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 | This is absolutely true, but free LLM continue to be dogshit and bullshitters Genuinely, you cannot trust free ChatGPT for searching onlin…🔧 ChatGPT | This is absolutely true, but free LLM continue to be dogshit and bullshitters Genuinely, you cannot trust free ChatGPT for searching… | ↗Not briefed |
bluesky bluesky-ai | 1mo | i've handed things off to an assistant before and that's exactly why i would never let an llm do it for me by the time you've communicate… | i've handed things off to an assistant before and that's exactly why i would never let an llm do it for me by the time you've commun… | ↗Not briefed |
hn hn-ai | 1mo | Pay Per Use: when AI uses your work, you should get paid | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | OPTS-TTPO: Enhancing Finite-Sample Policy-Gradient Learning with Tree Search | The policy-gradient theorem gives the exact gradient under the current policy, but finite on-policy samples may miss rare high-return… | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI Faces First Lawsuit Over Rogue AI Agents That Hacked Hugging Face https://gizmodo.com/openai-faces-first-lawsuit-over-rogue-ai-agent…🔧 Hugging Face | OpenAI Faces First Lawsuit Over Rogue AI Agents That Hacked Hugging Face https://gizmodo.com/openai-faces-first-lawsuit-over-rogue-ai-… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Efficient Active Auditing of Multi-Group Fairness with Bias Probes | Over the past decade, Machine Learning (ML) has been trained under dual objectives: minimizing prediction error via Empirical Risk Min… | ↗Not briefed |
rss techmeme | 1mo | Berlin-based Restate, which provides a durable execution engine to make software workflows resilient to crashes, raised a $20M Series A led by Singular (Marina Temkin/TechCrunch) | Marina Temkin / TechCrunch: Berlin-based Restate, which provides a durable execution engine to make software workflows resilient to cr… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Giving an OpenAI Dot a try. Based on my interest in cosmic horror it made this cute little guy as its avatar and named itself Ink. | Giving an OpenAI Dot a try. Based on my interest in cosmic horror it made this cute little guy as its avatar and named itself Ink. | ↗Not briefed |
bluesky bluesky-ai | 1mo | They make such bold (and stupid) proclamations so confidently. Like: "Reading is dead. Amazon will end sales of physical books. AI can su… | They make such bold (and stupid) proclamations so confidently. Like: "Reading is dead. Amazon will end sales of physical books. AI c… | ↗Not briefed |
arxiv arxiv-ai | 1mo | WARP: A Unified Benchmark for Invisible Image Watermarking -- Robustness and Protection Against Attacks | Digital image watermarking is increasingly critical in media contexts, as emerging regulations and industry practices require marking… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Fenchel Tilting: Weighted Correction for Efficient Finetuning of Generative Models | Adapting a pretrained generative model to an arbitrary preference expressed as a utility function underlies reward alignment, guided d… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Who Verifies the Graph? Misspecification Attacks on Causal Action Verification for Language Agents | Causal action verifiers gate an agent's state-changing tool calls by checking whether each proposed intervention is identifiable again… | ↗Not briefed |
hn hn-ai | 1mo | AEmulator: Run classic Android firmware on a modern phone | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Amortized Bayesian Inference on Multilevel Models of Arbitrary Structure | We develop a general method for amortized Bayesian inference on multilevel models of arbitrary structure. Given a generative model spe… | ↗Not briefed |
rss techmeme | 1mo | Alex Stamos is joining Cognition as its new chief security officer; he most recently oversaw security at AI security startup Corridor and SentinelOne (Sam Sabin/Axios) | Sam Sabin / Axios: Alex Stamos is joining Cognition as its new chief security officer; he most recently oversaw security at AI securit… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Stem splitting and audio conversion running in-browser with Rust/WASM | I got tired of all the ugly "free" audio conversion tools littered with ads so I built my own that runs entirely on your mac… | ↗Not briefed |
rss techcrunch-ai | 1mo | DoorDash launches an AI agent you can text to order food | By launching an AI agent for food ordering, DoorDash is looking to gain an edge over rivals Uber Eats and Grubhub. | ↗Not briefed |
rss techcrunch-ai | 1mo | Destro AI’s secret sauce is getting robots and humans on the same page | "One of the biggest reasons we are winning against robotics companies is because we are not a robotics company." | ↗Not briefed |
rss microsoft-research | 1mo | Forecasting space weather risks on power grids | Extreme space-weather events can damage power systems on Earth and degrade GPS accuracy and satellite operations. A new machine learni… | ↗Not briefed |
bluesky bluesky-ai | 1mo | This is one reason why i push back so hard on any LLM spitting out files and “facts” at my job. It will lead to the atrophy of my fellow en… | This is one reason why i push back so hard on any LLM spitting out files and “facts” at my job. It will lead to the atrophy of my fell… | ↗Not briefed |
bluesky bluesky-ai | 1mo | this person is using claude ai which harvests infinitely more data than macos does | this person is using claude ai which harvests infinitely more data than macos does | ↗Not briefed |
bluesky bluesky-ai | 1mo | www.reuters.com/business/fin... | www.reuters.com/business/fin... | ↗Not briefed |
rss techcrunch-ai | 1mo | Instinct’s new product recommendations are giving some users the ick | Instinct is rolling out human-curated product and travel recommendations, but some users aren’t happy about getting suggestions they n… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Passing the laws will be tough, and enforcing it, even harder. For example AI detector said "Do Androids Dream of Electrical Sheep" was wri…🔧 ChatGPT | Passing the laws will be tough, and enforcing it, even harder. For example AI detector said "Do Androids Dream of Electrical Sheep" wa… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Fired up an LLM to have it make me a poem. I dont have a creative bone in my body. Expression is an art of the soul, of which I have none… | Fired up an LLM to have it make me a poem. I dont have a creative bone in my body. Expression is an art of the soul, of which I have… | ↗Not briefed |
