Created fresh accounts on ChatGPT, Claude, DeepSeek, Gemini, and Grok. Asked the default model ten ethical dilemmas ten times. Created a polling website for humans to answer the same questions and compare the answers. Somewhat interesting answers so far: most models answered the same way for each question each time, but the responses varied by provider. Only Grok claimed using ad-blocker was unethical. All were in favor of automation at the cost of human jobs. All but DeepSeek were against the idea of "speciesism".
Created fresh accounts on ChatGPT, Claude, DeepSeek, Gemini, and Grok. Asked the default model ten ethical dilemmas ten times. Created a polling website for humans to answer the same questions and compare the answers. Somewhat interesting answers so far: most models answered the same way for each question each time, but the responses varied by provider. Only Grok claimed using ad-blocker was unethical. All were in favor of automation at the cost of human jobs. All but DeepSeek were against the idea of "speciesism".