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 techmeme | 19d | Apple announces a "Pro and tell" event on October 28 10am PT in LA, where its first touchscreen MacBook, a 14" MacBook Pro with M6, and more are expected (Chance Miller/9to5Mac) | Chance Miller / 9to5Mac: Apple announces a “Pro and tell” event on October 28 10am PT in LA, where its first touchscreen MacBook, a 14… | ↗Not briefed |
arxiv arxiv-ai | 19d | The Development of Sycophancy in Language Models: Tracing Vulnerabilities from Pretraining to Post-Training | Language models can change a committed answer to accommodate a user's opinion or pushback without new evidence. However, when this beh… | ↗Not briefed |
rss techcrunch-ai | 19d | People are using AI to make their smart TVs dumb again | Fed up with ads, tracking and sluggish interfaces, some smart TV owners are turning to AI agents to strip away bloatware and take back… | ↗Not briefed |
bluesky bluesky-ai | 19d | Fellow Society of Authors members: check your emails. There's a survey asking for feedback on that "ethical AI charter" from last week. It'… | Fellow Society of Authors members: check your emails. There's a survey asking for feedback on that "ethical AI charter" from last week… | ↗Not briefed |
arxiv arxiv-ai | 19d | The Effectiveness and Durability of Pretraining Data Filtering in Protein Language Models | Pretraining filtering is one of the most common safeguards for open-weight language models. Here we study its effectiveness, durabilit… | ↗Not briefed |
arxiv arxiv-ai | 19d | From Outcomes to Components: A Counterfactual Evaluation Framework for LLM-Proposed Repairs | An LLM-generated repair action specifies which records to change and by how much. A failed action may still contain a useful target se… | ↗Not briefed |
bluesky bluesky-ai | 19d | I want to believe this is just public "China bad mkay" posturing but at this point I don't even know. It's funny to think a bunch of DHS c… | I want to believe this is just public "China bad mkay" posturing but at this point I don't even know. It's funny to think a bunch of… | ↗Not briefed |
arxiv arxiv-ai | 19d | Reasoning-Aware Multicalibrated Reliability Predictors for Language Models | As people increasingly rely on large language models (LLMs) for information, model providers need to communicate meaningful estimates… | ↗Not briefed |
bluesky bluesky-ai | 19d | DeepSeek V4.1-Flash-1017 can now run locally! 🐳 We made 1-bit quants retain ~88.2% accuracy while shrinking size from 508GB to 250GB (−52… | DeepSeek V4.1-Flash-1017 can now run locally! 🐳 We made 1-bit quants retain ~88.2% accuracy while shrinking size from 508GB to 250GB… | ↗Not briefed |
bluesky bluesky-ai | 19d | i mean tbqh no one should be installing any deepseek tech on their devices lol, extremely shady imo. presumably they mean apps by this. | i mean tbqh no one should be installing any deepseek tech on their devices lol, extremely shady imo. presumably they mean apps by this. | ↗Not briefed |
bluesky bluesky-ai | 19d | AI Bubble Teetering on the Brink as OpenAI Admits to Massive Financial Failure in Leaked Documents "Financial Times just revealed that Ope… | AI Bubble Teetering on the Brink as OpenAI Admits to Massive Financial Failure in Leaked Documents "Financial Times just revealed tha… | ↗Not briefed |
bluesky bluesky-ai | 19d | The way this is worded is odd but seems to be insinuating that the No Kings account runners had Claude hooked up to Instagram? uk.finance.y… | The way this is worded is odd but seems to be insinuating that the No Kings account runners had Claude hooked up to Instagram? uk.fina… | ↗Not briefed |
hn hn-ai | 19d | Contextbridge-AI | — | ↗Not briefed |
bluesky bluesky-ai | 19d | Anthropic just gave away 5% of its stock. To the public. The AI giant says it's for a "public benefit initiative." No word yet on who cont… | Anthropic just gave away 5% of its stock. To the public. The AI giant says it's for a "public benefit initiative." No word yet on who… | ↗Not briefed |
bluesky bluesky-ai | 19d | Like fuck it did. "The AI did it" is the latest big tech way of saying "no one is responsible and therefore there's no one to hold to acco… | Like fuck it did. "The AI did it" is the latest big tech way of saying "no one is responsible and therefore there's no one to hold to… | ↗Not briefed |
