Definition
Token spin is token-consuming AI activity that produces insufficient learning, accepted work, or maintained value for its total cost. Common sources include idle agents, over-frequent scheduled jobs, repeated empty compaction, duplicated context, oversized tool results, unchanged retries, automations whose output nobody uses, and agents continuing after their strategy has plainly failed.
Gaspar's broader operator taxonomy separates tokens that teach, tokens that produce, and tokens that spin. Failed experiments can belong in the first category when they create reusable learning. A successful-looking automation can drift into spin when nobody consumes its output or its cost exceeds its maintained value.
Distinguish it from nearby terms
Token spin is not every failed run and not every expensive run. The term asks whether the activity creates learning or an accepted outcome. Token maxing describes an incentive or strategy that can cause spin; token burn is only the measured consumption.
Operator checks
- Weekend test: does spend continue when nobody is using the system?
- Output-use test: has anyone used the scheduled artifact during its review window?
- Input-output imbalance: is the system repeatedly ingesting large contexts while producing almost nothing?
- Changed-strategy test: after failure, did the next turn change evidence, method, or constraints?
- Spin-to-production ratio: what share of total consumption reaches accepted work, after preserving an explicit learning budget?
Check your understanding
Do not eliminate exploratory tokens merely because they shipped nothing. Eliminate recurring spend that produces neither accepted work nor retained learning.