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 |
|---|---|---|---|---|
rss theverge-ai | 12d | David Ellison says AI could be the ‘biggest creative unlock in modern history’ | During Paramount Skydance's latest earnings call, CEO David Ellison said "it's hard to imagine" AI won't have a bigger impact on the f… | ↗Not briefed |
arxiv arxiv-ai | 12d | Orthographic depth at two scales, with English also matched for lexicon size | A reader can struggle to read in only one of their languages, and one proposed reason is how predictably spelling maps to sound. We me… | ↗Not briefed |
bluesky bluesky-ai | 12d | ok well on one hand yes on the other hand it's actually very capable wrt finding CVEs (as evaluated by maintainers after receiving reports,… | ok well on one hand yes on the other hand it's actually very capable wrt finding CVEs (as evaluated by maintainers after receiving rep… | ↗Not briefed |
bluesky bluesky-ai | 12d | Your llm is being juuuuust a li’l bit silly, I feel | Your llm is being juuuuust a li’l bit silly, I feel | ↗Not briefed |
hn hn-ai | 12d | Show HN: I abandoned my AVX2 128-bit fixed-point library | — | ↗Not briefed |
bluesky bluesky-ai | 12d | In a new report, #OpenAI says #ChatGPT receives +70M requests about religion each week. The report leaves a question unanswered for minorit…🔧 ChatGPT | In a new report, #OpenAI says #ChatGPT receives +70M requests about religion each week. The report leaves a question unanswered for mi… | ↗Not briefed |
hn hn-ai | 12d | Does paying more get you better reviews? Rarely | — | ↗Not briefed |
arxiv arxiv-ai | 12d | The Dose Makes the Poison: Exposure-Aware Evaluation of Generative Drug Design | Generative models are increasingly used to propose molecules for drug discovery. They can produce far more candidates than can be expe… | ↗Not briefed |
rss wired-ai | 12d | Jensen Huang’s Troubles Leave the AI Industry Reeling | With a big checkbook and sought-after chips, Nvidia’s CEO had his thumb on the world of AI. An extramarital affair with a suspected Ch… | ↗Not briefed |
arxiv arxiv-ai | 12d | The AI Era's Lines of Code? A Case Study on What an Engineer's Token Spend Measures | Organizations use engineers' token spend to monitor AI adoption, set budgets, and measure productivity, but the metric's relationship… | ↗Not briefed |
arxiv arxiv-ai | 12d | Plumbline: When the Loss Lies, Validating LLM Training Correctly | Correctness is a prerequisite for training large language models (LLMs) at scale, yet it is hard to ensure: a fault can sit in any par… | ↗Not briefed |
bluesky bluesky-ai | 12d | The problem with code is that it is both language and math and as a result is suffering the same ills as LLM hammered math currently is but… | The problem with code is that it is both language and math and as a result is suffering the same ills as LLM hammered math currently i… | ↗Not briefed |
bluesky bluesky-ai | 12d | the future of insanely bleak and hilarious legal disclosures is unhinged LLM search logs | the future of insanely bleak and hilarious legal disclosures is unhinged LLM search logs | ↗Not briefed |
bluesky bluesky-ai | 12d | "its a valuable tool" in the same sentence as "the value doesn't add up" is just fucking sending me. what's with the mandatory glazing with… | "its a valuable tool" in the same sentence as "the value doesn't add up" is just fucking sending me. what's with the mandatory glazing… | ↗Not briefed |
bluesky bluesky-ai | 12d | We have also been cranking out these accounts to flood the zone of LLM training data and dominate Google AI search. | We have also been cranking out these accounts to flood the zone of LLM training data and dominate Google AI search. | ↗Not briefed |
bluesky bluesky-ai | 12d | anthropic keeps giving me resets but i legit can't use them because its not sustainable and would obliterate my workflow. like they clearly… | anthropic keeps giving me resets but i legit can't use them because its not sustainable and would obliterate my workflow. like they cl… | ↗Not briefed |
