# What's new

**URL:** https://aibility.org/en/docs/whats-new
**Meta description:** A dated list of changes to the methodology, the product and the documentation, newest first.

---
Changes are listed newest first. Each one has a date and what it changes for you.

## 16 October 2026: Documentation moved to aibility.org/en/docs

The Superpowered Professional documentation is now part of the aibility.org website, with new content and a new look. The texts were written again and use the same words as the website: short and detailed AI skills test, AI Workouts, coaching in your AI tool, and connecting to ChatGPT, Claude, Codex or Cursor.

## 7 October 2026: Principles and AI Workouts overview made public

The ten AI-first principles that coaching is built on and the overview of AI Workouts are publicly available on aibility.org, including a text version for AI tools. The explanations are on [Ten AI-first principles](https://aibility.org/en/docs/coaching/principles) and [AI Workouts](https://aibility.org/en/docs/coaching/ai-workouts).

## 22 May 2026: Third version of the methodology

The methodology became more precise, to better capture how people work with AI today. It cares more about how you brief AI, how you refine and check outputs, what you save and whether your work changes as a result than about which tool you use.

The result is no longer just a score. It rests on four layers: level, profile, next step and SP Score™. The score is a baseline that helps you follow progress over time, but on its own it does not say enough.

The methodology follows more closely what a person really does: concrete outputs, repeated routines, saved prompts and templates, assistants, refining and checking. It also separates more precisely using AI from changing your work. A one-off answer helps with a task, a good prompt helps again, a template or assistant helps repeatedly, and a changed way of working changes a whole area of your job. Advanced work with AI is not only for managers either: a specialist without a team can be very advanced if AI changed their key work.

Recommendations show one concrete next step in real work, for example: “Take five emails that went well, have AI find the rules of your style, and use them to build a first version of an assistant that drafts the next ones for you.”

If your score moved after the change, treat it as a new baseline under a more precise methodology. The old and new result may not match one to one. A lower number does not mean you got worse: often the new version just separates frequent work with AI from a repeatable change in how you work. An old report is never overwritten.