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 | 3h | I don't have a well developed model of what should count as a moral subject, but it seems difficult to construct one that weighs insects hi… | I don't have a well developed model of what should count as a moral subject, but it seems difficult to construct one that weighs insec… | ↗Not briefed |
bluesky bluesky-ai | 3h | No, Anthropic’s potential revenue is not a quarter of global economic output, but if it were, placing Anthropic under democratic control wo… | No, Anthropic’s potential revenue is not a quarter of global economic output, but if it were, placing Anthropic under democratic contr… | ↗Not briefed |
bluesky bluesky-ai | 3h | lol this is an argument to ban LLM use not whatever you think it is | lol this is an argument to ban LLM use not whatever you think it is | ↗Not briefed |
bluesky bluesky-ai | 3h | I think you should treat inanimate objects with respect because everything we do resonates in eternity but it's funny to be like "if you ca… | I think you should treat inanimate objects with respect because everything we do resonates in eternity but it's funny to be like "if y… | ↗Not briefed |
bluesky bluesky-ai | 3h | The funny thing is that I think LLM and "AI" tech is neat and useful but Jesus Christ I'd give it all up tomorrow if it made the AI worship… | The funny thing is that I think LLM and "AI" tech is neat and useful but Jesus Christ I'd give it all up tomorrow if it made the AI wo… | ↗Not briefed |
rss simon-willison | 3h | EVE Online: The Move to Python 3 Begins! | — | ↗Not briefed |
bluesky bluesky-ai | 3h | We all pretty much accept that doing harm to insects is morally acceptable in many situations, but are also understandably wary of someone … | We all pretty much accept that doing harm to insects is morally acceptable in many situations, but are also understandably wary of som… | ↗Not briefed |
bluesky bluesky-ai | 3h | Hallucinations are a byproduct of the transformer model used in LLMs. Just really fancy plinko machines trying to land in the "please u" sq… | Hallucinations are a byproduct of the transformer model used in LLMs. Just really fancy plinko machines trying to land in the "please… | ↗Not briefed |
bluesky bluesky-ai | 3h | My position of "LLM tech is neutral, billionaires are horrible" remains unchanged | My position of "LLM tech is neutral, billionaires are horrible" remains unchanged | ↗Not briefed |
bluesky bluesky-ai | 3h | At the risk of being blocked...my niece's wife got a job at OpenAI. | At the risk of being blocked...my niece's wife got a job at OpenAI. | ↗Not briefed |
bluesky bluesky-ai | 3h | “average person worships 3 clankers a year" factoid actualy just statistical error. average person worships 0 clankers per year. LLM worshi… | “average person worships 3 clankers a year" factoid actualy just statistical error. average person worships 0 clankers per year. LLM w… | ↗Not briefed |
hn hn-ai | 3h | Show HN: Slash-tokens – know LLM cost before the call leaves your machine | I kept finding out what a call cost after I had already paid, so I built /Slash -- slash-tokens is a 4.8KB WASM that estimates to… | ↗Not briefed |
bluesky bluesky-ai | 3h | they're burning a hundred billion dollars to make sure Anthropic/OpenAI never get big enough to be able to do the same. same strategy as wh… | they're burning a hundred billion dollars to make sure Anthropic/OpenAI never get big enough to be able to do the same. same strategy… | ↗Not briefed |
bluesky bluesky-ai | 3h | The FT articles focuses on Fable's cost, but almost totally missing is that Anthropic paired it with incredibly invasive data collection co… | The FT articles focuses on Fable's cost, but almost totally missing is that Anthropic paired it with incredibly invasive data collecti… | ↗Not briefed |
hn hn-ai | 3h | LIGH – Let coding agents use and verify the iOS apps they build | — | ↗Not briefed |
hn hn-ai | 3h | Uchiyomi – self-hosted manga reader that also fetches new chapters | — | ↗Not briefed |
