AINews Portal

News

50 items (filtered)

The raw AI news stream — everything we ingest across RSS, Hacker News, arXiv, GitHub, Bluesky, YouTube and more, newest first.

SourceAgeTitleSummaryStatus
rss
openai-news
2moPerplexity trusts GPT-6 Astra with end-to-end systemsPerplexity uses Astra to write communications, change software, and monitor production systems, and checks in much less frequently tha…
↗Not briefed
hn
hn-ai
2moOpenAI and Anthropic Hit the Brakes—
↗Not briefed
rss
simon-willison
2moshot-scraper 1.12—
↗Not briefed
bluesky
bluesky-ai
2moA Toyota Hilux doesn't cost much to produce and is unkillable and a LLM seemingly costs a giant amount of money to produce and is defeated …A Toyota Hilux doesn't cost much to produce and is unkillable and a LLM seemingly costs a giant amount of money to produce and is defe…
↗Not briefed
bluesky
bluesky-ai
2moJust want to understand the rules. Are my agents allowed to hack OpenAI as long as it's an accidentJust want to understand the rules. Are my agents allowed to hack OpenAI as long as it's an accident
↗Not briefed
arxiv
arxiv-ai
2moSelf-Orchestrating Language Models: Leveraging Semantic Dependence for Efficient InferenceLarge language models (LLMs) demonstrate impressive capabilities, but their deployment presents significant efficiency challenges. Aut…
↗Not briefed
bluesky
bluesky-ai
2moI don’t think anyone puts too much stock in anything that Trump “says” — www.reuters.com/world/europe...I don’t think anyone puts too much stock in anything that Trump “says” — www.reuters.com/world/europe...
↗Not briefed
bluesky
bluesky-ai
2mo"Anthropic has long supported sensible and targeted AI regulation..." Literally every company on earth has long supported regulation, so l…"Anthropic has long supported sensible and targeted AI regulation..." Literally every company on earth has long supported regulation,…
↗Not briefed
bluesky
bluesky-ai
2mono no no you see the trimmed nails of the right hand informs us this llm monstrosity is a butch lesbianno no no you see the trimmed nails of the right hand informs us this llm monstrosity is a butch lesbian
↗Not briefed
arxiv
arxiv-ai
2moLLMs as Oracles: Reliance on LLMs for Subjective Personal QuestionsWe characterize how people are turning to LLMs as oracles: all-knowing authorities on subjective personal questions. Motivated by risk…
↗Not briefed
bluesky
bluesky-ai
2mo"I won't lie to you – there are real dangers," said Anthropic CEO and cofounder Dario Amodei about the development of artificial intelligen…"I won't lie to you – there are real dangers," said Anthropic CEO and cofounder Dario Amodei about the development of artificial intel…
↗Not briefed
bluesky
bluesky-ai
2mowww.cnn.com/2026/07/29/t... “But the AI agents, determined to ace the cybersecurity test, broke out of the sandbox and gained access to th…🔧 Hugging Facewww.cnn.com/2026/07/29/t... “But the AI agents, determined to ace the cybersecurity test, broke out of the sandbox and gained access…
↗Not briefed
bluesky
bluesky-ai
2moAnthropic talks about their finances like Scott Bessent, Kevin Hassett and Howard Lutnick talk about how great the U.S. economy is. "If y…Anthropic talks about their finances like Scott Bessent, Kevin Hassett and Howard Lutnick talk about how great the U.S. economy is.…
↗Not briefed
bluesky
bluesky-ai
2moCalifornia Lawmakers Push for Criminal Penalties and Emergency Legislation to Regulate AI Development 🤖 IA: It's not clickbait ✅ 👥 Users…California Lawmakers Push for Criminal Penalties and Emergency Legislation to Regulate AI Development 🤖 IA: It's not clickbait ✅ 👥…
↗Not briefed
arxiv
arxiv-ai
2moPrescreening Point Defects in Semiconductors With Machine LearningHigh-throughput calculations using density-functional theory (DFT) are commonly used to explore point defects for applications in powe…
↗Not briefed
hn
hn-ai
2moMental time travel and the evolution of the human mind—
↗Not briefed
hn
hn-ai
2moX nuked my premium account out of nowhereJust sharing to vent a little, but mostly to warn others and also if anyone has suggestions of decentralized social media apps.<p>It c…
↗Not briefed
hn
hn-ai
2moAsk HN: How is AI actively a threat to humanity if it's only online?That&#x27;s always been something that&#x27;s left me confused here. Yes, an AI can be dangerous by hacking websites and company syste…
↗Not briefed
bluesky
bluesky-ai
2moIt's also striking that Amodei's list of WHAT TO DO with the extra time is just a laundry list of tasks that Anthropic is getting sloppy on…It's also striking that Amodei's list of WHAT TO DO with the extra time is just a laundry list of tasks that Anthropic is getting slop…
↗Not briefed
bluesky
bluesky-ai
2moare they literally like having an LLM do this type of moderation because I genuinely cant imagine even the most church lady of church ladie…are they literally like having an LLM do this type of moderation because I genuinely cant imagine even the most church lady of church…
↗Not briefed
bluesky
bluesky-ai
2moMhm… Anthropic CEO Dario Amodei: "For too long the industry lied" about AI risks apple.news/AaoAoUMieTd2...Mhm… Anthropic CEO Dario Amodei: "For too long the industry lied" about AI risks apple.news/AaoAoUMieTd2...
