LLMMedium
Philosophy AI Tentatively Produced Offensive Results for Certain Prompts
- Occurred
- Sep 15, 2020
- Published
- Sep 24, 2020
- Source
- AI Incident Database
Summary
Philosopher AI as built on top of GPT-3 was reported by its users for having strong tendencies to produce offensive results when given prompts on certain topics such as feminism and Ethiopia.
Developers
- Murat Ayfer
- OpenAI
Deployers
- Murat Ayfer
Harmed parties
- historically disadvantaged groups
Classification
Category confidence: 90%
Severity confidence: 85%
Rationale: GPT-3 based AI produced offensive content on specific topics, indicating a language model failure.
Classifier: Qwen/Qwen3.5-35B-A3B-FP8