Guides & directories, kept current

AI tooling isn’t settling into one shape — it’s splitting into several at once. A single task today might route through a fast model for drafting, a reasoning-heavy model for planning, and a dedicated image or video model for the final asset, with an agent stitching the handoffs together instead of a person copy-pasting between tabs. That multi-model pattern, not any single chatbot, is what “using AI” increasingly means in practice, and it’s the thread running through everything published here.

Autonomous and semi-autonomous agents are the sharpest example of that shift. Coding agents that can read a repository, make a plan, and run their own edits (see our Claude Code guide below) have moved from novelty to daily workflow for a lot of developers in 2026. The same pattern is showing up in image and video generation, where an agent chains together a research step, a prompt-refinement step, and a generation step rather than a person doing each manually. Google has pushed the underlying models hard in the same direction this year — Gemini 3 now sits behind AI Overviews and Search’s AI Mode, Gemini is built directly into Chrome for page summarization and tab comparison, and Nano Banana 2 became Google’s default image model across Search, Ads, and Cloud — which is exactly the kind of fast-moving ecosystem our Gemini guide tracks in detail.

We keep this blog to a small number of deep, maintained guides rather than a high volume of short posts. Google has been consistent on this point: the deciding factor for how AI-related content is treated isn’t whether AI assisted in producing it, but whether the page is original, accurate, and genuinely useful to the person reading it, as opposed to thin pages produced at scale mainly to attract search traffic. Every guide below is written and reviewed with that bar in mind, and revised as the tools, pricing, and models they cover change.