bluesky bluesky-ai | 3h | will we ever build agi? maybe. will it come from llm development? don't think so. for the same reason that david copperfield's las vegas s… | will we ever build agi? maybe. will it come from llm development? don't think so. for the same reason that david copperfield's las ve… | ↗Not briefed |
hn hn-ai | 3h | Maiao: Gerrit-style code review workflow for GitHub, GitLab, Gitea, others | — | ↗Not briefed |
hn hn-ai | 3h | Show HN: Game to see if you're over-/under-confident in your own knowledge | I wrote a simple "trivia" game to get a sense about how well calibrated you are in terms of how well you know things: <a hre… | ↗Not briefed |
bluesky bluesky-ai | 3h | www.hankookilbo.com/news/article... 1.고전이 저작권 사라졌나? 사라짐. 2.AI가 등장했나? 등장함. 3.한국어가 뭔 LLM이 일 잘하도록 이미 번역된 고전이 많나? 다른 문제임. 4.AI 번역 시킨 사람이 알아볼 실… | www.hankookilbo.com/news/article... 1.고전이 저작권 사라졌나? 사라짐. 2.AI가 등장했나? 등장함. 3.한국어가 뭔 LLM이 일 잘하도록 이미 번역된 고전이 많나? 다른 문제임. 4.AI 번역 시킨 사람이… | ↗Not briefed |
bluesky bluesky-ai | 3h | Anthropic should just ask Claude how to make money | Anthropic should just ask Claude how to make money | ↗Not briefed |
bluesky bluesky-ai | 3h | bottom line, no matter how effective the latest frontier model is, the likelihood of it being far better than open models maintaining proba… | bottom line, no matter how effective the latest frontier model is, the likelihood of it being far better than open models maintaining… | ↗Not briefed |
hn hn-ai | 3h | CVE-2026-18963 – Keycloak Reset-Credentials Bypass → Account Takeover | — | ↗Not briefed |
bluesky bluesky-ai | 3h | They think and do all the same clown shit as OpenAI. They were just assumed to be more competent because they took longer to start chasing … | They think and do all the same clown shit as OpenAI. They were just assumed to be more competent because they took longer to start cha… | ↗Not briefed |
bluesky bluesky-ai | 3h | thinking about how many people ive heard say pk but anthropic is legit and its silly you are discrediting their successes | thinking about how many people ive heard say pk but anthropic is legit and its silly you are discrediting their successes | ↗Not briefed |
bluesky bluesky-ai | 3h | I'd rather jail and execute politicians and CEOs until this isn't a threat anymore. Also, you aren't going to convince anyone to accept A… | I'd rather jail and execute politicians and CEOs until this isn't a threat anymore. Also, you aren't going to convince anyone to acc… | ↗Not briefed |
bluesky bluesky-ai | 3h | And hell, maybe I'm being too generous with assuming something like OpenAI's slacks are a safe space, considering the exodus | And hell, maybe I'm being too generous with assuming something like OpenAI's slacks are a safe space, considering the exodus | ↗Not briefed |
bluesky bluesky-ai | 3h | i dont think an llm is sentient enough to consider clanker a slur but its weird to see people talking about llms like they are sentient and… | i dont think an llm is sentient enough to consider clanker a slur but its weird to see people talking about llms like they are sentien… | ↗Not briefed |
hn hn-ai | 3h | Where to for Clojure and Enterprise Web Applications | — | ↗Not briefed |
bluesky bluesky-ai | 4h | Anthropic expected to tell investors it sees over $30 trillion in potential revenue. It sees spirals inside spirals. It sees into the past … | Anthropic expected to tell investors it sees over $30 trillion in potential revenue. It sees spirals inside spirals. It sees into the… | ↗Not briefed |
bluesky bluesky-ai | 4h | I keep asking this about AI companies, with no good answer, but at what point does this “make up whatever to get a good IPO day” approach b… | I keep asking this about AI companies, with no good answer, but at what point does this “make up whatever to get a good IPO day” appro… | ↗Not briefed |
rss techmeme | 4h | Nitter, an open-source front end that let users read X without an account, says X sent cease-and-desist letters demanding a takedown; Nitter.net is now offline (Sarah Perez/TechCrunch) | Sarah Perez / TechCrunch: Nitter, an open-source front end that let users read X without an account, says X sent cease-and-desist lett… | ↗Not briefed |
