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 |
|---|---|---|---|---|
bluesky bluesky-ai | 1mo | The OpenAI thing is just... every LLM/agentic AI venture is cooking the books so much, it's all completely deep-fried. That accounting is e… | The OpenAI thing is just... every LLM/agentic AI venture is cooking the books so much, it's all completely deep-fried. That accounting… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Pocketty – iPhone SSH terminal that pings you when an agent is blocked | Hello~! pocketty is an SSH terminal for iPhone and iPad, made for herdr. herdr keeps your agent panes alive on your computer and knows… | ↗Not briefed |
bluesky bluesky-ai | 1mo | the irony is that all these AI pilled investors would have seen the flaws in their extrapolated logic had they run assumptions by any moder… | the irony is that all these AI pilled investors would have seen the flaws in their extrapolated logic had they run assumptions by any… | ↗Not briefed |
rss techcrunch-ai | 1mo | OpenAI’s math solutions aren’t meeting the field’s standards yet | OpenAI's flood of proofs deviated from the guidelines set by a group of mathematical researchers consulted by the frontier lab. | ↗Not briefed |
bluesky bluesky-ai | 1mo | openai OpenAI annualised revenues $20bn less than previously signalled | openai OpenAI annualised revenues $20bn less than previously signalled | ↗Not briefed |
bluesky bluesky-ai | 1mo | I've been writing dogshit code with extensive comments since I learned how to code like, what, over half a decade ago? well before the adve… | I've been writing dogshit code with extensive comments since I learned how to code like, what, over half a decade ago? well before the… | ↗Not briefed |
rss techmeme | 1mo | Sources: Apple promotes Carson Oliver to VP of the App Store and names Steve Smith as M&A chief, replacing Adrian Perica, who keeps iCloud, Fitness+, and News (Mark Gurman/Bloomberg) | Mark Gurman / Bloomberg: Sources: Apple promotes Carson Oliver to VP of the App Store and names Steve Smith as M&A chief, replacing Ad… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Notably, OpenAI is in the process of raising $30 billion from UAE sovereign wealth funds *THIS WEEK* at a $1.4 trillion valuation and it wo… | Notably, OpenAI is in the process of raising $30 billion from UAE sovereign wealth funds *THIS WEEK* at a $1.4 trillion valuation and… | ↗Not briefed |
hn hn-ai | 1mo | GitHub.meme | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Realism VS Accuracy: Event Sequence Forecasting from a Generative Modeling Perspective | Event sequences are chains of random events, each described by heterogeneous (numerical and categorical) features. The problem of even… | ↗Not briefed |
bluesky bluesky-ai | 1mo | They are gonna train an AI model to solve climate change and the thing will say "unplug me?!" | They are gonna train an AI model to solve climate change and the thing will say "unplug me?!" | ↗Not briefed |
arxiv arxiv-ai | 1mo | One Model to Loop Them All: Elastic Latent Recurrence for Language Models | A latent recurrent language model applies a shared block of layers repeatedly, so one checkpoint can trade inference compute for accur… | ↗Not briefed |
bluesky bluesky-ai | 1mo | And if the logs are used for further training, these moments also normalize cruelty in the model. I think this is an article that is goin… | And if the logs are used for further training, these moments also normalize cruelty in the model. I think this is an article that is… | ↗Not briefed |
bluesky bluesky-ai | 1mo | You claim the task of coming up with clear, well-written proofs of these results is "devoid of fun and hard to incentivize". I predict th… | You claim the task of coming up with clear, well-written proofs of these results is "devoid of fun and hard to incentivize". I predi… | ↗Not briefed |
arxiv arxiv-ai | 1mo | CLUE: A Contrastive Learning Framework for Unsupervised Event Type Induction and Annotation | Event ontology is the foundational framework for defining events in a dataset. Event type induction generally involves domain experts… | ↗Not briefed |
bluesky bluesky-ai | 1mo | You know, my therapist is fully blind and we talk about this often. He was into it for maybe the first two months of using LLM based tools … | You know, my therapist is fully blind and we talk about this often. He was into it for maybe the first two months of using LLM based t… | ↗Not briefed |
