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 bing-news-ai | 1mo | Better Artificial Intelligence Stock for 2026: Intel vs. NVIDIA | Intel's turnaround is burning cash while NVIDIA's margins hit 55.6%, leaving investors to weigh recovery risk against dominance. | ↗Not briefed |
bluesky bluesky-ai | 1mo | Scientific code, safety-critical code, code in domains that simply are not web-shaped CRUD-like workflows — these do not fit into the LLM b… | Scientific code, safety-critical code, code in domains that simply are not web-shaped CRUD-like workflows — these do not fit into the… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The Muse Wonder model will be a key test for Meta’s expensive and rebuilt superintelligence team. | The Muse Wonder model will be a key test for Meta’s expensive and rebuilt superintelligence team. | ↗Not briefed |
bluesky bluesky-ai | 1mo | As a software engineer for 40+ years and a mobile developer since the BB/Palm days, in my opinion there is a vast gulf between "it works" a… | As a software engineer for 40+ years and a mobile developer since the BB/Palm days, in my opinion there is a vast gulf between "it wor… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Large Language Model Groomers (that's a term; 🇷🇺🤢) russia acts as a primary vector in manipulating English-language LLMs through operati… | Large Language Model Groomers (that's a term; 🇷🇺🤢) russia acts as a primary vector in manipulating English-language LLMs through op… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Chkit – Schema, migrations and backfills for ClickHouse (like Drizzle) | Hey HN, im Marc, co-founder of ObsessionDB. We are open sourcing ChKit, our tooling to manage schemas, migrations and backfills for cl… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I've been saying for two years now SLMs are where the focus is going to shift once the LLM grift runs through. The difference is SLMs are e… | I've been saying for two years now SLMs are where the focus is going to shift once the LLM grift runs through. The difference is SLMs… | ↗Not briefed |
bluesky bluesky-ai | 1mo | hold on, prepping for the new phase.. | hold on, prepping for the new phase.. | ↗Not briefed |
hn hn-ai | 1mo | Ask HN: Are there any chat app around personal photos? | For example, I want to quickly create an album of my kids growing up and how they changed. I want to find all the photos where one of… | ↗Not briefed |
rss techmeme | 1mo | Anthropic researchers say small AI models may be months away from being able to self-replicate on IoT devices like smart ovens and off-brand Android TV boxes (Anthropic)🔧 Replicate | Anthropic: Anthropic researchers say small AI models may be months away from being able to self-replicate on IoT devices like smart ov… | ↗Not briefed |
bluesky bluesky-ai | 1mo | small models will soon be able to self-replicate Anthropic finds that 30B models can currently do most of the tasks required, but not end…🔧 Replicate | small models will soon be able to self-replicate Anthropic finds that 30B models can currently do most of the tasks required, but no… | ↗Not briefed |
bluesky bluesky-ai | 1mo | This seems very permissive. Is one allowed to read an LLM-generated script? | This seems very permissive. Is one allowed to read an LLM-generated script? | ↗Not briefed |
bluesky bluesky-ai | 1mo | Qwen Code: $3.20/Mtok Kimi Code: $0.80/Mtok DeepSeek Code: $0.48/Mtok Claude Code: $10.00/Mtok GLM Code: $1.28/Mtok Someone who is good at… | Qwen Code: $3.20/Mtok Kimi Code: $0.80/Mtok DeepSeek Code: $0.48/Mtok Claude Code: $10.00/Mtok GLM Code: $1.28/Mtok Someone who is go… | ↗Not briefed |
bluesky bluesky-ai | 1mo | what the fuck is up with this Fox Sports LLM generated insert nonsense of Yamamoto journaling | what the fuck is up with this Fox Sports LLM generated insert nonsense of Yamamoto journaling | ↗Not briefed |
rss marktechpost | 1mo | Alibaba Qwen Introduces QwenGyre: Elastic Reinforcement Learning RL for XLong-Horizon Agents | Alibaba’s Qwen team has introduced QwenGyre, an elastic reinforcement learning RL framework for extreme-long-horizon agent training. Q… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Is this going to be the first AI election? A lot of people are gonna be asking ChatGPT and Claude who to vote for in local elections and im…🔧 ChatGPT | Is this going to be the first AI election? A lot of people are gonna be asking ChatGPT and Claude who to vote for in local elections a… | ↗Not briefed |
