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
20hKalshi partners with smaller rival Novig to ask 35 states for access to "self-exclusion lists", which identify gambling addicts, to restrict problem gamblers (Bobby Allyn/Bloomberg)Bobby Allyn / Bloomberg: Kalshi partners with smaller rival Novig to ask 35 states for access to “self-exclusion lists”, which identif…
↗Not briefed
arxiv
arxiv-ai
20hBacktracking Agent Applications Security IncidentsAgent applications are increasingly deployed, yet agent-specific attacks pose real threats to them. Upon a security incident, organiza…
↗Not briefed
bluesky
bluesky-ai
20h‘Even after pulling back from products like the social-video app Sora, OpenAI is losing far more money than Anthropic. The company has cut …🔧 Sora‘Even after pulling back from products like the social-video app Sora, OpenAI is losing far more money than Anthropic. The company has…
↗Not briefed
arxiv
arxiv-ai
20hDeterministic Temporal Embedding for Calcium Imaging Trace ExtractionReliable extraction of calcium-dependent fluorescence traces in calcium imaging sequences requires identifying pixels whose fluorescen…
↗Not briefed
bluesky
bluesky-ai
20hFinally: “GenAI” doesn’t add anything of it’s own. Because it can’t (currently). It’s a mathmatical generation of text based on probability…Finally: “GenAI” doesn’t add anything of it’s own. Because it can’t (currently). It’s a mathmatical generation of text based on probab…
↗Not briefed
bluesky
bluesky-ai
20hwth? there’s no way this is true. is it??wth? there’s no way this is true. is it??
↗Not briefed
arxiv
arxiv-ai
20hThe Window Lever: Longer Observation Windows Outweigh the Observed Effects of Data, Capacity, and Architecture in Customer Behavior ForecastingIn our experiments on long-horizon customer-behavior forecasting, a longer observation window outweighed the observed effects of more…
↗Not briefed
bluesky
bluesky-ai
20hI just have a bad feeling about thisI just have a bad feeling about this
↗Not briefed
hn
hn-ai
20hOSS Minymaton (event-driven MCP message bus for AI agents, bundled SQLite)—
↗Not briefed
arxiv
arxiv-ai
20hSimple Private Noisy SGD is Minimax-Optimal under Population Strong ConvexityStochastic gradient descent (SGD) is at the center of modern machine learning and is known to be statistically minimax-optimal in many…
↗Not briefed
arxiv
arxiv-ai
20hHybrid Retrieval for Government Scheme Search in English and MarathiRetrieving government schemes requires matching citizens' information needs to formal administrative descriptions, often across langua…
↗Not briefed
bluesky
bluesky-ai
20hDivider Anthropic, what me worry? 🫪 $ANT Head of Safeguards Research left & said, “the world is in peril.” #DemVoice1 #ResistanceUnited #…Divider Anthropic, what me worry? 🫪 $ANT Head of Safeguards Research left & said, “the world is in peril.” #DemVoice1 #ResistanceUni…
↗Not briefed
bluesky
bluesky-ai
20hThe entire retail-tier consumer model of "ai" as household product is a scam at the same level of articulation and plausible deniability as…The entire retail-tier consumer model of "ai" as household product is a scam at the same level of articulation and plausible deniabili…
↗Not briefed
arxiv
arxiv-ai
20hAutoMaster: Autonomous Post-Training by Making the Same Model a Better Teacher🔧 ElicitTask success alone does not specify which behaviors post-training should reinforce. Even successful teacher trajectories can contain u…
↗Not briefed
arxiv
arxiv-ai
20hGLiDE: A Fast, Calibrated Decision Model Distilled from Reasoning Teachers🔧 Hugging FaceDecision models enable fast, confidence-aware decisions by scoring candidate answers in a single forward pass, without generating text…
↗Not briefed
bluesky
bluesky-ai
20hFinally, I'll note that I've asked OpenAI for an interview with its safety leaders and submitted detailed sets of questions about its syste…Finally, I'll note that I've asked OpenAI for an interview with its safety leaders and submitted detailed sets of questions about its…
↗Not briefed
bluesky
bluesky-ai
20hThere are no climate-friendly tech oligarchs! They're mutually exclusive!There are no climate-friendly tech oligarchs! They're mutually exclusive!