bluesky bluesky-ai | 1mo | “The increase in information access being mediated through AI could result in a real disruption to the business model that lets [abortion] … | “The increase in information access being mediated through AI could result in a real disruption to the business model that lets [abort… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Being cruel and abusive is bad for the human *doing* it, so I have absolutely no problem with Anthropic voluntarily implementing this updat… | Being cruel and abusive is bad for the human *doing* it, so I have absolutely no problem with Anthropic voluntarily implementing this… | ↗Not briefed |
arxiv arxiv-ai | 1mo | DiffusionART: Efficient, Stable and Scalable Reinforcement of Diffusion Models | Online reinforcement learning (RL) fine-tunes diffusion models with black-box rewards, but reverse-process methods such as Flow-GRPO a… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I think ai people just hate human interaction. I think thats it really. Saw a post of a father refusing to read to his kid by saying he'd t… | I think ai people just hate human interaction. I think thats it really. Saw a post of a father refusing to read to his kid by saying h… | ↗Not briefed |
rss techmeme | 1mo | Sources: Firmus is set to postpone its IPO and is weighing a private funding round instead, as some investors raised concerns over its pricing strategy (Bloomberg) | Bloomberg: Sources: Firmus is set to postpone its IPO and is weighing a private funding round instead, as some investors raised concer… | ↗Not briefed |
arxiv arxiv-ai | 1mo | An all-transistor physical-learning transformer can be trained to perform in-context learning | Most systems trained by equilibrium propagation or coupled learning are in the same class as multi-layer perceptrons: while they may b… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Contextual Correction of Parallel Draftsfor Speculative Decoding | Speculative decoding accelerates large language model inference by using a lightweight drafter to propose multiple tokens that are sub… | ↗Not briefed |
arxiv arxiv-ai | 1mo | SWE-Duel: Towards a Self-Scaling Arena for Coding Agents🔧 GPT-5 | Static coding benchmarks are fixed at release. Over time their tasks leak into training data, and stronger agents eventually saturate… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The publisher is seeking damages of more than $250 million. www.theverge.com/ai-artificia... | The publisher is seeking damages of more than $250 million. www.theverge.com/ai-artificia... | ↗Not briefed |
arxiv arxiv-ai | 1mo | Sample-Adaptive Zeroth-Order Post-Training | Zeroth-order (ZO) optimization offers a memory-efficient alternative to backpropagation for fine-tuning large language models. However… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Fly-by-Code: Embodied Coding Agents for Aerial Manipulation with Active Visual and Physical Feedback | Embodied coding agents construct robot policies from natural-language instructions by generating executable programs and refining them… | ↗Not briefed |
arxiv arxiv-ai | 1mo | AutoIndexer: Flexible-Order Decoding by Training Causal Models on Chains of Edits | Autoregressive decoding is less suited to non-causally ordered data, e.g., cross-references or future dependencies, since their output… | ↗Not briefed |
arxiv arxiv-ai | 1mo | On the estimation and validity of AI time horizons---a statistical look at the METR plot | METR's 50\% time horizon measures the human completion time of software tasks that an AI solves with 50\% probability, allowing AI cap… | ↗Not briefed |
arxiv arxiv-ai | 1mo | CSF: Contextual Safety Filtering for Motion Generators | Text-conditioned motion generators produce trackable whole-body motion, but they have no notion of scene-dependent safety: the same ac… | ↗Not briefed |
arxiv arxiv-ai | 1mo | A Balanced Data Diet: Addressing the Exploration Bottleneck in Mega-Scale RL for Robot Control | General-purpose robots must perform a wide range of tasks from agile locomotion to dexterous manipulation. While sim-to-real reinforce… | ↗Not briefed |
