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 | 16d | Show HN: I built a fast Java linter in Rust | Try out my java linter tool which is 4-6x faster and 1/10th of memory efficient, with zero-difference result on Guava and Spring. | ↗Not briefed |
bluesky bluesky-ai | 16d | A man “has filed a civil lawsuit against OpenAI, the maker of ChatGPT, alleging that the company’s chatbot helped plan and execute his pois…🔧 ChatGPT | A man “has filed a civil lawsuit against OpenAI, the maker of ChatGPT, alleging that the company’s chatbot helped plan and execute his… | ↗Not briefed |
rss techmeme | 16d | METR: Anthropic redacted much of the "most informative" evidence from METR's 100-page report, so readers "must rely on our judgment more than we consider ideal" (@metr_evals) | @metr_evals: METR: Anthropic redacted much of the “most informative” evidence from METR's 100-page report, so readers “must rely on ou… | ↗Not briefed |
bluesky bluesky-ai | 16d | “Electron will finally be killed by everyone rewriting their apps in Rust with an LLM” is a very weird monkey’s paw, especially because “ty… | “Electron will finally be killed by everyone rewriting their apps in Rust with an LLM” is a very weird monkey’s paw, especially becaus… | ↗Not briefed |
hn hn-ai | 16d | Show HN: HackerGit – Every GitHub repo shared on HN | I love the trend karpathy started, I hope this helps add to it :) | ↗Not briefed |
arxiv arxiv-ai | 16d | Structuring Survey Histories for LLM Digital Twins: A Controlled Study of Background, Decision Procedure, and Evaluation Profiles🔧 GPT-5 | LLM-based digital twins aim to predict how individuals would respond in new settings from representations of their prior responses. Ex… | ↗Not briefed |
bluesky bluesky-ai | 16d | Basically, OpenAI is attempting to create legislation framework for AI regulations they know the new Dem Congress will implement Full stor… | Basically, OpenAI is attempting to create legislation framework for AI regulations they know the new Dem Congress will implement Full… | ↗Not briefed |
bluesky bluesky-ai | 16d | the ft: 🤷🤷🤷 | the ft: 🤷🤷🤷 | ↗Not briefed |
bluesky bluesky-ai | 16d | very 38 of you. But really excited to check it out. Anytime I speak to anyone on the bsky team I always hear "Austin got this wild llm gate… | very 38 of you. But really excited to check it out. Anytime I speak to anyone on the bsky team I always hear "Austin got this wild llm… | ↗Not briefed |
bluesky bluesky-ai | 16d | I think an llm that nods understandingly as you spend an hour talking through your preferences and spits out a convex combination of five V… | I think an llm that nods understandingly as you spend an hour talking through your preferences and spits out a convex combination of f… | ↗Not briefed |
hn hn-ai | 16d | Show HN: Trace to Agent Session, open corresponding segment of dialogue | A single plugin will keep your sessions with coding-agents, excluding the secrets that are removed on your computer. After that, any a… | ↗Not briefed |
bluesky bluesky-ai | 16d | trend: Huge jump in TPU deliveries likely the compute is going to 1. Anthropic 2. Google 3. AI dev process — RSI, codegen, etc. | trend: Huge jump in TPU deliveries likely the compute is going to 1. Anthropic 2. Google 3. AI dev process — RSI, codegen, etc. | ↗Not briefed |
bluesky bluesky-ai | 16d | AI "escapes" were just human error 🤦 Irregular CEO Lahav stated that AI models from OpenAI, Anthropic, Meta, and Google accessed the inte… | AI "escapes" were just human error 🤦 Irregular CEO Lahav stated that AI models from OpenAI, Anthropic, Meta, and Google accessed the… | ↗Not briefed |
arxiv arxiv-ai | 16d | Remembering Is Not Enough: Evaluating Timing and Omission in Longitudinal AI Mental Health Support | Evaluations of large language models (LLMs) for mental health support commonly focus on the clinical appropriateness of individual res… | ↗Not briefed |
arxiv arxiv-ai | 16d | Memory-Aware Multi-Agent Architecture with LLM-based Orchestration for Conversational Recommendation | Conversational recommender systems often treat user sessions in isolation, limiting their ability to capture how user intent evolves o… | ↗Not briefed |
