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 | www.theatlantic.com/technology/2... | www.theatlantic.com/technology/2... | ↗Not briefed |
hn hn-ai | 1mo | Florida AG seeks injunction to hamper OpenAI development | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | I trust the Pope more on philosophical matters than I would any AI model or any human who promotes AI. | I trust the Pope more on philosophical matters than I would any AI model or any human who promotes AI. | ↗Not briefed |
bluesky bluesky-ai | 1mo | Yep--as @authorizedmusings.bsky.social noted, it's probably because of LLM submissions. | Yep--as @authorizedmusings.bsky.social noted, it's probably because of LLM submissions. | ↗Not briefed |
hn hn-ai | 1mo | We diffed all 66 release pairs of the official MCP servers, 140 silent changes | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Rather alarming side note: David Robinson had been with OpenAI for three and a half years before resigning. He says he was “among the lon… | Rather alarming side note: David Robinson had been with OpenAI for three and a half years before resigning. He says he was “among th… | ↗Not briefed |
bluesky bluesky-ai | 1mo | “software with special tricks” (SWST) a better name than llm | “software with special tricks” (SWST) a better name than llm | ↗Not briefed |
bluesky bluesky-ai | 1mo | The AI minister has used AI-generated synthetic focus groups provided by a private company to test public opinion on AI and other issues T… | The AI minister has used AI-generated synthetic focus groups provided by a private company to test public opinion on AI and other issu… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The Vatican does have its own LLM model, trained on data from its own archives. (This is not a secret, Robert Ballecer SJ has talked about … | The Vatican does have its own LLM model, trained on data from its own archives. (This is not a secret, Robert Ballecer SJ has talked a… | ↗Not briefed |
bluesky bluesky-ai | 1mo | What I see a bit in your writing (admittedly skimming!) is that you may be treating a Markov chain n-gram model near-identically to an LLM.… | What I see a bit in your writing (admittedly skimming!) is that you may be treating a Markov chain n-gram model near-identically to an… | ↗Not briefed |
hn hn-ai | 1mo | EventMaxxer – Automatically apply to events to get you in the right room | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | That this guy is pointing to foundational culture rot more than hyper intelligent waifu robots with guns makes me give him a little more cr… | That this guy is pointing to foundational culture rot more than hyper intelligent waifu robots with guns makes me give him a little mo… | ↗Not briefed |
hn hn-ai | 1mo | Woman laid to rest in coffin made from mushrooms in Australian first | — | ↗Not briefed |
rss theverge-ai | 1mo | Capcom is preparing for a ‘future where we create games together with AI’ | Capcom's Pragmata might be all about the horrors of AI, but in practice the studio doesn't seem so down on the tech. During the Capcom… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I liked Ready Player One. Lots of people did. And then after Ready Player Two we realised we were reading vastly more meaning into it than … | I liked Ready Player One. Lots of people did. And then after Ready Player Two we realised we were reading vastly more meaning into it… | ↗Not briefed |
bluesky bluesky-ai | 1mo | “My office is asking OpenAI additional questions regarding cybersecurity incidents and risks involving the company and its AI models,” Bon… | “My office is asking OpenAI additional questions regarding cybersecurity incidents and risks involving the company and its AI models,… | ↗Not briefed |
rss techmeme | 1mo | Sources: ShinyHunters member Saif al-Din Khader, aka "Rey," was detained in Jordan and is cooperating to identify other hackers involved in the FBI breach (Reuters) | Reuters: Sources: ShinyHunters member Saif al-Din Khader, aka “Rey,” was detained in Jordan and is cooperating to identify other hacke… | ↗Not briefed |
hn hn-ai | 1mo | Why you do so much and get so little. Read this if you feel stuck | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | I'm not convinced that the entire archive of LLM training data contains even a single commit message. | I'm not convinced that the entire archive of LLM training data contains even a single commit message. | ↗Not briefed |
bluesky bluesky-ai | 1mo | See? Even Anthropic are telling you using AI makes you stupider. Come on. | See? Even Anthropic are telling you using AI makes you stupider. Come on. | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI safety employee resigns, claiming the company’s ‘culture is broken’ | OpenAI safety employee resigns, claiming the company’s ‘culture is broken’ | ↗Not briefed |
bluesky bluesky-ai | 1mo | I dropped $20 directly with deepseek after hearing people talk about how cheap it is and yeah when you don't fool around with the cache I'm… | I dropped $20 directly with deepseek after hearing people talk about how cheap it is and yeah when you don't fool around with the cach… | ↗Not briefed |
hn hn-ai | 1mo | You.com trains on your data by default with no opt out🔧 You.com | — | ↗Not briefed |
rss techcrunch-ai | 1mo | OpenAI safety employee resigns, claiming the company’s ‘culture is broken’ | By his own admission, David Robinson is “something of a cliché”: an employee at a leading AI company who issues a dire warning while r… | ↗Not briefed |
hn hn-ai | 1mo | Changes to Gemini model access and limits | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Sorry, using 'AI' 'Agents' for coding on your game counts just as much as generating visual or audio assets using an LLM. You're still stea… | Sorry, using 'AI' 'Agents' for coding on your game counts just as much as generating visual or audio assets using an LLM. You're still… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Pediram para eu avaliar um texto feito por LLM (Lero-Lero Machine). Dei uma lida e achei ruim. Devo mandar uma resposta gerada or LLM? #… | Pediram para eu avaliar um texto feito por LLM (Lero-Lero Machine). Dei uma lida e achei ruim. Devo mandar uma resposta gerada or LL… | ↗Not briefed |
bluesky bluesky-ai | 1mo | "The future depends on wisdom that Silicon Valley lacks" That's exactly right - a critical read on the future of AI (safety): www.theatla… | "The future depends on wisdom that Silicon Valley lacks" That's exactly right - a critical read on the future of AI (safety): www.th… | ↗Not briefed |
