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 | 2mo | A Scalable Framework for Automated NER Annotation Correction in Low-Resource Languages | Poor quality or noisy annotations in Named Entity Recognition (NER), as in any other NLP task, make it challenging to achieve state-of… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Booked a service appointment for my whip and had to use some LLM chatbot. I made sure to have the service department contact me because I w… | Booked a service appointment for my whip and had to use some LLM chatbot. I made sure to have the service department contact me becaus… | ↗Not briefed |
bluesky bluesky-ai | 2mo | No, that is not what they're proposing lol you just suck at reading comprehension Theyre saying terminator isn't real and is a marketing … | No, that is not what they're proposing lol you just suck at reading comprehension Theyre saying terminator isn't real and is a marke… | ↗Not briefed |
hn hn-ai | 2mo | Microsoft says AI rival Anthropic could have 'disastrous impact' on humanity | — | ↗Not briefed |
hn hn-ai | 2mo | The DeepMind Institute | — | ↗Not briefed |
arxiv arxiv-ai | 2mo | Clueing up LLMs with Tool-Augmented Deductive Reasoning | Despite recent advances in large language models (LLMs), performing logically consistent deductive reasoning over extended interaction… | ↗Not briefed |
hn hn-ai | 2mo | Show HN: ManyBot – Framework to build WhatsApp bots, without the boring part | Built on top of Baileys. The goal is to handle all the tedious work—opening sockets, authentication, reloading—and leave you to focus… | ↗Not briefed |
hn hn-ai | 2mo | System One Lite – typed decisions from a local LLM, with no generated tokens | — | ↗Not briefed |
hn hn-ai | 2mo | PS5 Linux lead quits: "a bunch of noobs using LLMs" that "they don't understand" | — | ↗Not briefed |
arxiv arxiv-ai | 2mo | Which LLM is Best for Translating Natural Language Goals to PDDL | Bridging the gap between human intent and machine execution remains a challenge in automated planning, where expressing goals in forma… | ↗Not briefed |
arxiv arxiv-ai | 2mo | "If I Had to Buy Just ONE: Galaxy S26 Ultra": Auditing AI-Generated Product Recommendations🔧 ChatGPT | Consumers increasingly use AI chatbots for advice on what to buy. With companies like OpenAI and Google monetising their AI through ad… | ↗Not briefed |
hn hn-ai | 2mo | Anatomy of a Texture | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | And, like, ultimately while the safety/speed balance is a little out of kilter, I think we do want US-aligned companies investigating and q… | And, like, ultimately while the safety/speed balance is a little out of kilter, I think we do want US-aligned companies investigating… | ↗Not briefed |
bluesky bluesky-ai | 2mo | i think a lot of people are assuming diffusion because they said “parallel”, but an AR LLM generates probabilities for thousands of possibl… | i think a lot of people are assuming diffusion because they said “parallel”, but an AR LLM generates probabilities for thousands of po… | ↗Not briefed |
hn hn-ai | 2mo | Show HN: OmnisBench, a re-gradable, open LLM routing benchmark on fresh tasks | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | Microsoft AI chief calls out Anthropic's approach to AI consciousness reut.rs/4h7Q2sP | Microsoft AI chief calls out Anthropic's approach to AI consciousness reut.rs/4h7Q2sP | ↗Not briefed |
rss techcrunch-ai | 2mo | Threads’ new features let podcasters promote shows and reach listeners | Threads is rolling out new tools for podcasters, including profile cards, episode links, transcripts, guest tags, posting reminders an… | ↗Not briefed |
hn hn-ai | 2mo | BurpSqueezer – Turn Burp Suite XML Dumps into Compact LLM-Ready Markdown | — | ↗Not briefed |
hn hn-ai | 2mo | Voila: Ship Web Apps Straight from Your LLM Agent | — | ↗Not briefed |
hn hn-ai | 2mo | Show HN: Swift-Qwen3.8-27B, -58.3% thinking, x1.95 speed, accuracy of xhigh🔧 Qwen 3 | Hi everybody, we post-trained Qwen 3.8 27B to be more efficient by figuring out which tokens were linked to overthinking and penalizin… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Beyond Truncation: Rethinking LLM Decoding as Ensemble Pruning | We introduce Mahalanobis-Ensemble Decoding (ME-Decoding), a novel Large Language Model (LLM) decoding framework that frames candidate… | ↗Not briefed |
bluesky bluesky-ai | 2mo | with this updated prompt I almost never have to edit the resulting alt text. the command reads from my clipboard and copies the alt text to… | with this updated prompt I almost never have to edit the resulting alt text. the command reads from my clipboard and copies the alt te… | ↗Not briefed |
arxiv arxiv-ai | 2mo | LocQE: Principled Domain Adaptation for Localisation Quality Estimation by Leveraging Post-Edits | Learned quality estimation (QE) models such as COMETKiwi are widespread and work well for general machine translation evaluation. Howe… | ↗Not briefed |
hn hn-ai | 2mo | FoxSchema: Database schema diff and migration tool | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | but I guess the idea with the confidence is that the model itself is capable of telling you when it's a good idea to use it. if that actual… | but I guess the idea with the confidence is that the model itself is capable of telling you when it's a good idea to use it. if that a… | ↗Not briefed |
bluesky bluesky-ai | 2mo | attaching context: Jacob Coxon is who quit anthropic and got tons of news saying AI will kill us. in AMA made clear he's one of those peopl… | attaching context: Jacob Coxon is who quit anthropic and got tons of news saying AI will kill us. in AMA made clear he's one of those… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Rethinking Critic Learning in PPO: Understanding and Mitigating Value Flattening | In reinforcement learning for large language models, Proximal Policy Optimization (PPO) commonly uses a critic to estimate state value… | ↗Not briefed |