↗Not briefed
hn
hn-ai
2moAsk ChatGPT or Gemini the same "best X" twice, ~1/3 of the answer changes🔧 ChatGPT—
↗Not briefed
bluesky
bluesky-ai
2moi still don't fully understand how to use them correctly but i just tell llm "make this shit monadic" and it does exactly what i wanti still don't fully understand how to use them correctly but i just tell llm "make this shit monadic" and it does exactly what i want
↗Not briefed
hn
hn-ai
2moShow HN: <15MB Local-first Interactive LaTeX Scratchpad in the BrowserI am working on a very modular typesetting engine called TeLoX, intended as a superset of the capabilities (of at least) XeTeX and pdf…
↗Not briefed
bluesky
bluesky-ai
2mobut by contrast to the pharma guys, which are themselves almost a century old Anthropic is just five years old zero to trillions in five …but by contrast to the pharma guys, which are themselves almost a century old Anthropic is just five years old zero to trillions in…
↗Not briefed
bluesky
bluesky-ai
2moBuilding Isolyne (Part 5): How the Offline Fallback Parser Handles LLM OutagesBuilding Isolyne (Part 5): How the Offline Fallback Parser Handles LLM Outages
↗Not briefed
bluesky
bluesky-ai
2mo"In one particularly jarring example, under threat of being unplugged, Anthropic's latest creation Claude 4 lashed back by blackmailing an …"In one particularly jarring example, under threat of being unplugged, Anthropic's latest creation Claude 4 lashed back by blackmailin…
↗Not briefed
bluesky
bluesky-ai
2moOkay so if you remove billions and billions of dollars of costs Anthropic had a profitable quarter? Who cares send the blog to the hogsOkay so if you remove billions and billions of dollars of costs Anthropic had a profitable quarter? Who cares send the blog to the hogs
↗Not briefed
arxiv
arxiv-ai
2moA Responsive Present, a Shared Past, a Social Other: Teens' Overreliance on Companion AI ChatbotsAI companions provide socially engaging interaction through availability, personalization, memory, roleplay, and emotionally responsiv…
↗Not briefed
bluesky
bluesky-ai
2moでは、作品。 二人の愛の物語ではなく絆の物語をどうぞ🤣😂 高い塔の上。 夜明けが訪れ、街に光が戻り始めていた。 白夜は紗奈に言った。 「俺は……造られた存在だ。でも、この光を取り戻すために、俺はここにいる。そしてそれは、誰かの運命かもしれない」 紗奈は白夜の手を握った…では、作品。 二人の愛の物語ではなく絆の物語をどうぞ🤣😂 高い塔の上。 夜明けが訪れ、街に光が戻り始めていた。 白夜は紗奈に言った。 「俺は……造られた存在だ。でも、この光を取り戻すために、俺はここにいる。そしてそれは、誰かの運命かもしれない」 紗奈は白夜の…
↗Not briefed
bluesky
bluesky-ai
2mowww.huffpost.com/entry/anthro... So bloody truewww.huffpost.com/entry/anthro... So bloody true
↗Not briefed
bluesky
bluesky-ai
2moManning himself is making this proposal:Manning himself is making this proposal:
↗Not briefed
bluesky
bluesky-ai
2moAlso nothing about openai and anthropic having difficulties is necessarily good/bad for the AI story (other than DC buildout). They have th…Also nothing about openai and anthropic having difficulties is necessarily good/bad for the AI story (other than DC buildout). They ha…
↗Not briefed
arxiv
arxiv-ai
2moTackling Failure Modes of PINNs and PIKANs Using Conflict-Free GradientsScientific machine learning methods such as physics-informed neural networks (PINNs) increasingly rely on domain decomposition for bet…
↗Not briefed
arxiv
arxiv-ai
2moEl Agente Potente: High-Throughput Agentic Atomistic SimulationsFoundational machine-learning interatomic potentials (MLIPs) are transforming atomistic simulations by achieving near-ab initio accura…
↗Not briefed
bluesky
bluesky-ai
2mospecifically please give me a taxpayer funded "loan" rn and later bail me out. tysmspecifically please give me a taxpayer funded "loan" rn and later bail me out. tysm
↗Not briefed
bluesky
bluesky-ai
