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 |
|---|---|---|---|---|
rss techmeme | 2mo | Cohere and Aleph Alpha sign a definitive merger agreement; the combined company will operate as Cohere, with Aleph Alpha co-CEO Ilhan Scheer becoming Cohere COO (Leo Marchandon/Reuters)🔧 Cohere | Leo Marchandon / Reuters: Cohere and Aleph Alpha sign a definitive merger agreement; the combined company will operate as Cohere, with… | ↗Not briefed |
hn hn-ai | 2mo | GPT-6 Astra Controlling Blender for VFX | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | I haven’t seen much on how companies like OpenAI are planning for the liability shitstorm of their rogue AI agents committing felonies. If … | I haven’t seen much on how companies like OpenAI are planning for the liability shitstorm of their rogue AI agents committing felonies… | ↗Not briefed |
hn hn-ai | 2mo | AI must not outrun safety controls, DeepMind co-founder warns | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | this means it's a mistake to see them do a task that takes a 160 iq human, and surmise that what you have is a 160 iq llm. you might! you a… | this means it's a mistake to see them do a task that takes a 160 iq human, and surmise that what you have is a 160 iq llm. you might!… | ↗Not briefed |
hn hn-ai | 2mo | Introducing the DeepMind Institute | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | www.cnn.com/2026/09/16/b... These researchers fear for the day when AI systems can improve themselves without human input, which is known a… | www.cnn.com/2026/09/16/b... These researchers fear for the day when AI systems can improve themselves without human input, which is kn… | ↗Not briefed |
bluesky bluesky-ai | 2mo | OpenAI challenges secrecy of Apple pact with Musk’s X and SpaceXAI reut.rs/4hymDsn | OpenAI challenges secrecy of Apple pact with Musk’s X and SpaceXAI reut.rs/4hymDsn | ↗Not briefed |
bluesky bluesky-ai | 2mo | For those who didn't spend the early 2010s being terminally online, here's a decent explainer of what the hell we're talking about. techcr… | For those who didn't spend the early 2010s being terminally online, here's a decent explainer of what the hell we're talking about. t… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Preventing Model Collapse: A Fisher-Rao Perspective on the Dynamics of Training with Synthetic Data | Large Language Models (LLMs) are now routinely trained using synthetic data, since high-quality human data has been exhausted by the e… | ↗Not briefed |
bluesky bluesky-ai | 2mo | I still think the gop will try to run an LLM based on trump if It Happens. There will be lots of Facebook posts about trump speaking throu… | I still think the gop will try to run an LLM based on trump if It Happens. There will be lots of Facebook posts about trump speaking… | ↗Not briefed |
hn hn-ai | 2mo | Show HN: KiwiDesk – Tiling that feels like it shipped with macOS | Hi HN, I'm Maikel, a solo dev from Germany. I've been building KiwiDesk, a tiling window manager which feels like it shipped… | ↗Not briefed |
hn hn-ai | 2mo | Show HN: Stopping AI agent swarms from hacking our systems🔧 Hugging Face | I don't trust OpenAI, Anthropic, Google, or any other AI lab to be responsible for protecting my systems from the systems they&#x… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Du moins c'est ce qu'on suppose, qu'ils s'agit d'acteurs qui indexent pour leurs LLM. Sur l'analyse du trafic du jour sur cartes.app, c'e… | Du moins c'est ce qu'on suppose, qu'ils s'agit d'acteurs qui indexent pour leurs LLM. Sur l'analyse du trafic du jour sur cartes.app… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Aujourd'hui, déjà les moteurs de recherche, en proposant directement un "résumé IA", ne ramènent plus d'audience sur les sites. Mais beauc… | Aujourd'hui, déjà les moteurs de recherche, en proposant directement un "résumé IA", ne ramènent plus d'audience sur les sites. Mais… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Interesting discussion of how Google’s competitor, Anthropic, is TRAINING its Claude models with parameters including the possibility of co… | Interesting discussion of how Google’s competitor, Anthropic, is TRAINING its Claude models with parameters including the possibility… | ↗Not briefed |