rss techcrunch-ai | 2mo | Next wave of VCs judging Startup Battlefield 200 contenders at TechCrunch Disrupt 2026 revealed | Meet the next five top-tier investors judging the Startup Battlefield 200 contenders live at TechCrunch Disrupt 2026. Register before… | ↗Not briefed |
bluesky bluesky-ai | 2mo | the new york post is carrying water for the effective accelerationists, preposterously attacking effective altruists as "the woke" "far-lef… | the new york post is carrying water for the effective accelerationists, preposterously attacking effective altruists as "the woke" "fa… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Echo: Learning-based Matching Decompilation using Trusted Back Translation🔧 GPT-5 | Neural decompilers can recover readable and recompilable source code from binaries, but their predictions remain difficult to trust. M… | ↗Not briefed |
bluesky bluesky-ai | 2mo | even after that if you are the kind of unhinged FREAK who gets excited for what an llm can do with your work, you save like $64M (citation … | even after that if you are the kind of unhinged FREAK who gets excited for what an llm can do with your work, you save like $64M (cita… | ↗Not briefed |
hn hn-ai | 2mo | Show HN: Ctxwitch – Git tells you what changed; this tells you what it'll do | — | ↗Not briefed |
arxiv arxiv-ai | 2mo | Toward Composable Network Digital Twins: A Subgraph-Based Latency Prediction Study | Modern networks must support changing topologies, configurations, and performance objectives, motivating fast and reliable performance… | ↗Not briefed |
hn hn-ai | 2mo | Switched to Claude Code's open-source rival, stopped worrying about rate limits | — | ↗Not briefed |
hn hn-ai | 2mo | Show HN: Ori, an agentic runtime for infrastructure that must ask before it acts | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | New in @towcenter.bsky.social's Deals & Disputes Tracker: 💰 OpenAI provides 400+ ChatGPT Edu subscriptions to CUNY & Northwestern J-schoo…🔧 ChatGPT | New in @towcenter.bsky.social's Deals & Disputes Tracker: 💰 OpenAI provides 400+ ChatGPT Edu subscriptions to CUNY & Northwestern J-… | ↗Not briefed |
hn hn-ai | 2mo | An LLM from First Principles | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | On je prvý chodiaci a dýchajúci LLM. | On je prvý chodiaci a dýchajúci LLM. | ↗Not briefed |
bluesky bluesky-ai | 2mo | Stoat had a single claude contribution that got reverted and they proceeded to ban AI from the codebase. Fluxer doesn't list agents on comm… | Stoat had a single claude contribution that got reverted and they proceeded to ban AI from the codebase. Fluxer doesn't list agents on… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Tracing individual knowledge trajectories in a changing field: the case of general relativity and gravitation | Historians have reconstructed the twentieth-century transformation of general relativity and gravitation (GRG) at the field level and… | ↗Not briefed |
bluesky bluesky-ai | 2mo | I would be strongly opposed to “AI regulation” where Brendan Carr gets to appoint a “Patriotism Ombudsman” to A\ with a mandate to ablate C… | I would be strongly opposed to “AI regulation” where Brendan Carr gets to appoint a “Patriotism Ombudsman” to A\ with a mandate to abl… | ↗Not briefed |
bluesky bluesky-ai | 2mo | i haven’t been paying attention to Grok Bot, but my suspicion is that openai could clone the idea then drop a more persistent version of So… | i haven’t been paying attention to Grok Bot, but my suspicion is that openai could clone the idea then drop a more persistent version… | ↗Not briefed |
bluesky bluesky-ai | 2mo | i stopped playing several seasons ago. fortnite's just boring. it's all collabs and has been for way too long. i'm fatigued of the battle … | i stopped playing several seasons ago. fortnite's just boring. it's all collabs and has been for way too long. i'm fatigued of the ba… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Novo and Anthropic announced today a collaboration to accelerate the development of new medicines and AI-driven software. | Novo and Anthropic announced today a collaboration to accelerate the development of new medicines and AI-driven software. | ↗Not briefed |
bluesky bluesky-ai | 2mo | i think yud's crash out about Anthropic focusing, in any amount, on them "cozying up to the left" is so stunningly political naive, that it… | i think yud's crash out about Anthropic focusing, in any amount, on them "cozying up to the left" is so stunningly political naive, th… | ↗Not briefed |
arxiv arxiv-ai | 2mo | RankGround: Efficient High-Resolution GUI Grounding via Lightweight Reranker-Guided Crop Selection | Graphical User Interface (GUI) grounding is a fundamental perception task for multimodal agents, enabling them to interpret natural la… | ↗Not briefed |
bluesky bluesky-ai | 2mo | We are receiving a number of enquiries about Anthropic settlement notices which have recently been sent out. Our Advisory team has put toge… | We are receiving a number of enquiries about Anthropic settlement notices which have recently been sent out. Our Advisory team has put… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Generalist-Specialist Mixture-of-Experts for Rare Pathology Detection in Multimodal Imaging🔧 Modal | AI models for multimodal medical imaging must balance modality-specific specialization with cross-modal shared representations, a trad… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Een uur per dag sluit ik mijn brein aan op een OpenAI trainingscentrum. Zo kunnen ze mijn denkkracht en ervaring uploaden naar hun chatbotp… | Een uur per dag sluit ik mijn brein aan op een OpenAI trainingscentrum. Zo kunnen ze mijn denkkracht en ervaring uploaden naar hun cha… | ↗Not briefed |
hn hn-ai | 2mo | SpartanX Solar lousy customer service | I cannot recommend SpartanX as a solar power partner. The package they put together was fair, but the service completely disappeared t… | ↗Not briefed |