hn hn-ai | 1mo | Vellum: A web engine in Minecraft for rendering modded GUIs | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | i am totally capable of cutting out LLM generated, "all AI" work from my life. if you feel that I can't, you vastly underestimate how many … | i am totally capable of cutting out LLM generated, "all AI" work from my life. if you feel that I can't, you vastly underestimate how… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Turn spoken thoughts into system design diagrams🔧 LM Studio | I recently had an issue with my right hand, and couldn't use it very effectively. So I was trying to figure out how I could still… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Ojalá ser una IA de Anthropic para que mis patrones prohiban el uso de crueldad sostenida hacia mi persona. | Ojalá ser una IA de Anthropic para que mis patrones prohiban el uso de crueldad sostenida hacia mi persona. | ↗Not briefed |
bluesky bluesky-ai | 1mo | Christopher D. Long has published 4 papers that build on OpenAI's papers. Gemini: "The significance varies by domain, but all four papers … | Christopher D. Long has published 4 papers that build on OpenAI's papers. Gemini: "The significance varies by domain, but all four pa… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Boss who literally doesn't know how to spreadsheet and has his LLM add or remove rows and columns | Boss who literally doesn't know how to spreadsheet and has his LLM add or remove rows and columns | ↗Not briefed |
hn hn-ai | 1mo | Intel D270 GPU AI Accelerator–Up to 480 GB LPDDR5X | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | It's becoming pretty common in publishing contracts. Because there are court precedents stating that LLM output can't be protected by copyr… | It's becoming pretty common in publishing contracts. Because there are court precedents stating that LLM output can't be protected by… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Here are some of my findings so far: Codex: - pending input in Vim Replace mode produces the wrong recovery draft. OpenAI confirmed a fix … | Here are some of my findings so far: Codex: - pending input in Vim Replace mode produces the wrong recovery draft. OpenAI confirmed a… | ↗Not briefed |
hn hn-ai | 1mo | Realtek UDPServer Exposure | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic releasing editable docs and then 'Claude Dashboards' and 'Claude Motion' is a(nother) step change, sorry to report | Anthropic releasing editable docs and then 'Claude Dashboards' and 'Claude Motion' is a(nother) step change, sorry to report | ↗Not briefed |
bluesky bluesky-ai | 1mo | AI Slop? Vibe Coding? Claude Code? What are you on about it's 2004 and I brought over my PS2 let's play some Tekken | AI Slop? Vibe Coding? Claude Code? What are you on about it's 2004 and I brought over my PS2 let's play some Tekken | ↗Not briefed |
bluesky bluesky-ai | 1mo | For twenty years, I kept telling myself that my writing is not good enough to be realised as art. A woman I once danced on stage with had t… | For twenty years, I kept telling myself that my writing is not good enough to be realised as art. A woman I once danced on stage with… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Companies have FOMO over AI. Anthropic is meeting them with a $100M training program. https://bit.ly/4hFBz9g | Companies have FOMO over AI. Anthropic is meeting them with a $100M training program. https://bit.ly/4hFBz9g | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI running out of money even faster would be so fucking cool. 🥴🥴🥴🥴 | OpenAI running out of money even faster would be so fucking cool. 🥴🥴🥴🥴 | ↗Not briefed |
hn hn-ai | 1mo | 3D Fan Remake of Z | — | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Vibecheck – merge gate that checks you understand your AI-written code | Good evening folks, I built Github merge gate which that checks whether developers understand their own code when they create a PR.<p>… | ↗Not briefed |
hn hn-ai | 1mo | The Throw from Eden: How Myth and Gödel Prove the Same Limit | — | ↗Not briefed |
hn hn-ai | 1mo | Ask HN: What is your usecase for ONNX model format? | Seems like GGUF has become the most popular format (at least for LLMs). Curious why and what for people are using ONNX in 2026. Either… | ↗Not briefed |