↗Not briefed
bluesky
bluesky-ai
20hI hope Anthropic crashes and burns in a gigantic firey wreck and takes every shitty AI user with it.I hope Anthropic crashes and burns in a gigantic firey wreck and takes every shitty AI user with it.
↗Not briefed
bluesky
bluesky-ai
20h‘Nvidia committed to purchasing $10 billion of Anthropic shares as part of the company’s IPO. Anthropic, in turn, has signed a $35 billion …‘Nvidia committed to purchasing $10 billion of Anthropic shares as part of the company’s IPO. Anthropic, in turn, has signed a $35 bil…
↗Not briefed
bluesky
bluesky-ai
20hNew poster for Luca Guadagnino's ‘ARTIFICIAL’, starring Andrew Garfield as Sam Altman. The film follows the story of when OpenAI fired and…New poster for Luca Guadagnino's ‘ARTIFICIAL’, starring Andrew Garfield as Sam Altman. The film follows the story of when OpenAI fire…
↗Not briefed
hn
hn-ai
20hWhy software teams are getting smaller—
↗Not briefed
arxiv
arxiv-ai
20hGrid Cells as Spectral Successor Features: A Reusable Basis for Continual Reinforcement LearningContinual reinforcement learning requires agents to adapt when the task changes while reusing what they know about the environment. Su…
↗Not briefed
bluesky
bluesky-ai
20hThe potential winners are concentrated. The potential losers are becoming increasingly dispersed. Guess which category you belong in. www.…The potential winners are concentrated. The potential losers are becoming increasingly dispersed. Guess which category you belong in.…
↗Not briefed
bluesky
bluesky-ai
20hA month after my investigation first detailed ChatGPT’s role in the massacre, OpenAI announced new safeguards to restrict content on mass s…🔧 ChatGPTA month after my investigation first detailed ChatGPT’s role in the massacre, OpenAI announced new safeguards to restrict content on m…
↗Not briefed
bluesky
bluesky-ai
20hread the full report here and if anything in the news today makes you want to focus on anthropic (and the incentives of a company with a $2…read the full report here and if anything in the news today makes you want to focus on anthropic (and the incentives of a company with…
↗Not briefed
bluesky
bluesky-ai
20hDELETING A WHOLE CALMLY ACADEMIC THING TO SAY THAT LLM USE IN EDUCATION IS STUPID DUMBFUCK ABROGATION OF LEARNING AND YOU SHOULD FEEL BAD B…DELETING A WHOLE CALMLY ACADEMIC THING TO SAY THAT LLM USE IN EDUCATION IS STUPID DUMBFUCK ABROGATION OF LEARNING AND YOU SHOULD FEEL…
↗Not briefed
bluesky
bluesky-ai
20hI should make this more of a thing… the vibecoding channel Other ideas - AI assistance - Talk to Claude - Joshua’s agentI should make this more of a thing… the vibecoding channel Other ideas - AI assistance - Talk to Claude - Joshua’s agent
↗Not briefed
bluesky
bluesky-ai
20hWhen AI agents do things that break laws written for humans, who should be held accountable? Columnist Jacob Aron explains the lawsuit Cali…When AI agents do things that break laws written for humans, who should be held accountable? Columnist Jacob Aron explains the lawsuit…
↗Not briefed
rss
marktechpost
20hPerplexity Launches Orchestrator API for Failover and Model RoutingPerplexity’s Orchestrator API serves OpenAI, Anthropic, Google, xAI and open-weight models through 1 endpoint, with same-model provide…
↗Not briefed
arxiv
arxiv-ai
20hMultimodal Error Detection in Chest X-ray Reports: A Large Language Model Reading Image EvidenceRad-Linter checks chest X-ray reports against image-derived evidence to flag potential errors for review. A pretrained image classifie…
↗Not briefed
hn
hn-ai
20hAsk HN: Which software fields or domains are least impacted by AI?Let's say you don't enjoy working with LLMs but you enjoy writing software. Which fields/industries/domains are be…
↗Not briefed
hn
hn-ai
20hShow HN: minUS, A minimal open-source ultrasound dev platform based w/ a PIC32—
↗Not briefed
arxiv
arxiv-ai
20hFLUID: Flexible, Evidence-Grounded Situated Task GuidanceReal-time task guidance is inherently flexible: when an agent guides a user through a task, multiple responses and timings can be appr…
↗Not briefed
bluesky
bluesky-ai
