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 |
|---|---|---|---|---|
hn hn-ai | 1mo | Claude Opus 5.5 for code review: More catches, different misses | — | ↗Not briefed |
hn hn-ai | 1mo | Claude Opus 5.5 | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | TraceVIC: Causal Reasoning over Code Evolution for Identifying Vulnerability-Inducing Commits | Software vulnerabilities are often discovered long after they are introduced, making it difficult to identify the vulnerability-induci… | ↗Not briefed |
hn hn-ai | 1mo | Anthropic now offering limit resets like Codex | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic absolutely beclowned since the advent of Sol et al. their game to lose and they stopped for an indulgent nap. let's see if they c… | Anthropic absolutely beclowned since the advent of Sol et al. their game to lose and they stopped for an indulgent nap. let's see if t… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Super excited about our Survey with Lihu Chen on the Role of Small Models in the LLM Era This topic is crucial in today's world direct.mit… | Super excited about our Survey with Lihu Chen on the Role of Small Models in the LLM Era This topic is crucial in today's world direc… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Train Where the Quantized Model Goes: On-Policy Distillation for Low-Bit Reasoning | Quantization-aware distillation (QAD) restores much of the short-form question-answering performance lost to sub-3-bit quantization, y… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI sued by Canadian province over role in mass shooting www.ft.com/content/30d2... | OpenAI sued by Canadian province over role in mass shooting www.ft.com/content/30d2... | ↗Not briefed |
arxiv arxiv-ai | 1mo | Optimal Sequential Annotations for Off-Policy Evaluation | Offline reinforcement learning and off-policy evaluation evaluates dynamic treatment rules based on retrospectively collected data pri… | ↗Not briefed |
rss bing-news-ai | 1mo | EASA Carries On With Cautious Approach To Artificial Intelligence | EASA Carries On With Cautious Approach To Artificial Intelligence is available to both Aviation Week & Space Technology and AWIN subsc… | ↗Not briefed |
bluesky bluesky-ai | 1mo | its crazy anthropic completely ripped off the big walk trailer music for opus 5.5 release | its crazy anthropic completely ripped off the big walk trailer music for opus 5.5 release | ↗Not briefed |
arxiv arxiv-ai | 1mo | When are bosonic Gaussian states classical to learn? | A fundamental question in physics is: When does classical behavior emerge from quantum systems? Bosonic Gaussian states provide a natu… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Beyond Repeated Sampling: Learning Search Policies for LLM Reasoning | Large language models increasingly tackle hard reasoning problems by spending more test-time compute, yet the dominant strategy remain… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic boosts Claude Opus 5.5 safety 🛡️ Anthropic released Claude Opus 5.5 with stricter cybersecurity safeguards to curb risky behavi… | Anthropic boosts Claude Opus 5.5 safety 🛡️ Anthropic released Claude Opus 5.5 with stricter cybersecurity safeguards to curb risky b… | ↗Not briefed |
bluesky bluesky-ai | 1mo | “The incentives are to put engineering resource or dev resource on what drives the bottom line or what drives stock valuations, rather than… | “The incentives are to put engineering resource or dev resource on what drives the bottom line or what drives stock valuations, rather… | ↗Not briefed |
bluesky bluesky-ai | 1mo | a16z launches homework-free tech academy Andreessen Horowitz is launching a free, year-long "Horowitz Andreessen Academy" for 50 high scho… | a16z launches homework-free tech academy Andreessen Horowitz is launching a free, year-long "Horowitz Andreessen Academy" for 50 high… | ↗Not briefed |
bluesky bluesky-ai | 1mo | While I hate the LLM industry, this is at last a viable use case I can get behind. “Hey dumbass, pay 300$ to have an AI buddy on your face… | While I hate the LLM industry, this is at last a viable use case I can get behind. “Hey dumbass, pay 300$ to have an AI buddy on your… | ↗Not briefed |
rss arstechnica-tech | 1mo | IT mistake erases 11 years of viewing history for hospitals’ maternity records | The English hospitals recovered patient care data only. | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Groupicorn directory of IOP group therapy programs | Hello fellow tech workers!<p>My name is Andrew, cs major from University of Pittsburgh 1996 and long time BBS SysOp from 1991-1993. I… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Gosh, I really love the opensource community. This morning I was thinking to myself, if my potato M1 Mac Mini was worth keeping on for jus… | Gosh, I really love the opensource community. This morning I was thinking to myself, if my potato M1 Mac Mini was worth keeping on fo… | ↗Not briefed |
hn hn-ai | 1mo | Code similarity detection using Tree-sitter | — | ↗Not briefed |
hn hn-ai | 1mo | ES Archive – a new memory system for AI, running natively on Mac | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI and Anthropic falling over themselves to take credit | OpenAI and Anthropic falling over themselves to take credit | ↗Not briefed |
hn hn-ai | 1mo | Claude Opus 5.5: Intelligence, Performance and Price Analysis | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | The LLM swimsuit competition will be epic. | The LLM swimsuit competition will be epic. | ↗Not briefed |
bluesky bluesky-ai | 1mo | reads so much better, kinda like luna has been. plus at least it's still pretty easy to tell when something's coming from a llm | reads so much better, kinda like luna has been. plus at least it's still pretty easy to tell when something's coming from a llm | ↗Not briefed |
