Cope Scores

Lab Framing Analysis

OpenAI: "ChatGPT Signals Shows Healthy Adoption"

HEAVY • 68

Displacement data framed as adoption. "Doing" presented as a feature. Geographic concentration called "global reach." The word "displacement" never appears.

Anthropic: "No Systematic Increase in Unemployment"

MODERATE • 52

Measured unemployment (lagging indicator) instead of hiring rates (leading indicator). Found hiring slowdowns but headlined the unemployment non-finding. Better than OpenAI but still cope.

Google: "AI Will Create More Jobs Than It Displaces"

TERMINAL • 85

Zero published displacement data. Unfalsifiable "new jobs" claim. Refuses to measure what OpenAI and Anthropic at least attempted.

Common Cope Patterns

"It's only 31% work-related"

TERMINAL • 82

The Volume Illusion. 31% of an exponentially growing base = MORE absolute work displacement than 51% in 2024.

"AI creates more jobs than it destroys"

TERMINAL • 88

165 work activities measured being automated. Zero data on new jobs. Unfalsifiable cope.

"People are just using it as a better search engine"

HEAVY • 72

46% of work messages are "doing." Only 34% are "asking." OpenAI's own data demolishes this.

"Workers are becoming more productive"

MODERATE • 45

True short-term. But productivity gains with a tool that works independently makes the human the bottleneck, then the overhead, then redundant.

"AI is a complement, not a substitute"

HEAVY • 65

Collapses when the "complement" performs 46% of work tasks independently. A complement assists. A substitute replaces. Doing = substitute.