Concept — Elastic Automators
# Elastic Automators
## Definition
**Elastic Automators** are language-model-driven workflow systems that behave less like artificial minds and more like flexible automation. The term names a category that current "AI agent" framing miscategorizes: most systems labelled "AI" are not minds in any meaningful sense — they are *automation made flexible enough to negotiate language*. The achievement is real (workflows that interpret messy input, classify, call tools, retry, present coherent responses), but it is not the achievement that "AI" or "artificial mind" implies.
The diagnostic vocabulary asks of any system: *what loop is running? what tools? what memory? what criteria? what failures hidden?* — replacing the consciousness-shaped questions ("does it know," "does it want") with practitioner questions that map onto how the systems actually operate.
## Coined by
Alexandru Mares (allemaar)
## First published
2026-04-27 (Zenodo paper v1.0.0).
## Canonical artifact
- **Position paper:** [Elastic Automators: A Diagnostic Vocabulary for Language-Model-Driven Workflow Systems](https://doi.org/10.5281/zenodo.19802018) — Zenodo, DOI [10.5281/zenodo.19802018](https://doi.org/10.5281/zenodo.19802018), 2026-04-27.
- Episode: [[2026-E0029 - Elastic Automators - Why Most AI Is Not Intelligence/_metadata|E0029 — Elastic Automators: Why Most "AI" Is Not Intelligence]]
- Future paper series: planned diagnostic-vocabulary expansions, case studies, workflow taxonomy.
## Bodies that develop this concept
- [[2026-E0033 - The Moment AI Stopped Being a Tool/_metadata|E0033 — The Moment AI Stopped Being a Tool]] (2026-05-05) — names the agentic-AI cohort (Computer Use, Agentforce, Operator) as elastic automation now flexible enough to negotiate outcomes; verb-shift framing extends the diagnostic vocabulary from query to operate
- [[2026-E0037 - I Caught an LLM at the Edge of Its World/_metadata|E0037 — I Caught an LLM at the Edge of Its World]] (2026-05-13) — uses *elastic automator* as the cold-probe term in the seed observation that became the [[token-substrate-hypothesis|TSH]] paper; the coined word is the diagnostic substrate the boundary movement is measured against
- [[2026-E0038 - What the Phone Did to Work/_metadata|E0038 — What the Phone Did to Work]] (2026-05-15) — gestures at *elastic automation* as the system category coming next, without naming the term in body prose; the canonical-body close anchors at `/writing/elastic-automators`; the piece treats the concept as the load-bearing answer to a question it deliberately leaves open
- [[2026-E0039 - The Word We Needed/_metadata|E0039 — The Word We Needed]] (2026-05-18) — names *elastic automator* twice in body prose as the plain word the narrator's scared friend needed all along; the inline link points to `/concepts/elastic-automators`; the story uses Ryle's category-mistake tool to dissolve the AI-as-mind framing the term replaces — concept introduced through anecdote and category argument, not technical exposition
## Related concepts
- [[younndai|YounndAI]] — sibling work
- [[yon|YON]] — notation that elastic automators can use natively
- [[textual-kinematics|Textual Kinematics]] — different angle on the same systems
- AI agents (the term Elastic Automators reframes)
- LLM workflows
- Tool use, orchestration
- Cognitive architecture (the framing Elastic Automators contrasts with)
- [[elastic-automators|Cluster: Elastic Automators]] — parent Cluster
## Why it matters
If we design, regulate, fund, and argue about "AI" under the wrong category, we ask the wrong questions and miss the practical ones. Elastic Automators offers a category that names what the systems actually are without inflating the claim. The term's value compounds: every essay, paper, or talk that uses it correctly trains the field to think about LLM-driven workflows under a more accurate frame.
## Status
`published` — paper live, vocabulary in active use across Alex's work and YounndAI products. Continuing to extend with future case studies and the diagnostic-loop-questions framework.