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 | ‘Uncertainty over the extent of its legal liabilities now hangs over Altman’s latest efforts to raise tens of billions in a private financi… | ‘Uncertainty over the extent of its legal liabilities now hangs over Altman’s latest efforts to raise tens of billions in a private fi… | ↗Not briefed |
rss bing-news-ai | 1mo | AI is being taught how to do your job. Will artificial intelligence replace us, or just change the way we work? | It's become a booming industry: teaching artificial intelligence how to get better at doing all kinds of jobs. As AI advances, many fe… | ↗Not briefed |
rss bing-news-ai | 1mo | AI is being taught how to do your job. Will artificial intelligence replace us, or just change the way we work? | It's become a booming industry: teaching artificial intelligence how to get better at doing all kinds of jobs. As AI advances, many fe… | ↗Not briefed |
rss bing-news-ai | 1mo | AI is being taught how to do your job. Will artificial intelligence replace us, or just change the way we work? | It's become a booming industry: teaching artificial intelligence how to get better at doing all kinds of jobs. As AI advances, many fe… | ↗Not briefed |
rss bing-news-ai | 1mo | AI is being taught how to do your job. Will artificial intelligence replace us, or just change the way we work? | It's become a booming industry: teaching artificial intelligence how to get better at doing all kinds of jobs. As AI advances, many fe… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The only argument I would make in defense of pirates here is that this is a fairly appropriate image to choose in that piracy was working f… | The only argument I would make in defense of pirates here is that this is a fairly appropriate image to choose in that piracy was work… | ↗Not briefed |
hn hn-ai | 1mo | Lockheed Taps OpenAI to Solve F-35 Challenges | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Hierarchical Reinforcement Learning with Stable Temporal Abstraction for Language Model Agents | Hierarchical reinforcement learning improves long-horizon control by organizing primitive actions around persistent subgoals and assig… | ↗Not briefed |
hn hn-ai | 1mo | An LLVM Pass for Automatic Skeletonization of MPI Applications | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Hallucination Across the Reasoning Lifecycle: Interface Visibility, Causal Evidence, and Release Control in Large Reasoning Models | Reasoning errors can propagate into later decisions and memory. This survey synthesizes 312 papers and first-party reports on text-bas… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Wow, holy crap. The current state of PPSSPP might be the perfect example of going off the deep end with AI slop in software dev. What I'm s… | Wow, holy crap. The current state of PPSSPP might be the perfect example of going off the deep end with AI slop in software dev. What… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I think it's only a matter of time that someone writes a piece of code into their fiction that will cause the LLM to choke if they ingest i… | I think it's only a matter of time that someone writes a piece of code into their fiction that will cause the LLM to choke if they ing… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Yeah having an LLM write all the rust for you and not actually learning it makes it very easy to violate the implicit contract between rust… | Yeah having an LLM write all the rust for you and not actually learning it makes it very easy to violate the implicit contract between… | ↗Not briefed |
bluesky bluesky-ai | 1mo | my ip got banned from the stackexchange sites and i'm pretty sure they think i'm an llm scraping. no!! i'm just a developer who is bad at p… | my ip got banned from the stackexchange sites and i'm pretty sure they think i'm an llm scraping. no!! i'm just a developer who is bad… | ↗Not briefed |
bluesky bluesky-ai | 1mo | we should start responding to AI art users by pretending we're one of them but shittalking them through the lens of FOMO. "what fucking mod… | we should start responding to AI art users by pretending we're one of them but shittalking them through the lens of FOMO. "what fuckin… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Dude at the bar is bragging about how he had Claude run a D&D game for a bunch of ai-generated “players” who then started their own “games”… | Dude at the bar is bragging about how he had Claude run a D&D game for a bunch of ai-generated “players” who then started their own “g… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Palantir built out a data system to track deportation targets. Now Anthropic is building a data system to track activists. Working people… | Palantir built out a data system to track deportation targets. Now Anthropic is building a data system to track activists. Working p… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Before teaching SI humanity, organizations must 1st remember how to treat humanity themselves. ―Robinson Clearly, we’re not there yet when… | Before teaching SI humanity, organizations must 1st remember how to treat humanity themselves. ―Robinson Clearly, we’re not there yet… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Writing as a Self-Organized Critical Process | We explain autocorrelation decay power laws omnipresent in texts by self-organized criticality. Specifically, we analyze the recently… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: [Squeegee] Auto-close stale windows on macOS, MIT | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Human-Like Attention? A Psychophysical Comparison of Visual Search in Humans and MLLMs🔧 Replicate | Visual search is a fundamental cognitive ability. This study investigates whether Multimodal Large Language Models (MLLMs) exhibit hum… | ↗Not briefed |
arxiv arxiv-ai | 1mo | AI Safety via Debate is Compromised by Cognitive Biases | Reinforcement learning from human feedback (RLHF) has played a central role in making large language models responsive to human instru… | ↗Not briefed |
bluesky bluesky-ai | 1mo | #AI #AI regulation #Anthropic #critics #Lobbying #Regulation #rhetoric #Washington 🌐 www.newsmart.ch | #AI #AI regulation #Anthropic #critics #Lobbying #Regulation #rhetoric #Washington 🌐 www.newsmart.ch | ↗Not briefed |
