A field guide to people working with AI
When humans and AI systems meet, work, learn, and shape each other, recurring patterns appear. This research programme by Andreas Ehstand names and describes those patterns — a careful, bilingual inventory of what is observed. Nothing prescribed, nothing sold.
What is this, really?
Three short answers. Skip to whichever fits how you read.
In one sentence
It is a research library of 1,128 named patterns that show up when a human and an AI system interact — in classrooms, in workplaces, in robotics labs, in everyday use.
In three sentences
Every page is one named pattern, described in plain English and German, written from observation rather than prescription. The patterns are sorted into 165 topical clusters and six thematic branches. You can browse the universe five different ways below — pick the one that matches how your mind likes maps.
What it is not
- It is not medical, legal, therapeutic or financial advice.
- It is not a commercial product or service.
- It is not a recommendation about what you should do.
- It is descriptive research, freely citable, attribution-required, non-commercial.
Who is it for
- Researchers who need a vocabulary for human-AI phenomena.
- Educators looking for shared language to talk about AI in class.
- Engineers and product designers who need precise terms for what users actually do.
- Anyone curious how a working coach started naming what they saw.
The Universe — explore by feel
Same nine corpora, five different visual maps. Click a node in any view to open the corresponding corpus. Try several — every map shows a different aspect.
What you can open
Nine independent collections you can read or download. Each opens in its own page.
Master Compendium
1,128 defined terms · bilingual EN+DEThe main library. One page per named pattern, each with a description, related terms, and disclaimer. Browse alphabetically or by topic.
Open the main libraryCompanion atlas
~2,141 academic pagesA second collection of term pages in a more academic format, with extra machine-readable export files.
Open companionPerformance framework
30 performance factors · interactiveAn interactive map of performance factors that apply equally to people, AI agents, robots and teams of any mix.
Open frameworkPeriodic table
100 selected terms · Mendeleev-styleA hundred of the most representative patterns, arranged like a periodic table. Hover for definitions.
Open periodic tableFounding essay
Founding methodological essayA short founding essay (in German) on collaborative thinking between humans and language models — the methodological grammar behind the whole library.
Read essaySister atlas
Companion landingA lighter sister entry point that mirrors the main library structure.
Open sisterKnowledge-graph view
100 core terms · graph perspectiveA knowledge-graph view of the core 100 patterns.
Open graph viewAbout the author
Background & affiliationsAndreas Ehstand's research background and academic anchors.
Open author pageThe branches of the work
Six thematic lenses that organise the library. Click any to open the topic page in the main library.
Core library
The descriptive baseline — the broadest collection of human-AI interaction patterns.
Open topic →Cross-references
The relationship layer — how patterns link to and depend on each other.
Open topic →Performance patterns
How performance factors from elite sport translate to humans, AI agents, robots and mixed teams.
Open topic →Robotics & teams
Multi-robot coordination, human-robot team dynamics, multi-agent orchestration patterns.
Open topic →Learning & teaching
Phenomena that show up when students and teachers work with AI tools.
Open topic →Work patterns
How everyday work — meetings, writing, planning, decision-making — shifts when AI is in the loop.
Open topic →Start with the main library.
If you only open one thing, open this. Browse by topic, by letter, or use the search.