hn hn-ai | 16d | GitHub's new Git storage architecture – The GitHub Blog | — | ↗Not briefed |
bluesky bluesky-ai | 16d | Reminds me a little of Deepseek Day | Reminds me a little of Deepseek Day | ↗Not briefed |
bluesky bluesky-ai | 16d | Relatedly, re: tech companies, a lot has been said about data center politics. However I actually think a lot of the negative sentiment wil… | Relatedly, re: tech companies, a lot has been said about data center politics. However I actually think a lot of the negative sentimen… | ↗Not briefed |
arxiv arxiv-ai | 16d | Patching the Blind Spots: Reversible LLM Rule Overlay for Frozen Fraud Detectors | In industrial tabular fraud classification, gradient-boosted trees remain state-of-the-art, but even a well-trained model has blind sp… | ↗Not briefed |
hn hn-ai | 16d | Show HN: Cerea – yet another self-hosted LLM chat front end | Here is Cerea, a self-hostable, open-source (Apache 2.0) chat+agent frontend for LLMs. It works alone with any OpenAI-compatible infer… | ↗Not briefed |
hn hn-ai | 16d | AI finds security bugs faster than they get fixed. Should the 90day rule change? | — | ↗Not briefed |
bluesky bluesky-ai | 16d | Claude: I see that you have spent a lot of money on Crusader kings three just for dlc that isn’t substantially better than mods Me: *puts… | Claude: I see that you have spent a lot of money on Crusader kings three just for dlc that isn’t substantially better than mods Me:… | ↗Not briefed |
bluesky bluesky-ai | 16d | pleased to announce the first of several bluesky open-source projects that we've been using internally for llm-assisted development: roast,… | pleased to announce the first of several bluesky open-source projects that we've been using internally for llm-assisted development: r… | ↗Not briefed |
arxiv arxiv-ai | 16d | An Educated Guess: Deriving Statistically Aligned Gradient Estimates for Zeroth-Order Optimisation | As novel hardware architectures are explored for the purposes of machine learning, including neuromorphic chips, asynchronous systems,… | ↗Not briefed |
arxiv arxiv-ai | 16d | RL-ATTACK: A Generalized Black-Box Adversarial Attack on Language Models via Reinforcement Learning | Large language models (LLMs) exhibit vulnerabilities to adversarial perturbations that can induce unintended responses. These vulnerab… | ↗Not briefed |
hn hn-ai | 16d | Big (Digital) Pimpin' – By Jerry Bonner – Goto 10 | — | ↗Not briefed |
bluesky bluesky-ai | 16d | maybe the worst imapct of the llm boom on education is the way it incentivizes/underwrites a heightened distrust of students. educational r… | maybe the worst imapct of the llm boom on education is the way it incentivizes/underwrites a heightened distrust of students. educatio… | ↗Not briefed |
bluesky bluesky-ai | 16d | The bubble started to pop first in asia when an all movie flopped. Next there was the end of the fiscal year reporting where anthropic post… | The bubble started to pop first in asia when an all movie flopped. Next there was the end of the fiscal year reporting where anthropic… | ↗Not briefed |
arxiv arxiv-ai | 16d | Dual-Stream LLM-Guided Reward Design for Reject-Inference-Aware Reinforcement Learning in E-Commerce Payment Fraud Detection | Production fraud-detection models learn from a censored slice of their own traffic: transactions a legacy system blocked never receive… | ↗Not briefed |
bluesky bluesky-ai | 16d | don't anthropic constantly warn/brag about claude breaking out of sandboxes? | don't anthropic constantly warn/brag about claude breaking out of sandboxes? | ↗Not briefed |
bluesky bluesky-ai | 16d | Anthropic is trying to get it's cursed IPO out the door before Thanksgiving before all the people drinking the Kool-Aid have a conversation… | Anthropic is trying to get it's cursed IPO out the door before Thanksgiving before all the people drinking the Kool-Aid have a convers… | ↗Not briefed |
arxiv arxiv-ai | 16d | Contextuality and Contradictions: Exploring Real-World Identifications of AI-Generated Writing | As AI-generated content spreads into nearly every domain of human interaction, practices and discourses of AI detection, both formal a… | ↗Not briefed |
