Patterns
Take the quiz: Human-in-the-loop or something else? This Pattern appears when the work output appears to be completed but the decision to release the work takes much longer. AI and Automation now makes the contrast more obvious.
the-chaos
The Blue blob stumbles across a Bubble debate but there also appears to be a lot of stuff inside the bubble not getting attention.
Signals
The Bubble Is Rarely The Whole Technology: The lazy question is whether AI is a bubble. The better question is which layer is inflated, which layer is useful, and who is left cleaning up the difference.
Guides
Bubble Layer Checker: A practical way to avoid lazy “bubble or not” debates by checking which layer is real, which layer is inflated and which layer creates work later.
the-chaos
There doesn't appear to be a difference in usage between the newest floor and the oldest ones.
Guides
Vendor Layer Memory Guide: A practical guide for checking what old systems, old data and old workarounds a shiny new AI or vendor layer is about to inherit.
Signals
New Layers Inherit Old Mess: A new AI layer does not float above the business. It lands on everything the business already built, ignored, customised or forgot.
Responsible use needs examples, not just a poster. Allowed AI Use Rules Starter helps turn the policy intent into practical examples people can follow.
This guide helps managers, HR teams, universities, early-career workers and team leads check whether a junior role is still designed for learning because AI may have changed the task mix.
If humans are babysitting the AI, that time belongs in the AI cost story. AI is now common enough that a lot of hidden work is becoming the new normal.
The boring table that enables upgrades, automation, and AI integration into workflows. Tools don’t equal capability. If you can’t list it, you can’t govern it. First choose an area of friction where you may be thinking of adding AI layer or other upgrades, and run these steps.
Technology changes but Patterns return. Take a quiz, follow the work and find a useful Guide.
The course may be complete. The workplace may still have locked cupboards. AI literacy needs work rules, not only training slides.
AI Literacy Is Knowing What You Are Allowed To Do At Work: A course can explain AI. Work still needs rules people can actually use on Tuesday.
Role-Based AI Literacy Evidence Log: A light evidence log for showing that people understand the AI rules for the work they actually do, not just the training slide they once saw.
The Blue blob sees worker trying to decide between multiple AI help desks. The help desks are willing to respond to any question.
AI Tool Duplication Check: A simple check for spotting when your AI toolkit is helping the work, and when the same question is just touring the building.
The Toggle Tax Is Real: When people ask the same question in three AI tools, the work has not disappeared. It has started travelling.
The Blue Blob found an agent traffic jam with lots of robots moving around. They appear to be very busy with various tasks.
Agent-to-Agent Interaction Checklist - Before agents talk to each other: List every agent; Define actions; Map conversations; Set stop points; Check loops; Name human owner
Agent Traffic Is Still Traffic: Agent traffic is still traffic. If many agents can move work, someone still needs the map.
The Blue Blob found a trainee desk with the human trainee waiting for work while the AI is being trained.
Apprenticeship Preservation Check - Before automating beginner work: Name the learning task; Decide assist or replace; Create practice loops; Protect low-stakes mistakes; Update entry expectations
Automation Can Remove The Practice Ground: A task can be boring for an expert and essential for a beginner.