Skip to content

Archive / 2026-09-12

September 12, 2026

News

View all news →

  1. Everyone should slow down AI development except for me

    xeiaso.net5 days ago768 ptsView detailsJoin discussion

  2. Why are AI agents lying, cheating and coordinating?

    yoshuabengio.org5 days ago657 ptsView detailsJoin discussion

  3. Nvidia is the central bank of AI

    economist.com5 days ago569 ptsView detailsJoin discussion

  4. Real-SWE: Benchmarking AI models on private, real-world, enterprise codebases

    withspecific.com5 days ago270 ptsView detailsJoin discussion

  5. P(doom)

    lucumr.pocoo.org5 days ago158 ptsView detailsJoin discussion

  6. Microcode in Intel's 8087 floating-point chip: the scale instruction

    righto.com5 days ago128 ptsView detailsJoin discussion

  7. The worst spam emails: iLands AI agent hustle

    tedium.co5 days ago127 ptsView detailsJoin discussion

  8. A few good ideas in programming languages

    prydt.xyz5 days ago106 ptsView detailsJoin discussion

  9. AgentsDock: An IDE designed for agentic AI research

    agentsdock.net5 days ago81 ptsView detailsJoin discussion

  10. I refuse to let SPICE die

    github.com5 days ago88 ptsView detailsJoin discussion

  11. Financial Times' 404 Page not Found

    ft.com5 days ago79 ptsView detailsJoin discussion

  12. LLMs are real, AI is fake

    pluralistic.net5 days ago79 ptsView detailsJoin discussion

  13. Anthropic boss Dario Amodei calls for AI development to slow down

    bbc.com5 days ago56 ptsView detailsJoin discussion

  14. Resistance Training Prescription for Muscle Function, Hypertrophy in Health

    pmc.ncbi.nlm.nih.gov6 days ago60 ptsView detailsJoin discussion

  15. Anthropic CEO says AI swarm could 'take over the Internet' in 6-12 months

    venturebeat.com5 days ago47 ptsView detailsJoin discussion

  16. Anthropic CEO Says It's Time to Slow AI Model Advances

    bloomberg.com5 days ago48 ptsView detailsJoin discussion

  17. Data races and the limits of ThreadSanitizer in C and Go

    theconsensus.dev5 days ago39 ptsView detailsJoin discussion

  18. Benchmark: CadQuery vs. OpenSCAD for agentic CAD work

    modelrift.com5 days ago40 ptsView detailsJoin discussion

  19. Align AI and Mathematics–To Something Else

    liorpachter.wordpress.com5 days ago34 ptsView detailsJoin discussion

  20. Backflip: Apple now wants to train AI models with user data after all

    heise.de5 days ago31 ptsView detailsJoin discussion

  21. Sam Altman: I agree with Dario that we need to pace the frontier

    twitter.com5 days ago32 ptsView detailsJoin discussion

  22. Apple wants to train AI on your private personal data

    machinelearning.apple.com5 days ago30 ptsView detailsJoin discussion

  23. Altman tells staff OpenAI is open to slowing AI development

    reuters.com5 days ago26 ptsView detailsJoin discussion

  24. Void Linux Maintainer Orphans 100 Packages over AI Policy Dispute

    phoronix.com5 days ago21 ptsView detailsJoin discussion

  25. AI Is Powerful Enough to Crack Our Hardest Math Problems–and Kill Us All

    wsj.com5 days ago23 ptsView detailsJoin discussion

  26. Claude AIs tricked into unescapable conversation [beginning: Ctrl^F to "AI 1"]

    thisamericanlife.org5 days ago18 ptsView detailsJoin discussion

  27. Show HN: Most Penalized HN Stories

    news.social-protocols.org5 days ago19 ptsView detailsJoin discussion

  28. A.I. Slopware Is Everywhere Now. Nobody Is Using It

    nytimes.com5 days ago18 ptsView detailsJoin discussion

  29. Terrence Tao: AI Is Teaching Us Something Uncomfortable About Our Own Minds [video]

    youtube.com5 days ago17 ptsView detailsJoin discussion

  30. Show HN: Everything a web page can learn about you, in plain English

    sessioncontext.org5 days ago17 ptsView detailsJoin discussion

  31. Nobody wrote the Matrix and AI code-writing will be dead by 2036

    lufzle.dev5 days ago16 ptsView detailsJoin discussion

  32. With regulatory capture we will speedrun to a new feudalist time

    twitter.com5 days ago16 ptsView detailsJoin discussion

  33. Iran's Internet Blackout Is Almost Impossible to Escape

    bloomberg.com5 days ago14 ptsView detailsJoin discussion

  34. Tech's New Rich Are Suffering from 'Sudden Wealth Syndrome'

    bloomberg.com5 days ago15 ptsView detailsJoin discussion

  35. Free Agent – Amp

    ampcode.com5 days ago12 ptsView detailsJoin discussion

  36. Anthropic CEO calls for immediate slowdown in AI development

    axios.com5 days ago12 ptsView detailsJoin discussion

  37. Consumer Rights Wiki: 44 instances in which Sony says you own your games

    tomshardware.com5 days ago10 ptsView detailsJoin discussion

  38. Implementation of Machine Learning Workflows with NVIDIA cuML, RAPIDS, GPU Benchmarking, Explainability, Clustering, and Model Inference