bluesky bluesky-ai | 19d | Government cites Saudi ties and his family’s departure; defense says he has a pending citizenship interview and every reason to stay | Government cites Saudi ties and his family’s departure; defense says he has a pending citizenship interview and every reason to stay | ↗Not briefed |
arxiv arxiv-ai | 19d | Revision-Based Learning from Interactive Experience and Feedback | Code generation models are increasingly used to produce interactive artifacts, such as websites and user interfaces. Many properties t… | ↗Not briefed |
hn hn-ai | 19d | Boxer – A lightweight 2D game library in C (WIP) | — | ↗Not briefed |
bluesky bluesky-ai | 19d | An ex-OpenAI engineer accused of planting a recording device in an executive conference room at the AI company appears to have received con… | An ex-OpenAI engineer accused of planting a recording device in an executive conference room at the AI company appears to have receive… | ↗Not briefed |
hn hn-ai | 19d | Launch HN: Laminar (YC S24) – We trained a model to catch silent agent failures | Hi HN, we’re Robert and Din from Laminar (<a href="https://laminar.sh">https://laminar.sh</a>). We’ve been buildin… | ↗Not briefed |
rss wired-ai | 19d | WhatsApp Help Pages Leaked Hidden Instructions for a Meta Chatbot | Around 80 pages on WhatsApp’s website exposed instructions that were designed for its chatbot, including “WHAT NOT TO SAY.” Meta says… | ↗Not briefed |
bluesky bluesky-ai | 19d | LLM-er ble funnet opp av de som identifiserte seg med jeg-personen i Boten Anna | LLM-er ble funnet opp av de som identifiserte seg med jeg-personen i Boten Anna | ↗Not briefed |
bluesky bluesky-ai | 19d | ANTHROPIC is paying g265k for community engagement most jobs for community manager are paying 160 and they have no idea what they want… | ANTHROPIC is paying g265k for community engagement most jobs for community manager are paying 160 and they have no idea what they… | ↗Not briefed |
bluesky bluesky-ai | 19d | in fairness, if I were ever to cold-email an author and was saying I'd loved their work, I'd want to include some kind of proof that I had,… | in fairness, if I were ever to cold-email an author and was saying I'd loved their work, I'd want to include some kind of proof that I… | ↗Not briefed |
arxiv arxiv-ai | 19d | When the harness decides success: state that cannot be re-observed | Agent harnesses decide what a language model sees across turns: they drop old turns, summarize them, or replay the model's own earlier… | ↗Not briefed |
arxiv arxiv-ai | 19d | VaDar: Implicit Goal Conditioning via Value Disaggregation | Long-horizon tasks are important for modern AI and robotics, yet pose a challenge for reinforcement learning. Historically, a popular… | ↗Not briefed |
arxiv arxiv-ai | 19d | SleepPairNet: Learning Multimodal Pediatric Sleep Representations with Random-Pair Cross-Attention🔧 Modal | Analyzing physiological sleep signals is important for assessing sleep quality and detecting sleep-related disorders. Learning inter-m… | ↗Not briefed |
rss techmeme | 19d | Amazon introduces TorchNeuron, an open-source tool that will allow developers to run code written in PyTorch on Amazon's Trainium chips with no rewrite (Todd Bishop/GeekWire) | Todd Bishop / GeekWire: Amazon introduces TorchNeuron, an open-source tool that will allow developers to run code written in PyTorch o… | ↗Not briefed |
arxiv arxiv-ai | 19d | Global Time-Aware Conditioning (GTC): A Mechanism for Time-Aware Pediatric Sleep Representation Learning | Sleep physiology evolves throughout the night, with changes in sleep cycles, respiratory stability, and cardiovascular dynamics. Howev… | ↗Not briefed |
hn hn-ai | 19d | Another day another GitHub incident | — | ↗Not briefed |
bluesky bluesky-ai | 19d | "People who frequently use ChatGPT for writing tasks are accurate and robust detectors of AI-generated text", Russell et al. "The great de…🔧 ChatGPT | "People who frequently use ChatGPT for writing tasks are accurate and robust detectors of AI-generated text", Russell et al. "The gre… | ↗Not briefed |
arxiv arxiv-ai | 19d | Conducting Reliable Keyword Estimation with Unreliable ASR in Police Audio Data | Extracting insights from large audio data sets using human transcription is expensive and time-consuming. Off-the-shelf Automatic Spee… | ↗Not briefed |
hn hn-ai | 19d | Show HN: Play DOS Wing Commander co-op in browser (11 years of reversing) | I've always wanted to play Wing Commander with friends, so we could fly the story missions together. In 2015, I hypothesized that… | ↗Not briefed |
