HUMANLM: Simulating Users with State Alignment Beats Response Imitation

Thu, 01 Jan 2026 00:00:00 +0000

Vicarious Offense and Noise Audit of Offensive Speech Classifiers

Wed, 01 Feb 2023 00:00:00 +0000

We ran a massive experiment: 9 different AI content moderation systems analyzed 92 million YouTube comments about US politics. The results were shocking—different AI systems flagged wildly different content as offensive, with almost no consistency. When we asked humans to label the same content, political identity was a huge factor: Democrats and Republicans disagreed strongly on what’s offensive, especially on hot topics like abortion and gun rights. This proves that “offensiveness” isn’t a fact that AI can learn—it’s a subjective judgment shaped by values. Current content moderation practices that treat one group’s perspective as “truth” are fundamentally unfair.

Computer Science - Machine Learning | Tharindu Cyril Weerasooriya

HUMANLM: Simulating Users with State Alignment Beats Response Imitation

Vicarious Offense and Noise Audit of Offensive Speech Classifiers