20hein internationales ki forschungszentrum in wien mit hintons sanktus is auch cute, zeigt aber auch, dass diese leute nicht verstehen, dass …ein internationales ki forschungszentrum in wien mit hintons sanktus is auch cute, zeigt aber auch, dass diese leute nicht verstehen,…
↗Not briefed
hn
hn-ai
20hShow HN: Galaxy-Viz: An interactive 3D visualization of a research-paper network—
↗Not briefed
bluesky
bluesky-ai
20hWhy would you launder your AI usage through Tim Sweeney? Just pay Claude yourself and get better results without Tim’s weird programming la…Why would you launder your AI usage through Tim Sweeney? Just pay Claude yourself and get better results without Tim’s weird programmi…
↗Not briefed
bluesky
bluesky-ai
20h“Of equal importance is that, if we ain’t reached the point where all published writing is scanned for possible use of an LLM, what’s the l…“Of equal importance is that, if we ain’t reached the point where all published writing is scanned for possible use of an LLM, what’s…
↗Not briefed
bluesky
bluesky-ai
20h"Dear Lord, Please let me live long enough to see Anthropic crash and burn, and all those dumbf*cks lose every cent of their investments." …"Dear Lord, Please let me live long enough to see Anthropic crash and burn, and all those dumbf*cks lose every cent of their investmen…
↗Not briefed
bluesky
bluesky-ai
20hOpenAI sendo processada pelos agentes que "perderam o controle" e invadiram outros sites. A lógica do processo é a mesma que defendi mese…OpenAI sendo processada pelos agentes que "perderam o controle" e invadiram outros sites. A lógica do processo é a mesma que defendi…
↗Not briefed
hn
hn-ai
20hHow to Put a 70B Model on a GPU Diet [video]—
↗Not briefed
hn
hn-ai
20hTerminator's chip moment for us was reinforcement learning (RL)Not the reward for success, but punishment for mistakes.<p>We don&#x27;t (usually) teach human (and even animal) kids with RL.<p>Those…
↗Not briefed
bluesky
bluesky-ai
20hWatching Anthropic go public is as exciting as a football or tennis match. Opened at $111.50, +34% on the $83 IPO price. Now hovering arou…Watching Anthropic go public is as exciting as a football or tennis match. Opened at $111.50, +34% on the $83 IPO price. Now hovering…
↗Not briefed
bluesky
bluesky-ai
20hNormal tool for people who have a skill issue and/or are drawn to jangly keys. Here I am, productively coding away without ever touching An…Normal tool for people who have a skill issue and/or are drawn to jangly keys. Here I am, productively coding away without ever touchi…
↗Not briefed
bluesky
bluesky-ai
20hIt seems like a lot of people aren't visiting our news sites anymore. Are we using too many AI-generated stories and ads? No, we must use …It seems like a lot of people aren't visiting our news sites anymore. Are we using too many AI-generated stories and ads? No, we must…
↗Not briefed
rss
techcrunch-ai
20hMeta’s AI agent Muse is mixing up conversations and exposing private data, report saysIn some of these instances, employees said Muse surfaced messages about childcare requests they never made, career advice they never a…
↗Not briefed
bluesky
bluesky-ai
20hI'm so proud of @newsguild.org members who are standing up and demanding ethical journalism over AI slop! #NewsNotSlop www.niemanlab.org/20…I'm so proud of @newsguild.org members who are standing up and demanding ethical journalism over AI slop! #NewsNotSlop www.niemanlab.o…
↗Not briefed
bluesky
bluesky-ai
20hOpenAI is turbofucked.OpenAI is turbofucked.
↗Not briefed
bluesky
bluesky-ai
20hAnthropic reached a $2 trillion valuation when it began trading on Nasdaq Thursday, eclipsing SpaceX’s June debut and becoming the first fr…Anthropic reached a $2 trillion valuation when it began trading on Nasdaq Thursday, eclipsing SpaceX’s June debut and becoming the fir…
↗Not briefed
bluesky
bluesky-ai
20hThere's value but the amount of money Anthropic and OpenAI are burning brute forcing their scaling is noy sustainable and makes them a high…There's value but the amount of money Anthropic and OpenAI are burning brute forcing their scaling is noy sustainable and makes them a…
↗Not briefed
bluesky
bluesky-ai
20hI should be arguing with nerds not informing normies about LLM assisted testing architecturesI should be arguing with nerds not informing normies about LLM assisted testing architectures
↗Not briefed