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 | 1mo | Echo State Network (ESN) for Signal Recovery in RF-Impaired IBFD MIMO Systems | In-band full-duplex (IBFD) multiple-input multiple-output (MIMO) systems enable simultaneous transmission and reception on the same fr… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Recommended reading. One thing I'm taking from this is that some of the cabinet officials and WH staff basically thought Fable was what so… | Recommended reading. One thing I'm taking from this is that some of the cabinet officials and WH staff basically thought Fable was wh… | ↗Not briefed |
hn hn-ai | 1mo | Chronic stress may trigger hidden inflammation that damages the heart | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | sad panda really needs to mandate AI/LLM translated/generated uploads and add some kind of easy site wide blacklist toggle. or just ban AI … | sad panda really needs to mandate AI/LLM translated/generated uploads and add some kind of easy site wide blacklist toggle. or just ba… | ↗Not briefed |
arxiv arxiv-ai | 1mo | XYEval: Agents say yes to bad advice | Effective communication between users and AI agents is essential for human-AI collaboration. The XY problem is a well-known communicat… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Sparse Regression Distilled from a Single Robust Fit | Robust linear fits can resist response contamination yet remain too dense or unstable for useful global explanations. We propose penal… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I'm glad he said poetry here, because that's the perfect example. An LLM absolutely cannot write poetry. Poetry is not just words. When I r… | I'm glad he said poetry here, because that's the perfect example. An LLM absolutely cannot write poetry. Poetry is not just words. Whe… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anyone talking about how if Anthropic or OAI gets hacked, it’s basically the end of the world. Injecting arbitrary tool calls into all mode… | Anyone talking about how if Anthropic or OAI gets hacked, it’s basically the end of the world. Injecting arbitrary tool calls into all… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Trump *really* doesn’t like Anthropic | Trump *really* doesn’t like Anthropic | ↗Not briefed |
arxiv arxiv-ai | 1mo | Measuring the Assistant's Harmlessness Preferences on the User Turn | Post-training turns a general next-token predictor into a chat model with a persistent assistant persona. If that persona is a charact… | ↗Not briefed |
arxiv arxiv-ai | 1mo | The Neural Forcing for Three-Dimensional Incompressible Navier-Stokes finite time blowup | We present a two-part neural framework for forced three-dimensional incompressible Navier--Stokes flow. Part~I develops the computatio… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Often LLMs are a bad, inefficient way to automate some drudgery. Sometimes the best use of an LLM is to solve a problem in a way that takes… | Often LLMs are a bad, inefficient way to automate some drudgery. Sometimes the best use of an LLM is to solve a problem in a way that… | ↗Not briefed |
hn hn-ai | 1mo | DAPO: An Open-Source RL System from ByteDance Seed and Tsinghua Air | — | ↗Not briefed |
rss techmeme | 1mo | China has made AI a national mission, but local workers' fear of displacement mirrors worries abroad, and potential for social unrest may force CCP to slow down (Michael Schuman/The Atlantic) | Michael Schuman / The Atlantic: China has made AI a national mission, but local workers' fear of displacement mirrors worries abroad,… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The similarity between LLM chatbot output and "psychic" or conman cold-reading statements seems worth noting: softwarecrisis.dev/letters/ll… | The similarity between LLM chatbot output and "psychic" or conman cold-reading statements seems worth noting: softwarecrisis.dev/lette… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Lightspeed, bidirectional WebSockets for Laravel (with asteroids demo) | Hi HN!<p>Laravel has worked with websockets for years but generally speaking it's a one way push architecture (i.e. http in, sock… | ↗Not briefed |
bluesky bluesky-ai | 1mo | this is real dunning-kruger stuff he's doing, and I'm not a computational linguist myself, but an answer to why LLMs don't think is that th… | this is real dunning-kruger stuff he's doing, and I'm not a computational linguist myself, but an answer to why LLMs don't think is th… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Density-Ratio Rescoring for Imbalanced Classification | Density-Ratio Rescoring (DRR) augments a classifier trained at the original class prior with a survey-raking dual score. Raking reweig… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Jevopt: Making intelligent compiler optimisation decisions with Jev | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | MCPGen: Benchmarking LLMs on Executable MCPWorkflow Development | We study whether LLMs can produce executable workflow artifacts that remain consistent across graph structure, tool implementation, sc… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Try Omarchy for Windows – Full Omarchy Desktop on Windows | No VM setup or configuration. Download TryOmarchy.exe, follow the prompts, and you'll be booted into a full Omarchy desktop withi… | ↗Not briefed |
bluesky bluesky-ai | 1mo | this is where I've settled, basically, it's just not worth arguing about it outside of an actual academic philosophy context, and from an e… | this is where I've settled, basically, it's just not worth arguing about it outside of an actual academic philosophy context, and from… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Matched-Input Estimates Differ in Sign Across Architectures: Auditing EEG Foundation Models on Motor Imagery | Pretrained EEG foundation models are increasingly proposed as general-purpose encoders for brain-computer interfaces, yet recent bench… | ↗Not briefed |
bluesky bluesky-ai | 1mo | “Look, I did a google!” Someone came up to me at an event recently and said “I’ve been talking to ChatGPT…” and unspooled a convoluted an…🔧 ChatGPT | “Look, I did a google!” Someone came up to me at an event recently and said “I’ve been talking to ChatGPT…” and unspooled a convolut… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Isn't it the case that most of the thinking we know about, in humans and in non-linguistic animals, doesn't involve words at all? I guess I… | Isn't it the case that most of the thinking we know about, in humans and in non-linguistic animals, doesn't involve words at all? I gu… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Not silly. I do wonder how much of the insight comes from you though. The AI-LLM will say whatever, prompted by you. | Not silly. I do wonder how much of the insight comes from you though. The AI-LLM will say whatever, prompted by you. | ↗Not briefed |