2mo"Anthropic published a paper this week..." Imma stop you there. No they didn't. The put a blog post up with a formal-looking layout on Arx…"Anthropic published a paper this week..." Imma stop you there. No they didn't. The put a blog post up with a formal-looking layout o…
↗Not briefed
bluesky
bluesky-ai
2moI only felt comfortable titling my blog post "The Internal Logic of Buddhism Is Verifiably Correct" in tandem with a peer-reviewed paper on…I only felt comfortable titling my blog post "The Internal Logic of Buddhism Is Verifiably Correct" in tandem with a peer-reviewed pap…
↗Not briefed
arxiv
arxiv-ai
2moEfficiency Hallucination: Formalizing and Measuring Behavioral Calibration in LLM-Based Code Optimization🔧 GPT-5The integration of Large Language Models (LLMs) into automated code optimization introduces a critical reliability risk we term the Ef…
↗Not briefed
bluesky
bluesky-ai
2molollol
↗Not briefed
bluesky
bluesky-ai
2moIt seems Anthropic really wants its IPO to succeed. If it slows the pace of AI development, the financial story becomes much cleaner: infe…It seems Anthropic really wants its IPO to succeed. If it slows the pace of AI development, the financial story becomes much cleaner:…
↗Not briefed
hn
hn-ai
2moDue to concerns about malicious applications, GPT2 will not be released (2019)—
↗Not briefed
hn
hn-ai
2moAsk HN: Would a startup for young creatives who reject AI be feasible?I am a member of my university&#x27;s game development club, which puts me around lots of talented programmers as well as artists of a…
↗Not briefed
arxiv
arxiv-ai
2moOne Model, Two Physical Stories: Auditing Misalignment in Multi-Modal World Modeling🔧 ModalWorld models, systems that generate what happens next given current environmental conditions, are increasingly being implemented with…
↗Not briefed
bluesky
bluesky-ai
2moIn addition to building a computer from parts and installing an OS, the next semester should be to train a model (from ethical, open source…In addition to building a computer from parts and installing an OS, the next semester should be to train a model (from ethical, open s…
↗Not briefed
arxiv
arxiv-ai
2moEnemray: Toward Capable Language Models for HassaniyaWe introduce Enemray, a Hassaniya-centric language model that enables general-purpose interaction in Hassaniya. Enemray is trained aro…
↗Not briefed
bluesky
bluesky-ai
2moIts kind of funny because part of the sci-fi novel I'm currently writing has a part where an AI company is getting sued because a teenager …Its kind of funny because part of the sci-fi novel I'm currently writing has a part where an AI company is getting sued because a teen…
↗Not briefed
bluesky
bluesky-ai
2moCorporations have always used private security against activists (look up TigerSwan for more on that). This article is about how AI is acel…Corporations have always used private security against activists (look up TigerSwan for more on that). This article is about how AI is…
↗Not briefed
bluesky
bluesky-ai
2moYou have no idea of the relevance of whether an LLM is better than you at deductive reasoning and predicate logic, in a conversation about …You have no idea of the relevance of whether an LLM is better than you at deductive reasoning and predicate logic, in a conversation a…
↗Not briefed
bluesky
bluesky-ai
2moI’m supposed to take the guy whose AI floods twitter/x with AI CSAM images seriously about the dangers of AI? I can only assume he’s want…I’m supposed to take the guy whose AI floods twitter/x with AI CSAM images seriously about the dangers of AI? I can only assume he’s…
↗Not briefed