Sources
Mario Nawfal
🇺🇸 STANFORD STUDY FINDS POPULAR AI MODELS SKEW LEFT ON POLITICAL TOPICS A new study from Stanford’s Hoover Institution finds that several widely used AI models show a left-leaning bias when responding to political prompts. Researchers gathered over 180,000 user judgments https://t.co/B4OMHWjtc1 https://t.co/DxUDO2VWPS
Entrepreneur A new study from Stanford University found that leading AI models from Google, Meta, and OpenAI lean left in their responses to political questions. OpenAI's ChatGPT was the most left-leaning of them all. Thoughts? https://t.co/TWyMrwimUa
Hoover Institution Based on a Hoover study using +180k user comparisons across 24 AI models, @JustinGrimmer, @AHall_Research, and @SeanJWestwood find that all major language models are perceived as having a left-leaning political bias. Read @CCreitzPolitics in @FoxBusiness: https://t.co/nQe4VZD4nd
