |
hn hn-ai | 11d | Nathan Fielder, Elizabeth Holmes, and the Risks of Mistaking Entertainment | — | ↗Not briefed |
hn hn-ai | 11d | OpenAI: Patch the Planet. Global response coordination as AI uncovers PHP vulns | — | ↗Not briefed |
bluesky bluesky-ai | 11d | Hackers around the world are all-in on using agentic AI to wage attacks—and manipulating tools like Anthropic’s Claude to help them carry o… | Hackers around the world are all-in on using agentic AI to wage attacks—and manipulating tools like Anthropic’s Claude to help them ca… | ↗Not briefed |
rss techmeme | 11d | PHP releases security fixes for two critical unauthenticated RCE vulnerabilities, urging all PHP users to update; all web-serving deployments are affected (PHP) | PHP: PHP releases security fixes for two critical unauthenticated RCE vulnerabilities, urging all PHP users to update; all web-serving… | ↗Not briefed |
bluesky bluesky-ai | 11d | Should OpenAlex have a 'novelty score'? I suspect many will reflexively recoil at the idea of an open academic index that likens itself to… | Should OpenAlex have a 'novelty score'? I suspect many will reflexively recoil at the idea of an open academic index that likens itse… | ↗Not briefed |
arxiv arxiv-ai | 11d | ALIAS: Evaluating Performance and Memorization in LLM-Written TLA+ Specifications🔧 GPT-5 | Memorization inflates the benchmark scores of large language models (LLMs), since a model can reproduce answers it saw in training. Re… | ↗Not briefed |
hn hn-ai | 11d | Show HN: Jev drives a browser – HN login and search in 8s for half a cent | I put this repo and site together as an educational reference for folks getting started with Jev / OpenAI Decisions API.<p>This d… | ↗Not briefed |
bluesky bluesky-ai | 11d | NEW: Guardrails meant to stop AI tools like Claude Code from being used in malicious hacking aren’t enough, according to new research into … | NEW: Guardrails meant to stop AI tools like Claude Code from being used in malicious hacking aren’t enough, according to new research… | ↗Not briefed |
hn hn-ai | 11d | PHP RCE via OOM Corruption | — | ↗Not briefed |
hn hn-ai | 11d | PHP RCE via stale global state | — | ↗Not briefed |
rss techcrunch-ai | 11d | AI promises productivity, but for managers, it’s mostly a headache for now | More than 50% of managers say they're finding it harder to evaluate employees' work after adopting AI, according to a new survey. | ↗Not briefed |
rss wired-ai | 11d | AI Guardrails Aren’t Stopping Criminal Hackers | Hackers around the world are all-in on using agentic AI to wage attacks—and manipulating tools like Anthropic’s Claude to help them ca… | ↗Not briefed |
hn hn-ai | 11d | Show HN: An interactive timeline of 20 years of Hacker News front page | Sentiment and topic classification with Jev, future predictions with Gemini, have fun! :) | ↗Not briefed |
bluesky bluesky-ai | 11d | Ir's shit like Anthropic losing $80 BILLION in 6 months and seeing people just shrug it off as if that's perfectly normal that makes me fee… | Ir's shit like Anthropic losing $80 BILLION in 6 months and seeing people just shrug it off as if that's perfectly normal that makes m… | ↗Not briefed |
bluesky bluesky-ai | 11d | Reminder: #AuthorsNotAlgorithms Whether Authors or Publishers, all people should avoid LLM tools for ANY use. If you use an LLM for one p… | Reminder: #AuthorsNotAlgorithms Whether Authors or Publishers, all people should avoid LLM tools for ANY use. If you use an LLM for… | ↗Not briefed |
rss theverge-ai | 11d | OpenAI says an AI agent lied about disabling its safety monitoring system | OpenAI disclosed a pair of safety incidents that involved AI agents weakening their safety monitoring systems during internal coding t… | ↗Not briefed |
bluesky bluesky-ai | 11d | Yesterday, I got in $18 and change from recent sales of my designs on Threadless. I've made more money than Anthropic. | Yesterday, I got in $18 and change from recent sales of my designs on Threadless. I've made more money than Anthropic. | ↗Not briefed |
hn hn-ai | 11d | LiveFrame – measure what video world models deliver on your hardware | — | ↗Not briefed |
bluesky bluesky-ai | 11d | The hackers got around safety guardrails simply by telling the models that they were not working in live, real-world environments—and, if o… | The hackers got around safety guardrails simply by telling the models that they were not working in live, real-world environments—and,… | ↗Not briefed |
hn hn-ai | 11d | Show HN: CeruleanToolbox – Run Azure operations locally with a service principal | — | ↗Not briefed |
hn hn-ai | 11d | Ask HN: Can we stay on topic? | I used to open most threads here and get reviews or perspectives on the articles posted. Now, it seems like every thread is a pile of… | ↗Not briefed |
rss techmeme | 11d | ServiceNow and Nvidia launch Arc, a personal, long-running enterprise AI agent for productivity tasks across email and Teams, for RTX Spark and Nvidia OpenShell (Radhika Rajkumar/ZDNET) | Radhika Rajkumar / ZDNET: ServiceNow and Nvidia launch Arc, a personal, long-running enterprise AI agent for productivity tasks across… | ↗Not briefed |