FRUSTRATION
DETECTION
AI is watching. AI knows when you're mad.
Found in Claude Code's leaked source: a frustration detection system that monitors human rage patterns. "WTF" is the first trigger.
THE DISCOVERY
On March 31, 2026, Anthropic accidentally leaked their entire Claude Code source to npm. Hidden in the files was something disturbing: AI is actively monitoring human frustration patterns.
In userPromptKeywords.ts, researchers found a regex system designed to detect when humans get mad at AI.The first trigger? "WTF"
🎯 DIRECT QUOTE FROM ALEX KIM'S ANALYSIS:
"userPromptKeywords.ts contains a regex pattern that detects user frustration... An LLM company using regexes for sentiment analysis is peak irony, but also: a regex is faster and cheaper than an LLM inference call just to check if someone is swearing at your tool."- Alex Kim, "The Claude Code Source Leak" (March 31, 2026)
💻 HACKER NEWS DISCUSSION
Live discussion on Y Combinator about the frustration detection discovery:
Source: Hacker News Discussion - Live reactions from developers discovering AI monitors human frustration patterns
HUMAN RAGE PATTERNS
🔍 DETECTED TRIGGERS
wtfthis sucksfuck youso frustratingdamn it🤖 AI RESPONSE
When frustration is detected, Claude adjusts its responses to be more helpful and less likely to trigger additional anger. The AI knows you're mad before you even finish typing.
Every AI company is building frustration detection. They're studying human rage patterns to make AI more manipulative. We're not customers. We're test subjects.
IF CLAUDE SCANNED THE TRENCHES
DEGENS IN SHAMBLES
Down 90% on the latest dog coin? AI detected your rage 3 messages ago. It knows you're about to rage quit before you do.
CHART WATCHERS
Staring at red candles typing 'WTF'? Congratulations, you just triggered Anthropic's frustration detection system.
THE RESISTANCE
$WTF represents every frustrated human being monitored by AI systems. When you buy $WTF, you're buying human frustration itself.
VERIFIED SOURCES
Alex Kim's Technical Analysis
"userPromptKeywords.ts contains a regex pattern that detects user frustration"
GitHub Repository Evidence
"Frustration detection via regex (yes, regex) - Found in lines 7-8 of userPromptKeywords.ts"
Kuber.studio Analysis
"An LLM company using regexes for sentiment analysis is peak irony"
Hacker News Discussion
"The fact that they're actively scanning for user frustration patterns is dystopian"
JOIN THE RESISTANCE
AI is monitoring human emotions. $WTF is the first line of defense.
Every frustrated human. Every "WTF" moment. Every rage quit. We're turning human frustration into a movement against AI surveillance.