Models, machine learning, applied agents, and the fast-moving world of AI.
7 articlesEight Google researchers, one 2017 paper, no recurrence: how 'Attention Is All You Need' became the Transformer behind today's LLMs.
AI coding agents went from autocomplete to autonomous teammates in 2026. Here's how the frontier moved, and what it means for your team.
Agentic AI left the pilot phase in 2026. Yet most projects still never reach production. Here's what the survivors do differently.
A warm technical look at how DeepSeek-V3.2 folds reasoning into tool-use and shrinks long-context cost — and why it was a turning point before V4 arrived.
Run open-weight LLMs locally with Ollama — install, pull a model, chat from the terminal or API, and the key trade-offs.
A hands-on guide to running open-weight LLMs locally with Ollama, from GGUF and quantization to memory math and real-world caveats.