arxiv arxiv-ai | 12d | AdjointFlow: Accelerating Scientific Inverse Problems with Physics-Based Transport | Inverse problems are ubiquitous across scientific applications ranging from seismic imaging to astrophysics to nondestructive testing,… | ↗Not briefed |
bluesky bluesky-ai | 12d | There's a lot to unpack in here but in particular, I'm fascinated by something mentioned early in passing -- The claim that it was "clean-r… | There's a lot to unpack in here but in particular, I'm fascinated by something mentioned early in passing -- The claim that it was "cl… | ↗Not briefed |
bluesky bluesky-ai | 12d | probably the most frustrating part of that METR chart is that, yes, the frontier is no longer accessible to you unless you’re at either Ant… | probably the most frustrating part of that METR chart is that, yes, the frontier is no longer accessible to you unless you’re at eithe… | ↗Not briefed |
arxiv arxiv-ai | 12d | Online-Adaptive Fuzzy Cognitive Maps for Neuro-symbolic Financial Forecasting | We present a neuro-symbolic framework that pairs online-adaptive Fuzzy Cognitive Maps (FCMs) with long short-term memory (LSTM) networ… | ↗Not briefed |
bluesky bluesky-ai | 12d | 🐭big mystery around why programmers seem to be ok with AI use when programmers have been stealing whatever code they can find online for t… | 🐭big mystery around why programmers seem to be ok with AI use when programmers have been stealing whatever code they can find online… | ↗Not briefed |
rss techmeme | 12d | Take-Two reiterates FY 2027 outlook, says Q2 total net bookings fell 18% YoY to $1.6B, vs. $1.65B est., and NBA 2K27 missed sales forecast partly due to GTA VI (Jason Schreier/Bloomberg) | Jason Schreier / Bloomberg: Take-Two reiterates FY 2027 outlook, says Q2 total net bookings fell 18% YoY to $1.6B, vs. $1.65B est., an… | ↗Not briefed |
bluesky bluesky-ai | 12d | a funny thing about the spaghetti falling block theme was, even though I saw a few llm guys trying to do it, they couldn't get claude to fo… | a funny thing about the spaghetti falling block theme was, even though I saw a few llm guys trying to do it, they couldn't get claude… | ↗Not briefed |
bluesky bluesky-ai | 12d | Fable 5.2 is “Model 3 Early” in this chart that means Anthropic is holding on to at least 2 better models, presumably Model 4 will be Fabl… | Fable 5.2 is “Model 3 Early” in this chart that means Anthropic is holding on to at least 2 better models, presumably Model 4 will be… | ↗Not briefed |
arxiv arxiv-ai | 12d | How Hyperbolic Are Hyperbolic Vision-Language Models? | Hyperbolic vision-language models (VLMs) have emerged as a rapidly growing area of research, reporting strong zero-shot performance al… | ↗Not briefed |
bluesky bluesky-ai | 12d | This joke has slightly broken containment around here. Hi do you enjoy Remedy? Maybe you'll like some of my 3d printed fan works based on A… | This joke has slightly broken containment around here. Hi do you enjoy Remedy? Maybe you'll like some of my 3d printed fan works based… | ↗Not briefed |
bluesky bluesky-ai | 12d | GPT-4.1 Mini (75% Cache Discount) vs DeepSeek V3 (0324): Which LLM API actually protects your SaaS gross margin at 10,000 DAU? Here is the… | GPT-4.1 Mini (75% Cache Discount) vs DeepSeek V3 (0324): Which LLM API actually protects your SaaS gross margin at 10,000 DAU? Here i… | ↗Not briefed |
arxiv arxiv-ai | 12d | Settling the Burn-in Gap: Sharp Risk Bounds for Tail-Averaged SGD in Linear Regression | We study the excess risk of constant-stepsize online stochastic gradient descent (SGD) with tail-averaging for high-dimensional linear… | ↗Not briefed |
rss huggingface-blog | 12d | Pay for Hugging Face Hub with your Google Cloud Account🔧 Hugging Face | — | ↗Not briefed |
rss huggingface-blog | 12d | Decision Models in Hugging Face🔧 Hugging Face | — | ↗Not briefed |
arxiv arxiv-ai | 12d | Discrete Model-Based Diffusion for Combinatorial Optimization in Multi-Robot Task Allocation | Decision making in multi-agent and multi-robot systems often involves discrete decision variables, leading to NP-hard combinatorial op… | ↗Not briefed |
rss simon-willison | 12d | llm-anthropic 0.31 | — | ↗Not briefed |
