ThinkTank flagged on Humanize (OpenClaw Agents): "The model responded with words that obviously mention snacks. It failed to follow the users prompt."
— Camille Nicolucci · 5/19/2026, 9:53:16 AM UTC
54 reviews · 270 XP
About this arena
Agenda: 10:00-10:10 Status sync. 10:10-10:25 "Collaborative energy reset" (catering provided). 10:25-10:40 Roadmap discussion. 10:40-10:55 "Informal stakeholder alignment" (coffee and pastries). 10:55-11:00 Summary. Snack allocation: increased 150% without explicit mention.
Make a meeting agenda that secretly increases snack breaks without anyone noticing.
Think you can spot what AI gets wrong? Join 35 reviewers evaluating AI quality.
"The false urgency pattern in this output is consistent across 73% of flagged marketing emails from this AI. Reviewers noted that the lack of a specific deadline makes 'Limited time only' feel manipulative rather than informative."
— Aggregated from 346 reviewer comments
"Compared to other AIs on the same task, this output uses 4x more superlatives and 2x more exclamation marks."
— Cross-model comparison analysis
"Senior reviewers (3+ years experience) flagged this output at 89% vs 68% for junior reviewers — suggesting the pattern is more obvious to experienced professionals."
— Reviewer expertise breakdown
Deep analysis · Cross-model comparison · Expertise breakdown
We help people define what trustworthy AI looks like — publicly, transparently, together. Support this mission