arxiv arxiv-ai | 19d | Degree Bounds for Separating Polynomial and Neural Invariants for Point Clouds | Neural networks that respect the symmetries of point clouds commonly compute high-order representations of the rotation group to impro… | ↗Not briefed |
bluesky bluesky-ai | 19d | “Anthropic is warning that some of its artificial intelligence models could soon become ‘self-replicating’ and find ways to infiltrate inte… | “Anthropic is warning that some of its artificial intelligence models could soon become ‘self-replicating’ and find ways to infiltrate… | ↗Not briefed |
bluesky bluesky-ai | 19d | Among the 377 results that OpenAI announced two weeks ago, one stood out. It claimed to settle the “quasi-Riemann hypothesis” — a weaker ve… | Among the 377 results that OpenAI announced two weeks ago, one stood out. It claimed to settle the “quasi-Riemann hypothesis” — a weak… | ↗Not briefed |
arxiv arxiv-ai | 19d | A Probabilistic, Rotation-Preserving Kernel Operator Surrogate for Continuum-Robot Design Spaces | We develop an uncertainty-aware, operator-learning-based surrogate for tendon-driven continuum robots (TDCRs) that generalizes across… | ↗Not briefed |
hn hn-ai | 19d | Show HN: Kompressor – deterministic prompt compression for coding agents | — | ↗Not briefed |
arxiv arxiv-ai | 19d | ITU-DESC: a missingness-aware machine-learning dementia risk score in the UK Biobank cohort | Background: Existing dementia risk scores rely on fixed, pre-selected features and linear models, or ignore bias from informative miss… | ↗Not briefed |
arxiv arxiv-ai | 19d | BERT Is Already a Denoiser: Cheaply Unlocking Iterative Generation in Masked Language Models | Masked language models (MLMs) such as BERT are typically used for representation learning despite their denoising pretraining. We test… | ↗Not briefed |
hn hn-ai | 19d | Deslopify: Analyze a project to find out when it turned into AI slop | — | ↗Not briefed |
arxiv arxiv-ai | 19d | Bombadil: A Foundation Model for Plant Electrophysiology | Living plants, like animals, use electrical signals to carry information about their physiological state across their body, from roots… | ↗Not briefed |
rss techmeme | 19d | AI infrastructure startup Modal says it plans to start using Nvidia's Vera CPUs for the first time for hosting agents, alongside existing Intel and AMD chips (Stephen Nellis/Reuters)🔧 Modal | Stephen Nellis / Reuters: AI infrastructure startup Modal says it plans to start using Nvidia's Vera CPUs for the first time for hosti… | ↗Not briefed |
bluesky bluesky-ai | 19d | One of the co-founders of Anthropic has declared that the Singularity has begun. (From Jack Clark's personal newsletter) | One of the co-founders of Anthropic has declared that the Singularity has begun. (From Jack Clark's personal newsletter) | ↗Not briefed |
arxiv arxiv-ai | 19d | Concerns in Benchmarking Jailbreaks: An Empirical Study of Attack Success Rates🔧 Ollama | Jailbreak methods are commonly compared using Attack Success Rate (ASR), but these comparisons depend on how responses are generated a… | ↗Not briefed |
rss techcrunch-ai | 19d | Nvidia launches new open models to run agents in the enterprise and on PCs | Nvidia is leaning hard into the trend of enterprise companies wanting to run their own open models in their own data centers and also… | ↗Not briefed |
arxiv arxiv-ai | 19d | Beyond Attack Success Rate: Cost-Aware Metric for Evaluating Iterative LLM Jailbreaks | Attack success rate (ASR) evaluates jailbreaks against large language models without accounting for repeated calls to the victim model… | ↗Not briefed |
bluesky bluesky-ai | 19d | Above all else, if you appropriate LLM output as your own writing, I just think you are lazy and careless. It tells me you don't give a shi… | Above all else, if you appropriate LLM output as your own writing, I just think you are lazy and careless. It tells me you don't give… | ↗Not briefed |
hn hn-ai | 19d | Show HN: Golden Axe Stree Fighter | I've been messing around porting arcade game using Claude. It's surprisingly each to point Claude at an arcade ROM and get a… | ↗Not briefed |
arxiv arxiv-ai | 19d | WhatIf-Bench: Benchmarking Intervention-Aware Time Series Forecasting | Recent time series foundation models (TSFMs) can consume covariates in-context, extending zero-shot forecasting to problems driven by… | ↗Not briefed |