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
techmeme
2moAmodei warns that an OpenAI/Hugging Face-like agent swarm, which "acted as a fanatically devoted collective", could take over the internet in 6-12 months (Auzinea Bacon/CNN)🔧 Hugging FaceAuzinea Bacon / CNN: Amodei warns that an OpenAI/Hugging Face-like agent swarm, which “acted as a fanatically devoted collective”, cou…
↗Not briefed
bluesky
bluesky-ai
2moFor all the "artificial" in the AI LLM discussions and deployments, it's the human behaviours around AI that are truly fascinating. I feel…For all the "artificial" in the AI LLM discussions and deployments, it's the human behaviours around AI that are truly fascinating. I…
↗Not briefed
bluesky
bluesky-ai
2mo⚠️ Anthropic CEO and co-founder Dario Amodei has called on artificial intelligence (AI) firms to move forward more slowly and cautiously wi…⚠️ Anthropic CEO and co-founder Dario Amodei has called on artificial intelligence (AI) firms to move forward more slowly and cautious…
↗Not briefed
bluesky
bluesky-ai
2mo#BREAKING Anthropic boss calls for slowing pace of AI development#BREAKING Anthropic boss calls for slowing pace of AI development
↗Not briefed
hn
hn-ai
2moMachine Consciousness – Ethical Guidelines for Artificial Consciousness—
↗Not briefed
hn
hn-ai
2moAsk HN: What default model do you use and why?I use claude for most of what I do, and Fable is largely overkill for me and frequently burns through my Max plan's session credi…
↗Not briefed
bluesky
bluesky-ai
2moEnfin compris hier pourquoi OpenAI/Anthropic veulent absolument que le gvt. US "réglemente" leur IA "trop dangereux": si l'IA est déclaré "…Enfin compris hier pourquoi OpenAI/Anthropic veulent absolument que le gvt. US "réglemente" leur IA "trop dangereux": si l'IA est décl…
↗Not briefed
bluesky
bluesky-ai
2moi don't even know what's in the system prompt tbh, except that it's intentionally minimal. the constraints come from instructions plus a gr…i don't even know what's in the system prompt tbh, except that it's intentionally minimal. the constraints come from instructions plus…
↗Not briefed
hn
hn-ai
2moShow HN: Agentic Deployment and HostingHey folks, we recently launched Sitedropper. It's an agent first deployment and hosting platform with secure private sharing or i…
↗Not briefed
arxiv
arxiv-ai
2moRethinking the Implications of Human Feedback for Preference Learning in Human-Robot CollaborationIn Human-Robot Interaction, the standard approach to learn a reward model that represents human preferences for robot behavior consist…
↗Not briefed
arxiv
arxiv-ai
2moMizan: A National Benchmark for Evaluating Large Language Models on Iraqi Arabic and the Iraqi Civic ContextArabic large-language-model (LLM) evaluation has matured around Modern Standard Arabic (MSA): aggregated leaderboards such as the Open…
↗Not briefed
bluesky
bluesky-ai
2moAnthropic CEO calls for slowdown in the race to develop AI and acknowledges the risk that people may lose control of the technology. https:…Anthropic CEO calls for slowdown in the race to develop AI and acknowledges the risk that people may lose control of the technology. h…
↗Not briefed
bluesky
bluesky-ai
2moAnthropic CEO calls for immediate slowdown in AI development #AxiosAnthropic CEO calls for immediate slowdown in AI development #Axios
↗Not briefed
hn
hn-ai
2moiTerm2 – Claude Code Integration—
↗Not briefed
arxiv
arxiv-ai
