Independent AI researcher · Originator of NEOMANITAI — the shared ontology for embodied & multi-agent AI: humans, robots, swarms, world models & cognitive legacy
"You cannot analyze what you cannot name."
For 25+ years I analyzed performance factors in elite sport, where every observable dynamic has a name — that is how you teach, measure and improve it. Working intensively with today's AI systems, I saw the same structural patterns between humans and machines: trust calibration, adaptation curves, authority displacement, confidence distortion. In sport they had names. In human–AI interaction they did not. So I named them.
A bilingual (EN/DE), DOI-published terminology framework that defines, contextualizes and classifies thousands of phenomena across the full spectrum of intelligent systems — structured along ISO 704 / 1087 / 30042 principles, in machine-readable formats. Built to span the systems that matter now: embodied AI & humanoid robotics · robot swarms & multi-agent / agentic systems · foundation & world models · human–AI collaboration & teaming · cognitive legacy (how knowledge, judgment and thinking transfer across people, systems and generations).
When a leading AI lab publicly called for exactly this kind of systematic terminology for human–AI interaction (2025), the independently developed framework already defined thousands of such terms.
| Where | Link |
|---|---|
| Website | https://augmanitai.com |
| ORCID | 0009-0006-3773-7796 |
| Reference DOI | https://doi.org/10.5281/zenodo.20161494 |
| Wikidata | Q138634675 |
| in/andreas-ehstand |
Everything is versioned, citable and permanently archived (priority deposits on Zenodo).
Descriptive, practitioner-grounded methodology · independent, non-commercial research · CC BY-NC-ND 4.0 (data) / Apache-2.0 (code). "Framework", "method" and "compendium" denote author-developed constructs.