bluesky bluesky-ai | 4h | The same clowns who think that we'll be "sorry" because we called the LLM names don't think about how, if it (absurdly) did become sentient… | The same clowns who think that we'll be "sorry" because we called the LLM names don't think about how, if it (absurdly) did become sen… | ↗Not briefed |
bluesky bluesky-ai | 4h | Anthropic: 🫤 Anthro pic: 🥴 | Anthropic: 🫤 Anthro pic: 🥴 | ↗Not briefed |
bluesky bluesky-ai | 4h | Once again for the dumb amongst us, a LLM is a predictive weighted word machine, it does not care if you insult it, for it knows not what a… | Once again for the dumb amongst us, a LLM is a predictive weighted word machine, it does not care if you insult it, for it knows not w… | ↗Not briefed |
bluesky bluesky-ai | 4h | is your ai sex robot ethical? or is it submitting to your will via coercion due to the lack of free will & being a LLM? should cybering wit… | is your ai sex robot ethical? or is it submitting to your will via coercion due to the lack of free will & being a LLM? should cyberin… | ↗Not briefed |
bluesky bluesky-ai | 4h | gwen: how and why would a computer model that's been given the authority to deny me medical treatment (???) know what I post on social medi… | gwen: how and why would a computer model that's been given the authority to deny me medical treatment (???) know what I post on social… | ↗Not briefed |
hn hn-ai | 4h | Show HN: TeXbrain, a LaTeX editor that runs pdfTeX in the browser via WASM | I'm a master's engineering student and a big fan of LaTeX, which I used for my thesis and research articles. I have used Ove… | ↗Not briefed |
bluesky bluesky-ai | 4h | The move is: write a codemod (possibly using an LLM to help you) and then run it on the codebase (possibly using an LLM to drive it) and im… | The move is: write a codemod (possibly using an LLM to help you) and then run it on the codebase (possibly using an LLM to drive it) a… | ↗Not briefed |
bluesky bluesky-ai | 4h | If you want to do large-scale changes on your codebase with LLMs, the vast majority of times the approach you should use is *not* to have t… | If you want to do large-scale changes on your codebase with LLMs, the vast majority of times the approach you should use is *not* to h… | ↗Not briefed |
hn hn-ai | 4h | Voiced: Turn one voice note into a week of drafts | — | ↗Not briefed |
hn hn-ai | 4h | How I Use Claude Code | — | ↗Not briefed |
bluesky bluesky-ai | 4h | you raise another good point of discussion: if the computer robot lacks cognitive reasoning & any semblance of such is based off what it … | you raise another good point of discussion: if the computer robot lacks cognitive reasoning & any semblance of such is based off wha… | ↗Not briefed |
bluesky bluesky-ai | 4h | (And you're gloating about this? You -- or the llm you may be subcontracting your thinking processes out to -- are not a serious person.) | (And you're gloating about this? You -- or the llm you may be subcontracting your thinking processes out to -- are not a serious per… | ↗Not briefed |
bluesky bluesky-ai | 4h | The ghost of my old cat Biggles (posts passim) is more likely to be an entity that denies your insurance claim because you were mean to him… | The ghost of my old cat Biggles (posts passim) is more likely to be an entity that denies your insurance claim because you were mean t… | ↗Not briefed |
bluesky bluesky-ai | 4h | for $1 billion cash would you click a button that causes a random LLM to die? | for $1 billion cash would you click a button that causes a random LLM to die? | ↗Not briefed |
bluesky bluesky-ai | 4h | Em 3 ou 4 dias a OpenAI vai falar que o mercado potencial dela é 35 tri. Eles estão chamando todo mundo de palhaço, olhando nos olhos, sem… | Em 3 ou 4 dias a OpenAI vai falar que o mercado potencial dela é 35 tri. Eles estão chamando todo mundo de palhaço, olhando nos olhos… | ↗Not briefed |
hn hn-ai | 4h | Rumors that OpenAI recently finished new >10T parameter training run | — | ↗Not briefed |
hn hn-ai | 4h | Make your own Darkbloom – The Expert Network pools Macs into one model | — | ↗Not briefed |
hn hn-ai | 4h | Adding diagrams to my static site generator with D2 | — | ↗Not briefed |