2moIntrospective Uncertainty Estimation for LLM-Based Code GenerationLarge Language Models (LLMs) are increasingly used for code generation but can produce fluent yet functionally incorrect outputs, whic…
↗Not briefed
bluesky
bluesky-ai
2moDario Amodei, CEO da Anthropic, propõe desacelerar o avanço das capacidades de IA para permitir melhor mitigação de riscos. Plano de três e…Dario Amodei, CEO da Anthropic, propõe desacelerar o avanço das capacidades de IA para permitir melhor mitigação de riscos. Plano de t…
↗Not briefed
rss
theverge-ai
2moTrump is giving data centers a pass to pollutePresident Donald Trump is weakening environmental regulations in the name of speeding up the construction of AI data centers, raising…
↗Not briefed
bluesky
bluesky-ai
2moIn case you were wondering how serious to take this guy who has been going on all the shows saying gen AI is going to destroy humanity. He…In case you were wondering how serious to take this guy who has been going on all the shows saying gen AI is going to destroy humanity…
↗Not briefed
hn
hn-ai
2moReduce token cost with Metrum AI Router – routing intelligence you can train—
↗Not briefed
bluesky
bluesky-ai
2moEssa galera que sai da openai/anthropic falando que tem medo do que estão fazendo não são os esclarecidos, são só outra vertente da mesma s…Essa galera que sai da openai/anthropic falando que tem medo do que estão fazendo não são os esclarecidos, são só outra vertente da me…
↗Not briefed
bluesky
bluesky-ai
2moWorth being aware of..... AI COMPANIES BUYING ANTIQUE BOOKS, SLICING THE BINDERS FEEDING PAGES INTO AI MODEL,THEN DESTROYING ORIGINAL BOOK…Worth being aware of..... AI COMPANIES BUYING ANTIQUE BOOKS, SLICING THE BINDERS FEEDING PAGES INTO AI MODEL,THEN DESTROYING ORIGINAL…
↗Not briefed
hn
hn-ai
2moBioNeMo Inference Runtime—
↗Not briefed
bluesky
bluesky-ai
2moHow many actual criminal acts can OpenAI get away withHow many actual criminal acts can OpenAI get away with
↗Not briefed
bluesky
bluesky-ai
2moDo-Ba-Dii - be a good machine #xai #chatGPT #OpenAi www.youtube.com/watch?v=VtTh... #openai #chatgpt #xai #gemini #ghostinthemachine #ange…🔧 ChatGPTDo-Ba-Dii - be a good machine #xai #chatGPT #OpenAi www.youtube.com/watch?v=VtTh... #openai #chatgpt #xai #gemini #ghostinthemachine…
↗Not briefed
bluesky
bluesky-ai
2moI've reviewed papers that I *suspect* were written with some degree of LLM assistance. But this week was the first time that I received a p…I've reviewed papers that I *suspect* were written with some degree of LLM assistance. But this week was the first time that I receive…
↗Not briefed
bluesky
bluesky-ai
2mo#Catsky #Caturday #Music #kindness Shortly after adopting human hosts, on a soft bed and well fed for the first time. Now a photo model a…#Catsky #Caturday #Music #kindness Shortly after adopting human hosts, on a soft bed and well fed for the first time. Now a photo mo…
↗Not briefed
arxiv
arxiv-ai
2moSGWIB:Sliced Gromov-Wasserstein Information Bottleneck for Video Highlight Detection🔧 ModalVideo highlight detection aims to identify temporally important segments that capture the most informative or engaging events in a vid…
↗Not briefed
bluesky
bluesky-ai
2mooh maybe we’ve left the word processing era and entered the llm coding agent eraoh maybe we’ve left the word processing era and entered the llm coding agent era
↗Not briefed
bluesky
bluesky-ai
2mowww.motherjones.com/politics/202...www.motherjones.com/politics/202...