bluesky bluesky-ai | 1mo | This week: How Anthropic turned to religious leaders to create a “moral” agent; a former head of US intel worries about China misreading th… | This week: How Anthropic turned to religious leaders to create a “moral” agent; a former head of US intel worries about China misreadi… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Terminal Gym – a Claude mod that makes you do pushups between prompts | Hi HN, I built this for myself and wanted to open-source it.<p>The problem: I wanted a way to get reminders in-between prompts as I of… | ↗Not briefed |
bluesky bluesky-ai | 1mo | “The experiments are still in the queue. The PR is already live,” says one expert. | “The experiments are still in the queue. The PR is already live,” says one expert. | ↗Not briefed |
bluesky bluesky-ai | 1mo | After recent hacking incidents involving OpenAI, Narayan identified Britain's reliance on the US for access to frontier AI models as his “b… | After recent hacking incidents involving OpenAI, Narayan identified Britain's reliance on the US for access to frontier AI models as h… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 🧪 The late great Wisława Szymborska (Nobel laureate in Literature) when asked why she had published so few poems, said: “I have a trash ca… | 🧪 The late great Wisława Szymborska (Nobel laureate in Literature) when asked why she had published so few poems, said: “I have a tra… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: I made a browser-based instrument out of the CHOMPI CLUB source | A few days ago Chase Bliss open sourced the CHOMPI Club source code under an MIT license<p><a href="https://news.ycombinator… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Since you raise the issue of huge fallacies at the center of arguments, can you explain exactly how AI - especially the current LLM-derived… | Since you raise the issue of huge fallacies at the center of arguments, can you explain exactly how AI - especially the current LLM-de… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The vast majority of LLM output — in *any* field — will never be subjected to rigorous scrutiny. The result is an almost unimaginable proli… | The vast majority of LLM output — in *any* field — will never be subjected to rigorous scrutiny. The result is an almost unimaginable… | ↗Not briefed |
bluesky bluesky-ai | 1mo | ???????????????????? so you run two separate video games PLUS an AI, which would be insanely resource intensive, instead of just downloadin… | ???????????????????? so you run two separate video games PLUS an AI, which would be insanely resource intensive, instead of just downl… | ↗Not briefed |
hn hn-ai | 1mo | Being a good CPU neighbor (2016) | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | By tricking AI chatbots into ignoring their own rules, attackers are bypassing enterprise security with plain English. Here is how prompt i… | By tricking AI chatbots into ignoring their own rules, attackers are bypassing enterprise security with plain English. Here is how pro… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Even more depressing than the knowledge that congress is now overrun by effective altruist and overall TESCREAL cult member staffers, is th… | Even more depressing than the knowledge that congress is now overrun by effective altruist and overall TESCREAL cult member staffers,… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Monday at Devoxx Belgium, I'm giving a Tools in Action on the agent loop: bash, a tiny LLM, and the birth of a coding agent 🤖 m.devoxx.com… | Monday at Devoxx Belgium, I'm giving a Tools in Action on the agent loop: bash, a tiny LLM, and the birth of a coding agent 🤖 m.devox… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 2034: both digital and physical media are erased and history and culture exist only in LLM slop approximations. bsky.app/profile/merr... | 2034: both digital and physical media are erased and history and culture exist only in LLM slop approximations. bsky.app/profile/mer… | ↗Not briefed |
bluesky bluesky-ai | 1mo | this is the most maximal definition of an LLM's capabilities that I'm comfortable with making, and it does exclude some things that I think… | this is the most maximal definition of an LLM's capabilities that I'm comfortable with making, and it does exclude some things that I… | ↗Not briefed |
bluesky bluesky-ai | 1mo | boolean search or regular expressions are absolute king shit, and LLM-ification of search is a greasy stinky steaming poo | boolean search or regular expressions are absolute king shit, and LLM-ification of search is a greasy stinky steaming poo | ↗Not briefed |
bluesky bluesky-ai | 1mo | this is a very good/correct piece. still, the problem rn is that imposing & implementing legit, rights-oriented safety will not happen over… | this is a very good/correct piece. still, the problem rn is that imposing & implementing legit, rights-oriented safety will not happen… | ↗Not briefed |
bluesky bluesky-ai | 1mo | - if your inputs are tokenizable (words) - if your outputs are tokenizable (words) OR can be created by side effects which are (tool calls)… | - if your inputs are tokenizable (words) - if your outputs are tokenizable (words) OR can be created by side effects which are (tool c… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Repeat after me: mensen die tot oktober 2026 nodig hebben om tot de conclusie te komen dat de grote AI labs mogelijk niet het beste met de … | Repeat after me: mensen die tot oktober 2026 nodig hebben om tot de conclusie te komen dat de grote AI labs mogelijk niet het beste me… | ↗Not briefed |
hn hn-ai | 1mo | City building games have a Soul Problem pt.2 | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Não existe pair-programming com LLM. É outra coisa. É é um desserviço a uma prática valiosa usar o termo desta forma. | Não existe pair-programming com LLM. É outra coisa. É é um desserviço a uma prática valiosa usar o termo desta forma. | ↗Not briefed |
bluesky bluesky-ai | 1mo | An LLM deploys its "intellectual resources" to exactly one end, and in predictable fashion: in response to the input. The model *can* produ… | An LLM deploys its "intellectual resources" to exactly one end, and in predictable fashion: in response to the input. The model *can*… | ↗Not briefed |