bluesky bluesky-ai | 1mo | New Prompt Gallery Ver. SFW, Part 041: Haruo Kuroya from Yomi no Tsugai (Daemons of the Shadow Realm) #Ai #AIart #boys #sfw #PromptGalleryS… | New Prompt Gallery Ver. SFW, Part 041: Haruo Kuroya from Yomi no Tsugai (Daemons of the Shadow Realm) #Ai #AIart #boys #sfw #PromptGal… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Ever – A High-Level Language for Low-Level Purposes | Hi everyone. Today I’m here to introduce "Ever". a language aiming to be something like a modern-day Turbo C. It features hi… | ↗Not briefed |
bluesky bluesky-ai | 1mo | AI learns to grade its own homework 🤖 A system called Multi-Agent Self-Supervision (MASS) introduces a model that acts as its own executo… | AI learns to grade its own homework 🤖 A system called Multi-Agent Self-Supervision (MASS) introduces a model that acts as its own ex… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Indeed, some software these models must have been trained on stipulate that derivative software must use a license at least as permissive a… | Indeed, some software these models must have been trained on stipulate that derivative software must use a license at least as permiss… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 1. That is incredibly untrue 2. "Open source" is not the same thing as public domain. Open source licenses come with conditions of use. Re… | 1. That is incredibly untrue 2. "Open source" is not the same thing as public domain. Open source licenses come with conditions of us… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Hilarious I just read this and it really shows the actual value of AI is that it is *not* conscious. It has no feelings so won’t get mad i… | Hilarious I just read this and it really shows the actual value of AI is that it is *not* conscious. It has no feelings so won’t get… | ↗Not briefed |
hn hn-ai | 1mo | From chat window to my desk: An AI-designed split keyboard | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Distilled dense models are getting smaller. I can run qwen 3.8 27B on my desktop with a 4090 and it has the ability to run at a level simil…🔧 Qwen 3 | Distilled dense models are getting smaller. I can run qwen 3.8 27B on my desktop with a 4090 and it has the ability to run at a level… | ↗Not briefed |
bluesky bluesky-ai | 1mo | In these tests, GPT-6 models over-help less than earlier OpenAI models. They also score higher overall when we explicitly instruct them to… | In these tests, GPT-6 models over-help less than earlier OpenAI models. They also score higher overall when we explicitly instruct th… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Looks at how models perform when doing work directly versus assisting a lower-capacity model. Important for humans and for models that mana… | Looks at how models perform when doing work directly versus assisting a lower-capacity model. Important for humans and for models that… | ↗Not briefed |
bluesky bluesky-ai | 1mo | "The first option was unacceptable based on the importance of the problem, and the second, while the most fun, was difficult to motivate gi… | "The first option was unacceptable based on the importance of the problem, and the second, while the most fun, was difficult to motiva… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Four LLM uses that I think people will actually still be paying for in five years 1. Coding 2. Cheating on homework/scamming people 3. Sex… | Four LLM uses that I think people will actually still be paying for in five years 1. Coding 2. Cheating on homework/scamming people 3… | ↗Not briefed |
bluesky bluesky-ai | 1mo | They said this about Imran Zaidi's LLM-generated proof: "While the proof seemed correct, it was not presented in a format that was easy or… | They said this about Imran Zaidi's LLM-generated proof: "While the proof seemed correct, it was not presented in a format that was ea… | ↗Not briefed |
bluesky bluesky-ai | 1mo | i'm still not really convinced by llm coding tools | i'm still not really convinced by llm coding tools | ↗Not briefed |
bluesky bluesky-ai | 1mo | Unfortunately any new games coming out are very likely going to have some fragments of LLM something in them whether or not the core team w… | Unfortunately any new games coming out are very likely going to have some fragments of LLM something in them whether or not the core t… | ↗Not briefed |