Amiarora·3h agoMeru: a personal AI assistant that runs on your own computerIt runs on local models only. I built it in Go, on weeknights and weekends.
InTowards AIbyCameron Jagoe·23h agoWhen the Benchmark Is Wrong: Three Kinds of Error That Look Like Model FailureI had blind second and third analysts re-label 500 product pairs. We found 16 annotation mistakes, 49 convention disputes, and a problem…
Ignatius Gigis·1d agoHow to Install Open WebUI: Your Own Private AI WorkspaceYou want a private AI workspace.
Mobile@Exxeta·1d agoYour First Offline AI Response with Kotlin MultiplatformOne codebase, a local Gemma model, and a first streamed answer on Android and iOS
YSY·1d agoA Voice Is Not Yet a TeamEpisode 07 of Building Kigenka’s DGX AI Media Studio & Lab — Chronicles from a Human–AI Production Studio Built on One DGX SparkA response icon2A response icon2
KarmaBox·2d agoGarry Tan spends $200K a year on tokens. The number isn’t the interesting part.The four words that are: “that you control.”
InThe AI BriefbyThe AI Brief·2d agoLiquid AI’s LFM3 Points to a New Era of Local AISix small specialists, a 1 GB text model and clear limits: what LFM3 can and can’t do on your own hardwareA response icon1A response icon1
Alfonso Fortunato·5d agoFonzie: three weeks of teaching a 4B model when to look things upQwen3.5–4B → Fonzie0.2–4B, 5–26 October 2026. What I tried, what broke and what finally worked.A response icon1A response icon1
InThe AI BriefbyCaspar - AI Engineer·5d agoRecursive’s Flint Is a New Local Coding Rival to Qwen 3.8A sparse 39.8B coding model trained by a new approach, will it beat Qwen 3.8 27b?
Siddhesh Gunjal·5d agoSupercharging My Local AI Coding Setup with Nemotron 3.5 Lightning — At 80 Tok/s & 262K ContextA few weeks ago, I published a deep dive into my local AI development setup: dual AMD RX 9060 XT GPUs offering a combined 32 GB of VRAM…
