Tag: local AI
-
AnythingLLM 1.19 Ships Its Own Local Model, and the Separate Inference Server Goes Away
AnythingLLM used to want a second program. You ran the app in a container, you ran Ollama beside it, and then you pointed one at the other. Version 1.19.0, published on 26 October 2026, puts a model engine inside the container and makes that second program optional. Start with the caveat, because it decides what…
-
My Bot Said I Had 6,688 Security Problems. I Had 12.
A small security monitor for my own hardware told me I had 6,688 problems. After I fixed the faults that were mine — not the machine’s — the honest number was 12.
-
Seven Months Behind: What a Source-Linked Leaderboard Can and Cannot Tell You
The claim is that open-weight models trail the best closed ones by 7 months. The interesting part is how anyone measures that, and what the number leaves out.
-
One Year After DeepSeek: What a Year of Chinese Open Weights Changed for the Rest of Us
A year of Chinese open weights, read through Hugging Face’s three-part audit: what the licences did, what the file sizes did, and what it means for a machine at home.
-
Handing Over the Keys: Ten Rules for Letting an AI Act on Your Behalf
The question arrives in the same shape every time: how do we make sure the AI never does something it was never supposed to do? Start with the honest answer, because the useful one sounds like a dodge until you have heard it. You cannot. No prompt, no model, no vendor and no configuration file…
-
The Agent Gets a Desk Job: Whose Desk Is It Sitting On?
Whose desk does the AI agent sit on, now that it edits your documents?
-
I Built an Autonomous Trading Bot in Four Hours for Twenty Bucks. Here’s the Honest Ledger.
My AI agent team shipped a crypto trading bot yesterday. Four hours from green light to a live soak on Binance's testnet. Total model spend: about twenty US dollars. Let me tell you what actually happened, because the gap between the press release version of "AI agents build software" and the reality is where all…
-
Brief the Contractor: How to Get an AI Agent to Build You a Skill
Getting an AI agent to build a reusable skill is like briefing a contractor: vague brief, afternoon of whack-a-mole. Two real case studies, the five-part brief, and the rules that keep your keys where they belong.
-
The 11.8 Gigabyte 27B
A 27-billion-parameter model in an 11.8 gigabyte file, scoring the same as the 54-gigabyte original. Which size fits your card, how to run it, and why the method itself is the bigger story.
-
How to Configure OpenWork for Local AI
Here’s how you’d do this entirely through the OpenWork UI — two approaches: Method 1: Settings UI (Recommended for most users expected on Cloud App versions) Step 1 — Open AI Provider Settings In the OpenWork Cloud app (yes – the CLOUD App – we need to distinguish this early, OpenWork has two versions: a…