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 |
|---|---|---|---|---|
bluesky bluesky-ai | 1mo | futurism.com/artificial-i... | futurism.com/artificial-i... | ↗Not briefed |
hn hn-ai | 1mo | OpenAI tricked me into switching to a business account | It's my fault for agreeing (I didn't realize what I was clicking) to move from personal to business account, but now I have… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OMG someone used AI to completely decoded and create all the Adobe Apps for free. I’m so not going to download and pirate off a failing com… | OMG someone used AI to completely decoded and create all the Adobe Apps for free. I’m so not going to download and pirate off a failin… | ↗Not briefed |
arxiv arxiv-ai | 1mo | FedAlphaEdit: Null-Space-Aligned Merging for Collaborative Knowledge Editing | Multiple institutions may each hold their own private knowledge edits and wish to integrate them into a single large language model wi… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Language Modeling is Monotone Compression | A long-standing hypothesis in artificial intelligence and neuroscience posits that intelligence is closely related to compression: the… | ↗Not briefed |
arxiv arxiv-ai | 1mo | NOMOS: Compiling Written Policies into Statically Verified Tool-Call Gates for LLM Agents | Tool-using LLM agents violate the policies they are deployed to enforce, often silently. Prior defenses hand-write rules, query an LLM… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I hated doing proofs in math and OpenAI should suffer for making me think about them again | I hated doing proofs in math and OpenAI should suffer for making me think about them again | ↗Not briefed |
arxiv arxiv-ai | 1mo | Intent Graph: Navigating the Analytical Reasoning Space for Exploratory Data Analysis | Exploratory data analysis (EDA) is rarely open-ended in practice: analysts work from high-level domain questions toward the concrete a… | ↗Not briefed |
hn hn-ai | 1mo | AI cluster networking reading list: RDMA, collectives, fabrics | — | ↗Not briefed |
hn hn-ai | 1mo | Zpu | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | A Graph Neural Network for Global Daily Fire Radiative Power Prediction at Medium-Range Lead Times | Skillful prediction of biomass-burning activity several days in advance is important for air-quality forecasting and aerosol predictio… | ↗Not briefed |
bluesky bluesky-ai | 1mo | You can write by hand at an entirely LLM-free workplace, and it'll make you enormously more productive. A very smart thing to do rn is to f… | You can write by hand at an entirely LLM-free workplace, and it'll make you enormously more productive. A very smart thing to do rn is… | ↗Not briefed |
bluesky bluesky-ai | 1mo | alongside it is largely garbage, openai still has generated a lean proof of finite-time NS blowup. that's not nothing, that's still a marve… | alongside it is largely garbage, openai still has generated a lean proof of finite-time NS blowup. that's not nothing, that's still a… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Letter: three fired OpenAI researchers urge AI labs to halt work that could impair AI monitoring and say their firings are "chilling those … | Letter: three fired OpenAI researchers urge AI labs to halt work that could impair AI monitoring and say their firings are "chilling t… | ↗Not briefed |
rss techmeme | 1mo | Letter: three fired OpenAI researchers urge AI labs to halt work that could impair AI monitoring and say their firings are "chilling those who remain at OpenAI" (Maxwell Zeff/Wall Street Journal) | Maxwell Zeff / Wall Street Journal: Letter: three fired OpenAI researchers urge AI labs to halt work that could impair AI monitoring a… | ↗Not briefed |
hn hn-ai | 1mo | AHM Statement on OpenAI's October 6 Release of Mathematical Documents | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Mid-Training Language Models on Raw Video | Multimodal large language models learn mostly from paired image-text data or annotated video, and raw web video is rarely used to furt… | ↗Not briefed |
bluesky bluesky-ai | 1mo | instead of a deluge of slop papers credited to some "internal model", only some of which correspond to lean proofs, what openai should be p… | instead of a deluge of slop papers credited to some "internal model", only some of which correspond to lean proofs, what openai should… | ↗Not briefed |
rss huggingface-blog | 1mo | The model that didn't exist, so you made it yourself | — | ↗Not briefed |
rss openai-news | 1mo | Disrupting AI-enabled “false front” operations | OpenAI disrupted two AI-enabled influence operations that used false-front journalists and a think tank to spread geopolitical messagi… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 1 - it's funny to hear these students talking to an LLM and sound just like it. 2 - it's also funny hearing people talk about capital and… | 1 - it's funny to hear these students talking to an LLM and sound just like it. 2 - it's also funny hearing people talk about capita… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 🎁 link Fucking AI, jfc 🙄 claude, google maps, all trails, and other “tools” nearly killed the lad | 🎁 link Fucking AI, jfc 🙄 claude, google maps, all trails, and other “tools” nearly killed the lad | ↗Not briefed |
bluesky bluesky-ai | 1mo | I'm more or less fine with local models with some baseline ethics around training data (e.g. opt-in via licensing terms) for personal proje… | I'm more or less fine with local models with some baseline ethics around training data (e.g. opt-in via licensing terms) for personal… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Prompts versus Rules: Auditing and Controlling Speech Naturalness Behaviors in Voice User Simulators | As voice agents gain more popularity commercially, the user simulators used to evaluate the deployed agents are also being developed t… | ↗Not briefed |
hn hn-ai | 1mo | Goose: Safer, faster and <memory than Rust/C++ for algos where flat data matters | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | I did something insanely ballsy. I emailed OpenAI asking them to run insanely hard math on the SUPER SECRET MATH MODEL | I did something insanely ballsy. I emailed OpenAI asking them to run insanely hard math on the SUPER SECRET MATH MODEL | ↗Not briefed |
arxiv arxiv-ai | 1mo | Distillation for Incrimination and Distillation for Capabilities | Powerful misaligned AI models might recognize alignment evaluations and strategically behave well on them, rendering direct audits uni… | ↗Not briefed |
