Attention-Economics Framing For Ai Output Quality

Issue 82 Edition 2026-03-23 4 min read

Not accepted General

Sources: 1 • Confidence: Medium • Updated: 2026-04-13 03:52

Key takeaways

In this corpus, "slop" is defined as content that requires more human effort to consume than it took to produce.
In this corpus, sending raw Gemini output to a coworker is characterized as disrespecting the recipient's time rather than exercising creative freedom.

In this corpus, "slop" is defined as content that requires more human effort to consume than it took to produce.
In this corpus, sending raw Gemini output to a coworker is characterized as disrespecting the recipient's time rather than exercising creative freedom.

What workflows are being discussed (e.g., requirements, design docs, code review notes, status updates), and how does the cost-asymmetry definition apply differently across them?
How is "effort to consume" measured (time, number of clarification cycles, error rate, rework, cognitive load), and what thresholds define unacceptable slop?
Does requiring summarization/cleanup before sharing AI output actually reduce downstream review time or errors in practice in the referenced context?
What is the baseline comparison for production effort (human-only vs AI-assisted), and does AI use change total system effort when including verification and iteration?
Are there explicit exceptions where sharing raw model output is acceptable (brainstorming, personal notes, early drafts) and how should those be labeled to avoid hidden costs?

Growing emphasis on recipient time as a cost metric could increase demand for enterprise AI features that enforce summarization, validation, and formatting before sharing outputs.
Workplace norms that label raw model output as disrespectful may drive adoption of AI governance and communication standards, including required labeling of draft versus final AI content.
If organizations operationalize slop as effort to consume versus effort to produce, vendors enabling measurement of review time, rework, and clarification cycles could see increased interest.

Organizations publish or adopt explicit policies requiring AI output cleanup, summarization, or verification before sharing internally, framed around protecting recipient time.
Pilots report measurable reductions in downstream review time, clarification cycles, or rework after introducing mandatory summarization or structured AI output templates.
Enterprise AI platforms add controls or defaults that block or warn on sharing unedited model output, or require labels indicating draft, brainstorm, or verified content.

No observable change in review time or error rates after enforcing cleanup and summarization, suggesting the norm adds process overhead without reducing recipient effort.
Teams treat raw AI output sharing as acceptable in common workflows without labeling, and no governance standards emerge beyond informal guidance.
Measurement of effort to consume proves impractical or too subjective, preventing thresholds for unacceptable slop and limiting implementation beyond rhetoric.