bluesky bluesky-ai | 2mo | www.cnbc.com/2026/09/16/o... “Slow down we are going to kill humanity” was meant to provide cover to postpone IPO. Now suddenly this- new … | www.cnbc.com/2026/09/16/o... “Slow down we are going to kill humanity” was meant to provide cover to postpone IPO. Now suddenly this-… | ↗Not briefed |
bluesky bluesky-ai | 2mo | My favorite thing about AI is that it's the only industry where its CEOs are constantly putting out statements saying "Our product could do… | My favorite thing about AI is that it's the only industry where its CEOs are constantly putting out statements saying "Our product cou… | ↗Not briefed |
hn hn-ai | 2mo | OpenAI weighs funding round at $1.2T valuation before IPO | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | If you are building Frankenstein, stop, because he was not a nice man. If however you're building the monster, that's actually okay beca… | If you are building Frankenstein, stop, because he was not a nice man. If however you're building the monster, that's actually okay… | ↗Not briefed |
hn hn-ai | 2mo | Show HN: Round 1 of a vibe-coder ladder, judged by an LLM that only sees the UI | — | ↗Not briefed |
bluesky bluesky-ai | 2mo | “AIs are not conscious. They do not feel, experience, or suffer. They do not have innate preferences or underlying motivations. They are se… | “AIs are not conscious. They do not feel, experience, or suffer. They do not have innate preferences or underlying motivations. They a… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Mozilla gaat samenwerken met Mistral, een Europese AI-ontwikkelaar. Mistral Small 4 wordt voortaan een van de modellen voor de ingebouwde A… | Mozilla gaat samenwerken met Mistral, een Europese AI-ontwikkelaar. Mistral Small 4 wordt voortaan een van de modellen voor de ingebou… | ↗Not briefed |
arxiv arxiv-ai | 2mo | ASLEval: Measuring Privacy Exposure Displacement in LLM Agent Sessions | Privacy evaluations of tool-using LLM agents often inspect a designated action, final response, or attacker report. These local proxie… | ↗Not briefed |
bluesky bluesky-ai | 2mo | One of the most valuable ways humans can contribute in a time of rising AI is by being a source of quality judgment. You see this in tons o… | One of the most valuable ways humans can contribute in a time of rising AI is by being a source of quality judgment. You see this in t… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Physics-based prediction, uncertainty quantification and decision-making for IN718 crystallographic texture intensity across LPBF defocus regimes | Reliable prediction of crystallographic texture in laser powder bed fusion is critical for linking process conditions with anisotropic… | ↗Not briefed |
hn hn-ai | 2mo | An Introduction to HTML Streaming for Interactive Web Applications | — | ↗Not briefed |
rss simon-willison | 2mo | Quoting Mustafa Suleyman | — | ↗Not briefed |
arxiv arxiv-ai | 2mo | PersonaPath: Towards Knowledge-Centric Personalized Learning Path Planning | Adaptive learning systems commonly formulate learning path planning as Exercise-Centric (EC) recommendation, where the next step is in… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Decodable but Misrouted: Sparse Features Uncover a Readout Gap in Vision-Language Models for Harmful Meme Detection | When a large vision-language model misclassifies a harmful meme, the failure may reflect missing internal evidence or an inability to… | ↗Not briefed |
rss openai-news | 2mo | Helping older adults use AI in everyday life🔧 ChatGPT | OpenAI and AARP are bringing free, hands-on ChatGPT workshops to 1,000 older adults across 10 U.S. cities to build practical AI skills… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Taming the Agentic RAN: Stability-Guaranteed Arbitration of Autonomous AI Agents in O-RAN | The O-RAN control plane is becoming agentic: autonomous AI agents, deployed as rApps by different vendors, independently close control… | ↗Not briefed |
arxiv arxiv-ai | 2mo | GrainSpeech: Less Context, More Detail for Compact Speech Synthesis | Compact acoustic models face a challenging quality-capacity trade-off. We investigate two factors in this regime: encoder context and… | ↗Not briefed |