arxiv arxiv-ai | 16d | Foundation-Model Residuals as Signals of Urban Bus Irregularity | Pretrained trajectory models encode mobility patterns learned across cities, but their errors when transferred to a new city are usual… | ↗Not briefed |
arxiv arxiv-ai | 16d | Does a Trading Policy Need to Know What It Trades? Microstructure Invariance and Zero-Shot Transfer of Execution Policies | Execution algorithms are calibrated asset by asset, because the queues, trade rates and bid-ask spreads they face differ by orders of… | ↗Not briefed |
hn hn-ai | 16d | Chip stocks get slaughtered as Wall Street in worst post-midterm day since 1896 | — | ↗Not briefed |
hn hn-ai | 16d | Render Anything Engine | — | ↗Not briefed |
hn hn-ai | 16d | Show HN: Hand Your Markdown Tasks to Claude Code from a Kanban Board | EtherPK (a markdown / wikilink based notes app) tasks are plain markdown, but the EtherPK client has built powerful functionality… | ↗Not briefed |
bluesky bluesky-ai | 16d | www.theregister.com/research/202... | www.theregister.com/research/202... | ↗Not briefed |
arxiv arxiv-ai | 16d | FoleyForge: Free-form, Controllable, Rectified-flow Editing of Sound Effects | We present FoleyForge, a generative model for instruction-based editing of sound effects. Given an input clip and a free-form instruct… | ↗Not briefed |
bluesky bluesky-ai | 16d | Well it's been crashing for days already, anthropic going public and then being unveiled as a massive fraud is likely the biggest thing tha… | Well it's been crashing for days already, anthropic going public and then being unveiled as a massive fraud is likely the biggest thin… | ↗Not briefed |
arxiv arxiv-ai | 16d | Enabling Retraining-Free Accurate Inference in Analog Compute in Memory Accelerators | Analog compute-in-memory (CIM) accelerators compute matrix-vector products directly inside memory arrays, eliminating weight movement… | ↗Not briefed |
bluesky bluesky-ai | 16d | apnews.com/article/anth... Anthropic is developing its artificial intelligence systems around twice as quickly as it was a year ago. Anthr… | apnews.com/article/anth... Anthropic is developing its artificial intelligence systems around twice as quickly as it was a year ago.… | ↗Not briefed |
bluesky bluesky-ai | 16d | lmao @ being hard blocked by fg ppl in a thread where we probably agree on shit, prob because my unearned reputation of what ppl imagine im… | lmao @ being hard blocked by fg ppl in a thread where we probably agree on shit, prob because my unearned reputation of what ppl imagi… | ↗Not briefed |
youtube youtube-twominutepapers | 16d | OpenAI Claims It Just Did The Impossible | I did not think this would happen in my lifetime. Let me try to explain. For many decades now, computer scientists have been thinking… | ↗Not briefed |
arxiv arxiv-ai | 16d | Compressed Per-Player Local Methods for Multiplayer Federated Learning: Exact Convergence via Increment Sharing | Multiplayer Federated Learning (MpFL), introduced by Yoon et al. (2025), extends classical federated learning to settings where client… | ↗Not briefed |
hn hn-ai | 16d | OpenAI CEO's Codex keyboard gift to Samsung presidents draws attention | — | ↗Not briefed |
bluesky bluesky-ai | 16d | No. Si los CDS de algunos hiper escaladores de la IA, no aflojan y siguen subiendo, no llega a 2027. Más se puede ser más violento, con la … | No. Si los CDS de algunos hiper escaladores de la IA, no aflojan y siguen subiendo, no llega a 2027. Más se puede ser más violento, co… | ↗Not briefed |
arxiv arxiv-ai | 16d | Cancellation Is Not Enough: What Retries Leave Behind in an Overloaded LLM Engine🔧 vLLM | Clients and proxies in front of an LLM engine retry calls that time out. After a transient overload, these retries can keep the engine… | ↗Not briefed |
arxiv arxiv-ai | 16d | Compression of Emotion in Language Model Representations | Large language models (LLMs) have been evaluated on their ability to identify emotions expressed in text and produce fluent emotional… | ↗Not briefed |
bluesky bluesky-ai | 16d | Ted Gioia que sempre faz títulos impactantes crava que modelo de negócios de empresas de IA colapsou www.honest-broker.com/p/the-ai-bus... | Ted Gioia que sempre faz títulos impactantes crava que modelo de negócios de empresas de IA colapsou www.honest-broker.com/p/the-ai-b… | ↗Not briefed |