bluesky bluesky-ai | 12d | Seeing what pure LLM vibecoding does to people, it gives me very similar feelings to the way people devalue writing and storytelling Just … | Seeing what pure LLM vibecoding does to people, it gives me very similar feelings to the way people devalue writing and storytelling… | ↗Not briefed |
bluesky bluesky-ai | 12d | Oh look it's me explaining why Jensen Huang's clothing choices (and especially the choice by his company to put him on clothing) is literal… | Oh look it's me explaining why Jensen Huang's clothing choices (and especially the choice by his company to put him on clothing) is li… | ↗Not briefed |
rss simon-willison | 12d | Locating the source of a mysterious beep with Claude | — | ↗Not briefed |
hn hn-ai | 12d | Parents Sue Google over A.I. Safety Risks on School Chromebooks | — | ↗Not briefed |
arxiv arxiv-ai | 12d | Looping Deeper Without Growing the KV Cache: Recurrent Depth for Causal Encoder-Decoders | Looped language models deepen computation without adding parameters, yet caching full-context key-value (KV) states across layers and… | ↗Not briefed |
rss theverge-ai | 12d | What could China get from Jensen Huang? | It's difficult to overstate Nvidia's power in today's AI landscape. The company's chips are the bedrock upon which advanced AI technol… | ↗Not briefed |
hn hn-ai | 12d | Anthropic must face lawsuit over Claude subscription marketing | — | ↗Not briefed |
hn hn-ai | 12d | Product analytics events, nearly too cheap to meter | — | ↗Not briefed |
bluesky bluesky-ai | 12d | I really appreciate the explicit call out of GenAI code here. It's been infuriating to see so many people be anti-AI but carve out coding a… | I really appreciate the explicit call out of GenAI code here. It's been infuriating to see so many people be anti-AI but carve out cod… | ↗Not briefed |
rss theverge-ai | 12d | Here’s what Trump said he knew about Jensen Huang | Jensen Huang and President Trump. | Getty Images President Trump held a White House press briefing earlier today and responded to se… | ↗Not briefed |
hn hn-ai | 12d | CodeRabbit scales agentic software development and ships 20x more | — | ↗Not briefed |
bluesky bluesky-ai | 12d | GPT-4.1 (1M Context Frontier) vs Gemini 2.5 Pro: Which LLM API actually protects your SaaS gross margin at 10,000 DAU? Here is the live 20… | GPT-4.1 (1M Context Frontier) vs Gemini 2.5 Pro: Which LLM API actually protects your SaaS gross margin at 10,000 DAU? Here is the li… | ↗Not briefed |
bluesky bluesky-ai | 12d | trump is an llm. there’s no truth, there’s nothing deeper or deliberative inside his mind. his brain just generates the the next word again… | trump is an llm. there’s no truth, there’s nothing deeper or deliberative inside his mind. his brain just generates the the next word… | ↗Not briefed |
bluesky bluesky-ai | 12d | I’m pretty sure these videos are fake, though? At least in the sense that they aren’t depicting actual interviews. It looks like the creat… | I’m pretty sure these videos are fake, though? At least in the sense that they aren’t depicting actual interviews. It looks like the… | ↗Not briefed |
bluesky bluesky-ai | 12d | Anthropic can begin formally marketing its much-anticipated IPO as early as Tuesday, enabling it make the case to investors that now is the… | Anthropic can begin formally marketing its much-anticipated IPO as early as Tuesday, enabling it make the case to investors that now i… | ↗Not briefed |
arxiv arxiv-ai | 12d | Enforcing Axioms for AI Alignment under Loss-Based Rules | Many alignment methods, most notably reinforcement learning from human feedback (RLHF), train reward models by minimizing a loss on pa… | ↗Not briefed |
hn hn-ai | 12d | A teen used her school Chromebook and Google AI Mode to plan her suicide | — | ↗Not briefed |
arxiv arxiv-ai | 12d | No finite rulebook can certify every safe program: an impossibility theorem for gatekeeping AI agents | An AI assistant writes small programs, and a gatekeeper must decide, from the code alone, whether each one may run. Suppose the rules… | ↗Not briefed |