bluesky bluesky-ai | 1mo | There are some students in my library having some LLM talk to them about the value workers contribute to a company, and how to graph the ca… | There are some students in my library having some LLM talk to them about the value workers contribute to a company, and how to graph t… | ↗Not briefed |
bluesky bluesky-ai | 1mo | US and Aussie parts of the site not fully editorially independent so they're still impacted by, for example, their lucrative 'strategic par… | US and Aussie parts of the site not fully editorially independent so they're still impacted by, for example, their lucrative 'strategi… | ↗Not briefed |
bluesky bluesky-ai | 1mo | ai generated math papers associated with lean certs might be better understood simply as context that helped steer an llm to produce the le… | ai generated math papers associated with lean certs might be better understood simply as context that helped steer an llm to produce t… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Why LLM Agents Favor Their Group: Stakes, Observed Norms, and Reputation | Language-model agents favor their own group because they have watched their members favor each other. The group label alone does littl… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I don’t know enough about math to say anything novel and useful about the OpenAI math dump. But I am looking forward to seeing what exper… | I don’t know enough about math to say anything novel and useful about the OpenAI math dump. But I am looking forward to seeing what… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Curating Always-Loaded Context for LLM Agents: A Capacitated Assortment Model with Censored Feedback | At the start of every session, LLM agents load a fixed context file, such as $\texttt{AGENTS.md}$. Each loaded token in the file is ch… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I feel like OpenAI could do a little bit better job acting like they give any amount of a shit whatsoever about the results they are publis… | I feel like OpenAI could do a little bit better job acting like they give any amount of a shit whatsoever about the results they are p… | ↗Not briefed |
arxiv arxiv-ai | 1mo | How Narrative Wrapping Affects LLM Refusal: A Cross-Language Benchmark and Defense | Safety-aligned language models often refuse a harmful request stated directly but answer the same request inside a role-play or narrat… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic did announce some sort of wet lab program 😅😩 | Anthropic did announce some sort of wet lab program 😅😩 | ↗Not briefed |
hn hn-ai | 1mo | I built a ridiculous game to test decision models vs. 9 LLMs | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Reconstruction of Multiscale Plasma Dynamics Across Operating Regimes | Reconstructing spatially resolved plasma dynamics from few sensors is essential for diagnostics, reduced-order modelling and control,… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Work wants a local LLM for opencode, etc… eh, whatever, fine. Spin up an ollama container, import a qwen 3.8 GGUF, ask it “hello, who are …🔧 Ollama🔧 Qwen 3 | Work wants a local LLM for opencode, etc… eh, whatever, fine. Spin up an ollama container, import a qwen 3.8 GGUF, ask it “hello, who… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Low-rank tensor structure of precipitation and its application to satellite-reference merging | The intermittent and variable nature of precipitation makes its accurate estimation over extended domains difficult, yet its spatiotem… | ↗Not briefed |
rss techmeme | 1mo | Samsung reports preliminary Q3 operating profit up 782% YoY to ~$80.1B, vs. ~$81.2B est., and revenue up 127% to ~$145.7B, lower than expected (Yoolim Lee/Bloomberg) | Yoolim Lee / Bloomberg: Samsung reports preliminary Q3 operating profit up 782% YoY to ~$80.1B, vs. ~$81.2B est., and revenue up 127%… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anyone who's read technical documentation generated by LLM knows exactly this play | Anyone who's read technical documentation generated by LLM knows exactly this play | ↗Not briefed |
arxiv arxiv-ai | 1mo | F$^3$NO: Frequency-Decomposed Finite-Time Flow-map Neural Operators with Cross-Scale Conditioning | Neural operators enable fast PDE forecasting, but repeated predictions accumulate errors and fine-scale structures remain difficult to… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Invariant-Measure Reasoners: Stable Representations for Latent Reasoning | Latent reasoning models repeatedly update a latent state using the same recurrent block. As the recurrent depth increases, the sequenc… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Beijing measures the success of its AI rollout in terms of adoption and diffusion across its economy and beyond. This has a key safety impl… | Beijing measures the success of its AI rollout in terms of adoption and diffusion across its economy and beyond. This has a key safety… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I don't know if I talked about it here, but I let my newsletter peeps know the reasoning--Amazon, like the other grifting companies with LL… | I don't know if I talked about it here, but I let my newsletter peeps know the reasoning--Amazon, like the other grifting companies wi… | ↗Not briefed |
bluesky bluesky-ai | 1mo | “Calling everything existential may be provocative but obscures the more thoughtful policy and model development work that will be necessar… | “Calling everything existential may be provocative but obscures the more thoughtful policy and model development work that will be nec… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Region-Aware CLS Token Augmentation for Fine-Grained Image Retrieval | Image retrieval methods often rely on a single global semantic descriptor extracted from an image, e.g., the [CLS] token in vision tra… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Omni-Diffusion-Distill: Few-Step Distillation of Unified Multimodal Diffusion Large Language Models | Unified multimodal diffusion large language models (dLLMs) offer a single architecture for both image generation and multimodal unders… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Multi-Bandwidth Distribution Matching Distillation: On the Equivalence of Distribution Matching Distillation and Drifting Models | Researchers are exploring effective one-step generative model continuously, and, Drifting Models (Deng et al., 2026), demonstrate grea… | ↗Not briefed |