Skip to content

Archive / 2026-08-22

August 22, 2026

News

View all news →

  1. Why your local LLM feels dumber than it is

    forum.level1techs.com26 days ago510 ptsView detailsJoin discussion

  2. Munder Difflin – Agent harness to run an office of your clones

    munderdiffl.in27 days ago311 ptsView detailsJoin discussion

  3. A Friendly Introduction to Racket

    geometridae.bearblog.dev26 days ago268 ptsView detailsJoin discussion

  4. Anthropic appears to be A/B testing reduced effort levels in Claude Code

    twitter.com26 days ago213 ptsView detailsJoin discussion

  5. AI Chip Architectures

    jepeake.com26 days ago146 ptsView detailsJoin discussion

  6. Z80 – The 1970s Microprocessor Still Alive (2021)

    computer.org27 days ago149 ptsView detailsJoin discussion

  7. NanoGPT Speedrun Frontier

    primeintellect.ai26 days ago139 ptsView detailsJoin discussion

  8. Show HN: Make your logo extra bright on HDR screens

    soverybright.com26 days ago78 ptsView detailsJoin discussion

  9. Four Years Ago, a Crypto Boss Went Missing. Now His Successor Has

    nytimes.com26 days ago58 ptsView detailsJoin discussion

  10. Digging the grave of my skills: Hollywood creatives training AI to do their jobs

    theguardian.com26 days ago57 ptsView detailsJoin discussion

  11. Embedded AI

    nostarch.com27 days ago39 ptsView detailsJoin discussion

  12. Anthropic IPO filing will show AI backlash as a risk factor, sources say

    cnbc.com26 days ago37 ptsView detailsJoin discussion

  13. Software Engineering in the Agentic Era

    simonwillison.net26 days ago35 ptsView detailsJoin discussion

  14. AI has failed to win people's trust. Its makers? less trusted

    euronews.com26 days ago28 ptsView detailsJoin discussion

  15. Microsoft Blames Windows Gaming Issues on RGB Lighting Devices

    bleepingcomputer.com26 days ago22 ptsView detailsJoin discussion

  16. Dutch regulator fines Uber €825M for letting AI deactivate driver accounts

    nltimes.nl26 days ago21 ptsView detailsJoin discussion

  17. A New Kind of Science by Stephen Wolfram: Monster Raving Egomania

    bactra.org26 days ago20 ptsView detailsJoin discussion

  18. Palantir's Karp – frontier AI labs that are 'trying to drug addict us'

    cnbc.com26 days ago19 ptsView detailsJoin discussion

  19. Fast and Hard Code

    lucumr.pocoo.org26 days ago17 ptsView detailsJoin discussion

  20. In 2000, Ted Kaczynski advised against math career due to future AI progress

    twitter.com26 days ago15 ptsView detailsJoin discussion

  21. Show HN: Structural code grep across public GitHub repositories

    grep.codemod.com26 days ago15 ptsView detailsJoin discussion

  22. Htmx live is cool. Datastar is fast. This cow is raw and strong

    stario.dev26 days ago13 ptsView detailsJoin discussion

  23. 80% of developers find AI coding more addictive than helpful

    zdnet.com26 days ago13 ptsView detailsJoin discussion

  24. Linus Torvalds Endures a Debug Session from Hell, "Enormously Helped" by AI

    phoronix.com26 days ago11 ptsView detailsJoin discussion

  25. Delivery drivers will get minimum wage in Australia's 'world-first' deal

    theguardian.com26 days ago11 ptsView detailsJoin discussion

  26. Apple degrades privacy and encryption of iMessage with AI integrations

    twitter.com26 days ago10 ptsView detailsJoin discussion

  27. The crisis of AI-generated mathematics

    arxiv.org26 days ago10 ptsView detailsJoin discussion

  28. The Developer’s Guide to NeMo Guardrails for Enterprise AI Safety

    In this tutorial, we explore how to design production-grade safety for LLM-based applications using the NeMo Guardrails framework. We move beyond simple prompt filtering to implement a layered architecture, featuring deterministic PII redaction, retrieval filtering, output masking, and policy-based tool gating. By int…

    marktechpost.com26 days agoView details

  29. Decoding AI’s Open-Source Course Maps Three Ways to Run an Agent Loop and the Provider Economics Behind Each

    Most teams treat ‘which model’ as the important decision. The harness engineering literature keeps pointing somewhere else. In LangChain’s Terminal-Bench experiment, changing only the harness—same model throughout—moved a coding agent from roughly 30th place into the top 5. That result reframes the question. If the ha…

    marktechpost.com26 days agoView details

Models

View all models →

  1. sensenova/SenseNova-U1.5-8B-MoT

    any-to-any · transformers · safetensors · neo_chat

    huggingface.co27 days ago196 ptsView details

  2. lylogummy/Anima-3.8B

    region:us

    huggingface.co26 days ago75 ptsView details

  3. fastino/gliner2.5-multi-v1

    token-classification · gliner2 · safetensors · extractor

    huggingface.co26 days ago67 ptsView details

  4. lvladikov/Krea2-Turbo-Distill-4step-LoRA

    text-to-image · diffusers · lora · text-to-image

    huggingface.co27 days ago16 ptsView details

  5. MohamedAhmedAE/llava-medical-8B-clip-vit-stage2

    safetensors · llava · region:us

    huggingface.co27 days ago5 ptsView details

  6. TAOTAO777/ai-girlfriend-natsume

    safetensors · gguf · endpoints_compatible

    huggingface.co27 days ago3 ptsView details

  7. mradermacher/Neuron-46x4B-i1-GGUF

    transformers · gguf · Neuron-46x4B-Instruct

    huggingface.co27 days agoView details

  8. qf-iquest/MyAwesomeModel-TestRepo

    feature-extraction · transformers · pytorch · bert

    huggingface.co27 days agoView details

  9. mradermacher/CalibForge-35B-A3B-i1-GGUF

    transformers · gguf · en

    huggingface.co27 days agoView details

Open source

View all open source →

  1. ggml-org/llama.cpp b10588

    <details open> common: json.h: fix clang lto (#27575) </details> **Website:** - <https://llama.app> **Attestations:** - <https://github.com/ggml-org/llama.cpp/attestations/42368523> **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/dow…

    github.com26 days agoView details

  2. ggml-org/llama.cpp b10577

    <details open> common : fix draft-mtp with embeddings (#26352, #27299) (#27400) * common: fix draft-mtp with embeddings (#26352) * --whitespace --------- Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@huggingface.co> </details> **Website:** - <https://llama.app> **Attestati…

    github.com27 days agoView details

  3. ggml-org/llama.cpp b10576

    <details open> sycl : add Q2_K reordered MMVQ and ESIMD kernels (again) (#27490) * Revert "Revert "sycl : add Q2_K reordered MMVQ and ESIMD kernels (#26336)" (#…" This reverts commit 7a0e42fd01fb0acda644e4f04b1f1acbbb9e23ba. * add gate params </details> **Website:** - <https://l…

    github.com27 days agoView details