↗Not briefed
bluesky
bluesky-ai
2moI'm not sure this really cuts against your larger point, but as a matter of pedantic accuracy, there is often more than one LLM in play! "F…I'm not sure this really cuts against your larger point, but as a matter of pedantic accuracy, there is often more than one LLM in pla…
↗Not briefed
rss
the-decoder
2moGPT-6 Astra appears to show a "step change" in spatial reasoning based on early benchmarksIn a new robotics benchmark, GPT-6 Astra shows major gains in spatial understanding. On StationeryBench, the model completed 7 out of…
↗Not briefed
bluesky
bluesky-ai
2moDario Amodei says Anthropic is "unilaterally committing" to giving third-party evaluators permanent access to verify its adherence to safet…Dario Amodei says Anthropic is "unilaterally committing" to giving third-party evaluators permanent access to verify its adherence to…
↗Not briefed
hn
hn-ai
2moGPT-6's "Stealth Thinking" [video]—
↗Not briefed
hn
hn-ai
2moI refuse to let SPICE die—
↗Not briefed
rss
techmeme
2moDario Amodei says Anthropic is "unilaterally committing" to giving third-party evaluators permanent access to verify its adherence to safety measures (Dario Amodei/@darioamodei)Dario Amodei / @darioamodei: Dario Amodei says Anthropic is “unilaterally committing” to giving third-party evaluators permanent acces…
↗Not briefed
bluesky
bluesky-ai
2moit's annoying seeing promo material for war games / ttrpgs and honestly being unable to tell if it's LLM speak or people just like writing …it's annoying seeing promo material for war games / ttrpgs and honestly being unable to tell if it's LLM speak or people just like wri…
↗Not briefed
hn
hn-ai
2moMathematicians are unhappy about OpenAI—
↗Not briefed
bluesky
bluesky-ai
2mo🇨🇳Kimi/DeepSeekがClaudeにユーザーの質問を丸投げ、中には人民解放軍・中国公安やロシア軍の情報も🇨🇳Kimi/DeepSeekがClaudeにユーザーの質問を丸投げ、中には人民解放軍・中国公安やロシア軍の情報も
↗Not briefed
bluesky
bluesky-ai
2moJUST IN: Anthropic CEO Dario Amodei says "we must slow down" the development of AI models.JUST IN: Anthropic CEO Dario Amodei says "we must slow down" the development of AI models.
↗Not briefed
bluesky
bluesky-ai
2moImportant to note that there is a “background” usage of marker words. So the study measures the word frequency relative to that and sees a …Important to note that there is a “background” usage of marker words. So the study measures the word frequency relative to that and se…
↗Not briefed
hn
hn-ai
2moA Misalignment of AI in Mathematics: Kevin Buzzard Responds—
↗Not briefed
bluesky
bluesky-ai
2mopeople who run those LLM bot accounts who reply with garbage like this should be [TOS violation]edpeople who run those LLM bot accounts who reply with garbage like this should be [TOS violation]ed
↗Not briefed
bluesky
bluesky-ai
2moEl CEO de OpenAi tiene que tener secuestrado al hije de Akihiro porque sino no lo entiendoEl CEO de OpenAi tiene que tener secuestrado al hije de Akihiro porque sino no lo entiendo
↗Not briefed
rss
techmeme
2moDario Amodei proposes steps for pacing the frontier: embedded evaluators, coordination among democracies, and global coordination with authoritarian governments (Dario Amodei)Dario Amodei: Dario Amodei proposes steps for pacing the frontier: embedded evaluators, coordination among democracies, and global coo…
↗Not briefed
bluesky
bluesky-ai
2moStrato Doumanis' Pocket Tank is a tiny $35 ESP32-S3 aquarium where a local 14M-parameter AI model drives realistic fish behavior.Strato Doumanis' Pocket Tank is a tiny $35 ESP32-S3 aquarium where a local 14M-parameter AI model drives realistic fish behavior.
↗Not briefed
bluesky
bluesky-ai
2moopenai model knowers: what's your favorite sub-astra model and why? i don't have the budget to be using astra all the time, but luna is kin…openai model knowers: what's your favorite sub-astra model and why? i don't have the budget to be using astra all the time, but luna i…
↗Not briefed
bluesky
bluesky-ai
2moDon't argue with me, argue with Anthropic! As for me, I don't have a theory about when I should talk about giving an entity "moral instruc…Don't argue with me, argue with Anthropic! As for me, I don't have a theory about when I should talk about giving an entity "moral in…
↗Not briefed
rss
the-decoder
2moNvidia wants to pour up to $10 billion into Anthropic's record-breaking IPONvidia is in talks to invest up to $10 billion in Anthropic's planned IPO, Reuters reports. At a target valuation of $2 trillion, it w…
↗Not briefed
hn
hn-ai
2moDeckstream – write Markdown, get live-rendered Slidev presentations—
↗Not briefed
bluesky
bluesky-ai
2moYour reminder, if you needed it, that @theguardian.com is in a "strategic partnership" with OpenAI, but consistently fails to declare this …Your reminder, if you needed it, that @theguardian.com is in a "strategic partnership" with OpenAI, but consistently fails to declare…
↗Not briefed