Loom 0.7: When the Answer Is on Page 40 of a PDF
Loom 0.7 fixes its most annoying limit with long documents: instead of the first 8000 characters, the LLM now gets the passage that actually answers the question. Plus a loom_grep MCP tool for exact …
Artificial Intelligence
Language models, agents, local AI and everything that is changing the way we work. Here you’ll find analysis of what’s new (minus the hype), hands-on guides for running models on your own computer and a few thoughts on where we’re heading.
They Cloned Photoshop with AI. Is This the End of Software as We Know It?
In just a few days, open-source clones of Photoshop, Illustrator and Premiere written with AI showed up on GitHub, along with browser versions of Halo and GTA derived from the original games. …
The Day After: AI Companies Prepare for the Worst (While Others Say It Won't Happen)
According to Axios, Anthropic, OpenAI and other AI companies are privately rehearsing the “day after” an AI-driven catastrophe. On TV, Zeta Global’s founders call it an exaggeration …
Loom Grows Up: Optional Hybrid Semantic Search, Multilingual Retrieval, and an HTTP Server
Loom gets an upgrade: optional hybrid semantic search (BM25 + embeddings) arrives to handle the cases where keyword search alone falls short - paraphrases and multilingual queries. On top of that, …
Anthropic, Fable 5 and Mythos 5 Blocked by the US Government Over a Jailbreak
The US government blocked foreign access to Anthropic’s Fable 5 and Mythos 5, citing national security and an alleged jailbreak. Anthropic is complying, but openly disputes the decision. Here is …
DeepSeek V4 vs NVIDIA: Huawei Ascend, Costs and Architecture
I used DeepSeek V4 for a few days and the verdict is clear: excellent performance, ridiculously low costs, and an architecture that should make Jensen Huang nervous. Here is everything you need to …
Loom: Local LLM Memory with MCP Server and CLI, No Docker
Loom is an open source tool written in Go that turns your files into a memory layer LLMs can query. Inspired by Karpathy’s llm-wiki pattern, but with its own architecture: precomputed summaries, …
GitHub Will Train Copilot on Your Code: Opt Out by April 24, 2026
Starting April 24, 2026, GitHub will use Copilot interaction data to train its AI models. If you do not opt out, your code becomes training material. Here is what is happening, how to protect …
ContextDoc: Your Coding Agent Finally Stops Breaking Things
ConceptDoc grew up, changed its name, and — most importantly — learned to stay quiet when it has nothing useful to say. Discover how ContextDoc, .ctx files, and the new ctx-watch and ctx-run tools …
Coding Agent Prompts: A Guide to the Perfect Prompt
Want a coding agent to write truly good code? Giving it orders isn’t enough. Learn how to guide it with the right prompt, a clean codebase, and a few tricks from a seasoned AI tamer.
OpenClaw – Complete Guide, Part 3: First Missions and Cost Optimization
Part three of the OpenClaw guide: how to define the agent’s mission, use ClawHub skills, automate with cron jobs, and manage costs with a multi-model strategy.
OpenClaw Channel Setup: Telegram, WhatsApp and Security
Part two of the OpenClaw guide: how to connect Telegram and WhatsApp, configure the allowlist, protect the dashboard, and defend against prompt injection.