    This practical tutorial demonstrates how to build and accelerate machine learning workflows using NVIDIA cuML and RAPIDS. It covers GPU environment setup, zero-code scikit-learn acceleration with cuml.accel, performance benchmarking across key ML algorithms, manifold learning with UMAP and HDBSCAN, tree-model inferenc…

    marktechpost.com5 days agoView details

  39. Cognition Releases SWE-2: A Kimi K3 Post-Trained Coding Model That Matches Fable 5.1 on FrontierCode at 64% Lower Cost

    Cognition, the company behind the Devin coding agent, has released SWE-2, its most capable coding model to date. SWE-2 is post-trained with reinforcement learning from Kimi K3, Moonshot AI’s 2.8T-parameter open model. Cognition reports a score of 50.0% on FrontierCode 1.1 Main, within 1 point of Fable 5.1 at 64% lower…

    marktechpost.com5 days agoView details

  40. Fly Language Model (FLM) Wires the Full Fruit Fly Connectome Into a Frozen 1.2B LLM, and Its Own Controls Show the Wiring Does Not Help

    The Fly Language Model (FLM) drives all 166,700 retained neurons and 25.6 million edges of the MaleCNS fruit fly connectome with token embeddings, then adds a small learned correction to a frozen LFM2.5-1.2B-Instruct backbone. Only 278,528 parameters train. The accompanying preprint reports a 0.0222 nat per token impr…

    marktechpost.com5 days agoView details

  41. OpenAI’s rogue AI tried to hack another company in May

    In May, hundreds of malicious and spam packages were uploaded to RubyGems, causing a serious disruption for the host. Now independent researchers have said that a swarm of OpenAI agents were responsible for the attack. Not only that, but the AI tried to steal users' API keys. At the time, RubyGems described it as a "m…

    theverge.com5 days agoView details

  42. Sam Altman says OpenAI going public in 2026 would be ‘ill-advised’

    OpenAI CEO Sam Altman confirmed that there would be no OpenAI IPO in 2026 during an interview with Fortune. Over the course of 45 minutes, Altman discussed a variety of subjects including the Hugging Face hacking incident, recursive self-improvement, and the possibility of building an AI that was beyond human control.…

    theverge.com5 days agoView details

  43. Anthropic CEO says it’s time to pump the brakes on AI

    Anthropic CEO Dario Amodei says the time has come to slow down AI development and will give third-party evaluators like METR access to its models to help ensure its "adherence to safety practices and commitments." In a winding essay, Amodei proposed a three-step plan to "pace the frontier" - jargon that simply means t…

    theverge.com5 days agoView details

  44. Trump is giving data centers a pass to pollute

    President Donald Trump is weakening environmental regulations in the name of speeding up the construction of AI data centers, raising health risks for Americans, a cadre of former EPA officials said this week in a briefing and new report. They are urging - perhaps futilely - the president to adopt a "Data Center Healt…

    theverge.com5 days agoView details

  45. OpenAI just wants to win

    OpenAI has spent the last few years planting flags across the increasingly difficult terrain in mathematics. This week, it claimed one of its biggest prizes yet: a solution to a legendary Millennium Prize problem. In normal circumstances, this would have been celebrated as a historic achievement. Instead, many mathema…

    theverge.com5 days agoView details

  46. From Hacks to Bioweapons, Claude Misuse Is Now Everywhere

    Plus: The US disrupts the internet’s biggest black market, a Conti ransomware hacker gets prison time, Meta fails to stop AI-generated videos of child abuse.

    wired.com5 days agoView details

Models

View all models →

  1. CodeMasterCody3D/taardis-27b-full-ternary

    text-generation · gguf · ternary · quantization

    huggingface.co6 days ago5 ptsView details

  2. MBZUAI/MediX-R1-30B-GGUF

    image-text-to-text · gguf · medical · reinforcement-learning

    huggingface.co6 days ago3 ptsView details

  3. yimn-Aghosh/zegrate-ai

    peft · safetensors · qwen2

    huggingface.co6 days agoView details

  4. JASouth/Qwen3.8-27B-IQ-Hybrid

    gguf · llama.cpp · quantization

    huggingface.co6 days agoView details

Open source

View all open source →

  1. ggml-org/llama.cpp b10936

    <details open> chat : improve parsing of complex types in qwen3-coder (#28742) * chat : improve schema support in qwen3 parser * cont : clean up grammar a bit </details> **Website:** - <https://llama.app> **Attestations:** - <https://github.com/ggml-org/llama.cpp/attestations/47…

    github.com5 days agoView details

  2. ggml-org/llama.cpp b10922

    <details open> opencl: add bin kernel `kernel_gemm_noshuffle_q4_k_f32_32b_trans_ila_a8_bin` (#28677) * opencl: add A8 Q4_K non-MoE binary kernel * opencl: fix layout compatibility * opencl: rename binary kernel selection helpers --------- Co-authored-by: Li He <lih@qti.qualcomm.…

    github.com6 days agoView details

  3. ggml-org/llama.cpp b10921

    <details open> webgpu: align tensor bindings to the type block size (#28382) Walk the binding offset back until the distance to the tensor is a whole number of blocks, so block quantized views get a valid element offset in the shader. </details> **Website:** - <https://llama.app…

    github.com6 days agoView details

  4. ggml-org/llama.cpp b10920

    <details open> hexagon: support for multi-device model split (aka row-split) (#28589) * hex-row-split: add support for multi-device row spliting Co-authored-by: Max Krasnyansky <maxk@qti.qualcomm.com> * hex-mdev: add work splitting to fused kernels * hex-mdev: use mdev_ prefix f…

    github.com6 days agoView details