hn hn-ai | 1mo | Xeno-Interpretability: Investigating the Alien Minds of LLMs | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | godspeed to everyone that wants an llm to do their banking and all that | godspeed to everyone that wants an llm to do their banking and all that | ↗Not briefed |
bluesky bluesky-ai | 1mo | "The best investment would be a problem gambler." | "The best investment would be a problem gambler." | ↗Not briefed |
bluesky bluesky-ai | 1mo | We literally just published this about OpenAI today: Humans are training ChatGPT by reading users' prompts -- which yes, can contain sensit…🔧 ChatGPT | We literally just published this about OpenAI today: Humans are training ChatGPT by reading users' prompts -- which yes, can contain s… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Measuring the Serving Stack Instead of the Model: Hidden Confounds in Local Tool-Use Evaluation🔧 Ollama🔧 vLLM | A coding agent must emit a valid tool call--a parseable invocation of a tool in the provided schema--before the harness can execute it… | ↗Not briefed |
hn hn-ai | 1mo | How do you best protect against copycats? | Someone's created pixel-perfect copies of my website (prepfully.com).<p>Whoever it is, is using the same domain name, but differe… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic releases cheaper AI model ahead of IPO ft.trib.al/7tHpgjj | Anthropic releases cheaper AI model ahead of IPO ft.trib.al/7tHpgjj | ↗Not briefed |
bluesky bluesky-ai | 1mo | 🚨 Breaking: Vance/Oz to kick ~760,000 off of ACA coverage and/or subsidies based on "fraud" claims "predicted" in part by an AI model: ac… | 🚨 Breaking: Vance/Oz to kick ~760,000 off of ACA coverage and/or subsidies based on "fraud" claims "predicted" in part by an AI model… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Paul brothers are investing heavy into gambling, AI, drones, and defense protos.com/paul-brother... | Paul brothers are investing heavy into gambling, AI, drones, and defense protos.com/paul-brother... | ↗Not briefed |
hn hn-ai | 1mo | Show HN: graf (1000x faster graphify in Rust) | I liked the idea of graphify, but for large repos the search queries were visibly very slow.<p>I dug into it and saw an opportunity to… | ↗Not briefed |
rss techmeme | 1mo | Sources: Apple is developing prototypes for a screenless health and fitness tracker similar to Whoop but hasn't decided whether to release the product (Mark Gurman/Bloomberg) | Mark Gurman / Bloomberg: Sources: Apple is developing prototypes for a screenless health and fitness tracker similar to Whoop but hasn… | ↗Not briefed |
rss bing-news-ai | 1mo | Is it ‘killer robots’ or ‘super intelligence’? At the UN, leaders see both | Trump was referring to a 2018 UN report that outlined the danger of human-caused climate change for long-term breaching of the interna… | ↗Not briefed |
hn hn-ai | 1mo | A framework for frontier AI and the dawning of a new age | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Detecting GPT-Assisted Writing Using Interpretable Stylometric Features🔧 ChatGPT | Distinguishing GPT-assisted from independently authored student writing has become a critical challenge in academia. This paper evalua… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Chis Nicholson (Executive Director of Digital Antiquity / tDAR) and I wrote an article with a critique of the Advisory Council on Historic … | Chis Nicholson (Executive Director of Digital Antiquity / tDAR) and I wrote an article with a critique of the Advisory Council on Hist… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic says Opus 5.5 matches Fable 5.1 "on most tasks" while costing about 40% less to run than Opus 5; Opus 5.5 costs $4/1M input and $… | Anthropic says Opus 5.5 matches Fable 5.1 "on most tasks" while costing about 40% less to run than Opus 5; Opus 5.5 costs $4/1M input… | ↗Not briefed |
rss bing-news-ai | 1mo | Is it ‘killer robots’ or ‘super intelligence’? At the UN, leaders see both | Trump was referring to a 2018 UN report that outlined the danger of human-caused climate change for long-term breaching of the interna… | ↗Not briefed |
arxiv arxiv-ai | 1mo | PROSWIN: Probabilistic Solar Wind Speed Forecasting Using Deep Distributional Regression From Solar Images | Accurately predicting fast solar wind conditions is challenging, as uncertainties are large and unquantified by traditional single-val… | ↗Not briefed |
rss techmeme | 1mo | Anthropic says Opus 5.5 matches Fable 5.1 "on most tasks" while costing about 40% less to run than Opus 5; Opus 5.5 costs $4/1M input and $20/1M output tokens (Matthias Bastian/The Decoder) | Matthias Bastian / The Decoder: Anthropic says Opus 5.5 matches Fable 5.1 “on most tasks” while costing about 40% less to run than Opu… | ↗Not briefed |
arxiv arxiv-ai | 1mo | From Alignment to Access Control: A Framework for GenAI Policy Enforcement | Generative AI (GenAI) applications have flourished enabling users to chat with large language models, and to create agents to act on t… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Gotta invent a name for a new category of guy who only listens to common-sense advice if the LLM tells him to. | Gotta invent a name for a new category of guy who only listens to common-sense advice if the LLM tells him to. | ↗Not briefed |
bluesky bluesky-ai | 1mo | Rat brain AI lands on AWS The Biological Computing Company (TBC) will offer its "rat brain" AI model to select Amazon Web Services custome… | Rat brain AI lands on AWS The Biological Computing Company (TBC) will offer its "rat brain" AI model to select Amazon Web Services cu… | ↗Not briefed |
hn hn-ai | 1mo | Go-System-One: Jev-Like Results in Pure Go and SIMD/PTX | — | ↗Not briefed |
rss bing-news-ai | 1mo | Is it ‘killer robots’ or ‘super intelligence’? At the UN, leaders see both | Visions of the future straight out of comic books are playing out at the United Nations as world leaders talk about the threat — or pr… | ↗Not briefed |