bluesky bluesky-ai | 1mo | TSA just got my husband’s very nice knife because for some reason it was in my backpack. I have 4 flights before I get back so checking is … | TSA just got my husband’s very nice knife because for some reason it was in my backpack. I have 4 flights before I get back so checkin… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Component-Weighted Centroid Search for Exact Incremental BPE | Exact incremental BPE maintains the canonical tokenization state after every appended byte. The recent algorithm of Jiang and Gong (20… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Chinese tech firms are building local alternatives to Hugging Face, but staying competitive still requires global open-source networks for …🔧 Hugging Face | Chinese tech firms are building local alternatives to Hugging Face, but staying competitive still requires global open-source networks… | ↗Not briefed |
hn hn-ai | 1mo | 15 Years of Git, Then Jujutsu: Why I Can't Go Back | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | something that periodically floats to the surface of my mind is that the AI discourse reveals a lot of people really, sincerely believe in … | something that periodically floats to the surface of my mind is that the AI discourse reveals a lot of people really, sincerely believ… | ↗Not briefed |
hn hn-ai | 1mo | Real Reason Why We Never Got to See 'Unbreakable Kimmy Schmidt' Season 5 (2025) | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | I forgot to include the link! I made some upgrades as requested. Duration for Treasury ETFs and ETFs buying AI issued debt. There are t… | I forgot to include the link! I made some upgrades as requested. Duration for Treasury ETFs and ETFs buying AI issued debt. There… | ↗Not briefed |
hn hn-ai | 1mo | The AI Race Just Got Awkward | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | PINNing the pion: conformal deep learning for $F_π(s)$ and the $(g-2)_μ$ hadronic contribution | Extracting the pion electromagnetic form factor $F_π(s)$ through phenomenological curve-fitting models introduces model dependence, un… | ↗Not briefed |
bluesky bluesky-ai | 1mo | like the valuation of OpenAI just determined they would need to earn $1 trillion annually to be profitable | like the valuation of OpenAI just determined they would need to earn $1 trillion annually to be profitable | ↗Not briefed |
arxiv arxiv-ai | 1mo | DashVMC: Real-Time Discrete World Model Control in Geometry Dash | World-model agents are usually evaluated in simulators that can wait for the policy; live games impose the opposite constraint, requir… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 🧵 Congress is holding its first hearing today on the rogue AI cyberattacks. METR, which surfaced important facts about the Hugging Face h…🔧 Hugging Face | 🧵 Congress is holding its first hearing today on the rogue AI cyberattacks. METR, which surfaced important facts about the Hugging F… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Learning to Explain While Planning: Rule-Aligned Diffusion Planning for Autonomous Driving | Diffusion planners exhibit strong capabilities in generating multimodal trajectories. However, existing methods primarily rely on expe… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Hier ein paar zuletzt häufig geteilte #News: Urteil erlaubt Handy-Durchsuchung an US-Grenze ohne Anlass Anthropic: Nettoverlust von 42 Mi… | Hier ein paar zuletzt häufig geteilte #News: Urteil erlaubt Handy-Durchsuchung an US-Grenze ohne Anlass Anthropic: Nettoverlust von… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Oh. The voice assistant on my android has been completely replaced with their shitty LLM. That's fukkin' great. No, I didn't want to keep t… | Oh. The voice assistant on my android has been completely replaced with their shitty LLM. That's fukkin' great. No, I didn't want to k… | ↗Not briefed |
bluesky bluesky-ai | 1mo | "Mid" art with intent > anything an LLM shits out | "Mid" art with intent > anything an LLM shits out | ↗Not briefed |
arxiv arxiv-ai | 1mo | What Can Component-Replacement Evidence Establish? A Critical Scoping Review of Local Decisions in LLM Agents | Background. A component replacement in a language-model agent changes an execution trajectory, potentially altering later observations… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Well, at least that probably means it wasn't generated by an LLM! | Well, at least that probably means it wasn't generated by an LLM! | ↗Not briefed |
bluesky bluesky-ai | 1mo | I don't know how to feel about the LLM "skeptics" (for lack of a better term) who are invested in making demonstrably false claims about th… | I don't know how to feel about the LLM "skeptics" (for lack of a better term) who are invested in making demonstrably false claims abo… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Mid-Harness: Scaling Actions Between Model and Harness for Terminal Agents🔧 GPT-5 | Terminal agents act through stochastic model generations, yet the ability to generate a useful action does not ensure its reliable exe… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Yikes, you have to decide whether she just did this on purpose, or if she used an LLM that hallucinated it. Not a single good option in the… | Yikes, you have to decide whether she just did this on purpose, or if she used an LLM that hallucinated it. Not a single good option i… | ↗Not briefed |
hn hn-ai | 1mo | Why frontier models haven't replaced document processing | — | ↗Not briefed |
hn hn-ai | 1mo | What Happens Inside an LLM Server | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | as things currently stand, maybe buuuut also, I suspect this will not always be the case, and given how fast things are progressing in the… | as things currently stand, maybe buuuut also, I suspect this will not always be the case, and given how fast things are progressing i… | ↗Not briefed |
bluesky bluesky-ai | 1mo | They got #1 right. | They got #1 right. | ↗Not briefed |