arxiv arxiv-ai | 1mo | Increasing Skill Level Recruits Deeper Attention Layers in a Frozen Chess Transformer | Chess involves complex reasoning in a deterministic environment, which makes it a useful setting for studying the mechanisms of comput… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Time-Incremental Continued Pretraining of LLMs: Knowledge Updates Without Catastrophic Forgetting | Large language models (LLMs) drift out of date the moment their pretraining ends, yet retraining from scratch is prohibitively expensi… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Learning-Based 3D Reconstruction of Power Networks from Aerial Point Clouds | This paper presents an end-to-end framework for reconstructing overhead power utility network topology and extracting span-level physi… | ↗Not briefed |
bluesky bluesky-ai | 1mo | LLM-Skalierung macht aus Sprachmodellen immer bessere Kompressoren und Rekombinatoren vorhandenen Wissens, aber keine Erkenntnisarchitektur… | LLM-Skalierung macht aus Sprachmodellen immer bessere Kompressoren und Rekombinatoren vorhandenen Wissens, aber keine Erkenntnisarchit… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: jevals – replacing LLM judges with typed Jev decisions | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | The controversy isn't whether the AI could do the math very well--it's whether OpenAI stole *HEAVILY AI-assisted* work and then used their … | The controversy isn't whether the AI could do the math very well--it's whether OpenAI stole *HEAVILY AI-assisted* work and then used t… | ↗Not briefed |
arxiv arxiv-ai | 1mo | ReVeal: A Reconstruction-Aware Real-to-Sim Framework for VLA Policy Evaluation | Simulation-based evaluation provides a scalable and repeatable alternative to real-world evaluation of vision-language-action (VLA) po… | ↗Not briefed |
bluesky bluesky-ai | 1mo | if google managed to make their regular search work as well as it did in 2015 tomorrow, 60% of the valuation of llm companies would be eras… | if google managed to make their regular search work as well as it did in 2015 tomorrow, 60% of the valuation of llm companies would be… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Unregulated AI competition between the US and China threatens global infrastructure, prompting calls for US-China cooperation to manage the… | Unregulated AI competition between the US and China threatens global infrastructure, prompting calls for US-China cooperation to manag… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Reality monitoring LLM version | Reality monitoring LLM version | ↗Not briefed |
arxiv arxiv-ai | 1mo | A discrete generative model of neuronal spiking activity on microelectrode arrays | Generative models of neural activity could help characterize tissue dynamics, compare experimental conditions, and simulate population… | ↗Not briefed |
bluesky bluesky-ai | 1mo | If you use Grammarly and don't like AI, you should know that it trains its AI model on your writing. As long as you use their product there… | If you use Grammarly and don't like AI, you should know that it trains its AI model on your writing. As long as you use their product… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Multivariate quantile regression via Kolmogorov-Arnold Networks | This paper introduces a novel algorithm for predicting conditional joint distributions of vector-valued targets in stochastic systems… | ↗Not briefed |
bluesky bluesky-ai | 1mo | once told Claude i was a bimbo and after the second time i said it, it started yelling at me. Closest thing to genuine anger i’ve ever seen… | once told Claude i was a bimbo and after the second time i said it, it started yelling at me. Closest thing to genuine anger i’ve ever… | ↗Not briefed |
arxiv arxiv-ai | 1mo | GDN Tree-Scan: Served Tree Verification for Recurrent-Hybrid Language Models🔧 vLLM | Tree speculative decoding verifies multiple candidate continuations in one target forward pass. For attention-only transformers, the v… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Mini-Runbot – Ephemeral Odoo Enviroments Using Python and Docker | I build a Mini-Runbot, an open source PoC for creating ephemeral enviroments for Odoo from a Git commit.<p>This is a mini implementati… | ↗Not briefed |
bluesky bluesky-ai | 1mo | A threat intelligence report released by Anthropic this month identified non-state actors using Claude to develop weaponry, including a swa… | A threat intelligence report released by Anthropic this month identified non-state actors using Claude to develop weaponry, including… | ↗Not briefed |
bluesky bluesky-ai | 1mo | “but i talk to the llm and it can hold a conversation!” no, an llm quite literally cannot “hold a conversation.” that is an illusion bc y… | “but i talk to the llm and it can hold a conversation!” no, an llm quite literally cannot “hold a conversation.” that is an illusion… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Die Enshittification von LLM-Anbietern möge beginnen: www.buchodi.com/chatgpt-now-...🔧 ChatGPT | Die Enshittification von LLM-Anbietern möge beginnen: www.buchodi.com/chatgpt-now-... | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic says today's most advanced models are not reliable enough to autonomously select and attack targets. The company's attempts to im… | Anthropic says today's most advanced models are not reliable enough to autonomously select and attack targets. The company's attempts… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Good lord. www.politico.com/news/magazin... | Good lord. www.politico.com/news/magazin... | ↗Not briefed |
bluesky bluesky-ai | 1mo | Paddy Walker, an expert in AI ethics, states that modern warfare is characterized by an ambiguity that is difficult to codify or teach an L… | Paddy Walker, an expert in AI ethics, states that modern warfare is characterized by an ambiguity that is difficult to codify or teach… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Depressing thought: someone, somewhere, used an LLM to summarize your blog post rather than simply reading it :/ | Depressing thought: someone, somewhere, used an LLM to summarize your blog post rather than simply reading it :/ | ↗Not briefed |
hn hn-ai | 1mo | ComfyUI workflow that writes and performs a whole radio drama locally🔧 ComfyUI | — | ↗Not briefed |