The AmplifierLearn
The Landscape update
The book says the brand names change every six months. This is where we check, every six months, and say what did — and what didn't.
Planned for release phase 2. The page is complete as a draft; interactive parts run in the browser and do not yet save to an account.
In brief
Each update covers the same eight categories as the tool directory, notes what entered and left, and ends with what the change means for each of the eight steps — usually "nothing", which is the point of a framework built on categories.
September 2026 — Issue 0: the baseline
Published 12 September 2026 · next issue March 2027
This is the baseline against which future updates will be read. The categories are the eight in the directory. The names are the ones the book carried in April 2026 plus what has since become a strong example of its category. Nothing in this issue changes any step of the framework.
What has moved fastest: agents. In April the book's mentor calls them "the new one — we're early in that, but it's where everything is heading." That remains true, and it remains the category where the Decision Filter's fourth question — cost of being wrong — matters most, because an agent that acts is an agent that can act wrongly. The framework's advice is unchanged: verifiable, low-cost, reviewable sequences first.
What has not moved: the categories themselves. Text-to-text, code, image, video, voice, documents, agents, vision. Every tool that has appeared in the last six months fits one of them, which is the evidence for teaching selection thinking rather than tools.
For the framework: no step changes. Step 4's five characteristics still sort every new arrival. Step 5's two-step prompting works across every conversational tool tested.