bluesky bluesky-ai | 1mo | good news: I have the perfect test case for my system 1 agent safety rig bad news: that perfect case is deepseek cheating on one of the ben… | good news: I have the perfect test case for my system 1 agent safety rig bad news: that perfect case is deepseek cheating on one of th… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Measuring and Reducing Cross-Vendor Mismatch in Language Models | Running the same language model on different graphics processing unit (GPU) vendors can produce different logits, even when the model… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Your next coworker might well be an AI agent—and will require a whole new model of workplace interactions. | Your next coworker might well be an AI agent—and will require a whole new model of workplace interactions. | ↗Not briefed |
bluesky bluesky-ai | 1mo | redditors when the llm says its “thinking” | redditors when the llm says its “thinking” | ↗Not briefed |
hn hn-ai | 1mo | Command injecting a reference tool to copy a source file | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Taylor Representations for Model-Free RL in Networked MDPs | In Networked Markov Decision Processes, transition dynamics are often unknown and the state--action space grows rapidly with the numbe… | ↗Not briefed |
hn hn-ai | 1mo | One-click mod installer and manager for DLSS 5 | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Of all the things to say about an emulator project dev using an llm i think this is the funniest | Of all the things to say about an emulator project dev using an llm i think this is the funniest | ↗Not briefed |
bluesky bluesky-ai | 1mo | "Absent policy steering, AI’s modeled effects increase the carbon intensity of the global economy and reinforce fossil fuel incumbency—outc… | "Absent policy steering, AI’s modeled effects increase the carbon intensity of the global economy and reinforce fossil fuel incumbency… | ↗Not briefed |
rss techmeme | 1mo | Sources: Apple and Google pitch state bills with self-attested age checks and no private lawsuits; Meta backs ones making app stores, not platforms, verify ages (Owen Dahlkamp/Politico) | Owen Dahlkamp / Politico: Sources: Apple and Google pitch state bills with self-attested age checks and no private lawsuits; Meta back… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Você não soube? A parceria do Brasil com a China rendeu a produção do Lulatron, o ciborgue que funciona a base de energia solar. Ele foi pr… | Você não soube? A parceria do Brasil com a China rendeu a produção do Lulatron, o ciborgue que funciona a base de energia solar. Ele f… | ↗Not briefed |
arxiv arxiv-ai | 1mo | The Poisoned Conversation: Privacy-Leaking Watermarks in Unified Multimodal Models | Multimodal models are increasingly shifting toward unified architectures that understand and generate text, images, and other modaliti… | ↗Not briefed |
bluesky bluesky-ai | 1mo | well i asked AI and claude told me i didnt kill your entire family | well i asked AI and claude told me i didnt kill your entire family | ↗Not briefed |
bluesky bluesky-ai | 1mo | We've written code that can pantomime human-like speech. I feel like it really diminishes humanity to equate that with being a person. This… | We've written code that can pantomime human-like speech. I feel like it really diminishes humanity to equate that with being a person.… | ↗Not briefed |
hn hn-ai | 1mo | A browser-native classic Visual Basic VB6 IDE | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | VERA: Verdict-Conditioned Reliability for Adaptive LLM Judges | Accurately estimating judgment reliability is a central challenge in adapting LLM judges to newly verified feedback while preserving p… | ↗Not briefed |
bluesky bluesky-ai | 1mo | fwiw the original plan for this was to use a local multimodal LLM w layers removed for image captioning, but after talking about it w @femm… | fwiw the original plan for this was to use a local multimodal LLM w layers removed for image captioning, but after talking about it w… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Groupwise Distortion Guarantees for Preference-Based Alignment | Preference-based alignment methods such as reinforcement learning from human feedback (RLHF) and Nash learning from human feedback (NL… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Just saw a reddit post of a guy who asked if someone could remake some old nes game for VR and 3 different peoples in the replies said "use… | Just saw a reddit post of a guy who asked if someone could remake some old nes game for VR and 3 different peoples in the replies said… | ↗Not briefed |
bluesky bluesky-ai | 1mo | "Eine KI, die sich selbst immer weiter verbessert, klingt nach Science-Fiction – laut dem Chefwissenschaftler von OpenAI ist sie es bald ni… | "Eine KI, die sich selbst immer weiter verbessert, klingt nach Science-Fiction – laut dem Chefwissenschaftler von OpenAI ist sie es ba… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Amiben béna kacsa vagyok: önmagam jellemezni. Amiről a gyereknek holnapra angol beadandót kell írnia: az apja. Aki írja a beadandót: az apj… | Amiben béna kacsa vagyok: önmagam jellemezni. Amiről a gyereknek holnapra angol beadandót kell írnia: az apja. Aki írja a beadandót: a… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The OP should go read the bug tracker for the Bun Rust rewrite, which is full of SEGVs, because the LLM just translated the whole thing to … | The OP should go read the bug tracker for the Bun Rust rewrite, which is full of SEGVs, because the LLM just translated the whole thin… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: DoubtBench – test whether Jev knows when humans disagree | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | "Bevor die KI-Unternehmen einer Superintelligenz beibringen, wie man die Menschheit gut behandelt, müssen sie sich zuerst daran erinnern, d… | "Bevor die KI-Unternehmen einer Superintelligenz beibringen, wie man die Menschheit gut behandelt, müssen sie sich zuerst daran erinne… | ↗Not briefed |
bluesky bluesky-ai | 1mo | “An environment where things like this can happen is no place to grow artificial minds that could be smarter than we are and that might not… | “An environment where things like this can happen is no place to grow artificial minds that could be smarter than we are and that migh… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Transformers Explained Visually | Transformers Explained Visually | ↗Not briefed |
bluesky bluesky-ai | 1mo | real answer: so you understand the formal model at a deep and intuitive level granting you the ability to compose the right formal concepts… | real answer: so you understand the formal model at a deep and intuitive level granting you the ability to compose the right formal con… | ↗Not briefed |