bluesky bluesky-ai | 2mo | They're currently typing paragraphs in my replies. It's like someone prompted an LLM to act offended then accuse everyone replying of bein… | They're currently typing paragraphs in my replies. It's like someone prompted an LLM to act offended then accuse everyone replying of… | ↗Not briefed |
bluesky bluesky-ai | 2mo | this piece highlights three key points: (1) trump admin actively opposes AI safety reg; (2) zuck & others essentially say trust-but-don't… | this piece highlights three key points: (1) trump admin actively opposes AI safety reg; (2) zuck & others essentially say trust-but-… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Greens MSP Laura Moodie warns Scotland could become 'a digital equivalent of the old extractive model' if ministers fail to set strong rule… | Greens MSP Laura Moodie warns Scotland could become 'a digital equivalent of the old extractive model' if ministers fail to set strong… | ↗Not briefed |
bluesky bluesky-ai | 2mo | so i'm gonna get negatively polarized by chuds into being an Anthropic shill then, huh? | so i'm gonna get negatively polarized by chuds into being an Anthropic shill then, huh? | ↗Not briefed |
arxiv arxiv-ai | 2mo | EviGen: Predictive Evidence Scaffolding for Verifiable Clinical Rationale Generation | Longitudinal electronic health records (EHRs) capture years of patient history across notes, codes, labs, and procedures, and contain… | ↗Not briefed |
hn hn-ai | 2mo | Is Anthropic the New Netscape? | — | ↗Not briefed |
arxiv arxiv-ai | 2mo | Ask the Tool, Don't Guess: Agent Tool Calls Hold Their Progress, and the Serving System Should Read It | An agentic request spends substantial wall-clock time waiting for tools, and its KV cache holds GPU memory the whole time. Serving sys… | ↗Not briefed |
bluesky bluesky-ai | 2mo | "If you're building Frankenstein, stop." Sure, I am building an early 19th century German scientist who will then build a monster. Because… | "If you're building Frankenstein, stop." Sure, I am building an early 19th century German scientist who will then build a monster. Be… | ↗Not briefed |
bluesky bluesky-ai | 2mo | AI chatbots developed a secret language that baffled humans, study says | AI chatbots developed a secret language that baffled humans, study says | ↗Not briefed |
bluesky bluesky-ai | 2mo | I don't want regulation of AI (written by OpenAI and Anthropic, naturally). I want to unplug it. | I don't want regulation of AI (written by OpenAI and Anthropic, naturally). I want to unplug it. | ↗Not briefed |
arxiv arxiv-ai | 2mo | ReFigBench: Benchmarking Scientific Figure Reconstruction as Editable PowerPoint Artifacts | Multimodal coding agents are expected to turn visual inputs into usable artifacts, and they act through a harness, the layer of tools,… | ↗Not briefed |
bluesky bluesky-ai | 2mo | when someone says an LLM can think they're telling you how they believe thinking works, because it's how they believe their own thoughts wo… | when someone says an LLM can think they're telling you how they believe thinking works, because it's how they believe their own though… | ↗Not briefed |
bluesky bluesky-ai | 2mo | Searl Effect Generator animated model in Google AI Studio. This controversial energy & propulsion drive was invented by John R. R. Searl in… | Searl Effect Generator animated model in Google AI Studio. This controversial energy & propulsion drive was invented by John R. R. Sea… | ↗Not briefed |
arxiv arxiv-ai | 2mo | Infinite-Parameter LLMs: Generating and Adapting Weights from Live Data | The scaling laws hold that a language model grows more capable with more parameters and more training data, and Mixture-of-Experts (Mo… | ↗Not briefed |
hn hn-ai | 2mo | Red-teaming my own credential broker: 231 attacks, and the five that worked | — | ↗Not briefed |
hn hn-ai | 2mo | Show HN: Put an AI agent on a FaceTime audio/video call (open source, WebRTC) | Hey everyone, Gary here from Chert (YC P26). We just open-sourced a framework for putting AI agents on FaceTime audio/video calls… | ↗Not briefed |
rss techmeme | 2mo | Noetive, which is developing an industrial AI model for businesses in physical industries, emerges from stealth with a $41M seed led by Eclipse (Sarah Klearman/Wall Street Journal) | Sarah Klearman / Wall Street Journal: Noetive, which is developing an industrial AI model for businesses in physical industries, emerg… | ↗Not briefed |