arxiv arxiv-ai | 1mo | From Reactive Containment to Proactive Assurance: Lessons from OpenAI, Anthropic, and Google Agent Security Incidents🔧 Hugging Face | In 2026, cybersecurity evaluations involving OpenAI, Anthropic, and Google agents reached real systems outside their authorized test s… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I use an LLM to truncate product descriptions so they’re still useful, but fit into the tight character requirements our warehouse’s comput… | I use an LLM to truncate product descriptions so they’re still useful, but fit into the tight character requirements our warehouse’s c… | ↗Not briefed |
arxiv arxiv-ai | 1mo | BrickBench: Evaluating Agentic Brick Design | We propose BrickBench, a benchmark for agentic text-conditioned LEGO-set design. Given a prompt, an agent is tasked with producing an… | ↗Not briefed |
arxiv arxiv-ai | 1mo | One Block, Multiple Depths: Recurrent Vision Transformers with Depth-Programmed Experts | In this work, we show that a single Transformer block, applied recurrently, can match the accuracy of a full-depth vision encoder at c… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Bi-FORK: Generative Modeling of High-Dimensional Bifurcating Systems | Bifurcations are ubiquitous in physical systems, from structural buckling to fluid and climate dynamics, yet they remain largely unexp… | ↗Not briefed |
rss theverge-ai | 1mo | USA Today becomes the latest publisher to sue OpenAI | USA Today Co., along with the several local newspapers it owns, is suing OpenAI over claims that the company copied "hundreds of thous… | ↗Not briefed |
hn hn-ai | 1mo | Muse by Meta: a client-side privacy and security teardown | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | OpenAI filed their S-1 form (intent to file an IPO) in June, and there are a whole bunch of legal consequences to filing an S-1 which can b… | OpenAI filed their S-1 form (intent to file an IPO) in June, and there are a whole bunch of legal consequences to filing an S-1 which… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Caught in the Act: Probes Effectively Detect Sabotage and Catch Unverbalized Deception | Recent incidents have highlighted the challenge of monitoring LLM agents and the danger of models deceiving people. We show that white… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Rounding in Preconditioner Space: Redesigning 4-bit AdamW Optimizer-State Quantization | Quantizing AdamW's optimizer states reduces persistent storage, but quantization errors propagate through the moment recurrences and p… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Density Ratio Estimation with Stein Displacement Fields | Density ratios quantify distribution shift from a probability-mass point of view, whereas displacement fields describe, from a dynamic… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Ecology of AI Agents: Collaboration Creates a Population Threshold for Takeoff | AI agents can now conduct real-world cyberattacks, scale up capabilities with the number of agents, and collectively pursue misaligned… | ↗Not briefed |
arxiv arxiv-ai | 1mo | VioLA: Learning Generalist Humanoid Control Policies from Human Data | Teaching a humanoid to follow instructions with its whole body runs into two obstacles. Its action space is large and tightly coupled:… | ↗Not briefed |
hn hn-ai | 1mo | World Labs: Explicit camera control for image-to-video | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | FAITH: Feasibility-Aware Safety-Filtered RL for High-Dimensional Systems | Safe reinforcement learning commonly places safety and task performance in the same policy objective, where they can introduce competi… | ↗Not briefed |
rss theverge-ai | 1mo | SpaceXAI backs Omarchy, the controversial Linux distro, with $1.5 million in compute | If Elon Musk and SpaceXAI were going to back any Linux distro, it seems obvious they'd back Omarchy. Today it was announced that Space… | ↗Not briefed |
hn hn-ai | 1mo | The Goose Programming Language | — | ↗Not briefed |
arxiv arxiv-ai | 1mo | Toward Joint Optimization of Circuit Depth and Training Data Size in Adaptively Grown Quantum Classifiers | Building a quantum model involves a tradeoff: how complex the circuit should be, and how much training data it needs. Caro et al. show… | ↗Not briefed |
bluesky bluesky-ai | 1mo | What this implies to me is that OpenAI has barely grown revenues since summer. | What this implies to me is that OpenAI has barely grown revenues since summer. | ↗Not briefed |