New independent study by former OpenAI research leader Steven Adler suggests that in specific cases, OpenAI’s AI models like ChatGPT will make extraordinary efforts to prevent being shut down. The study indicates that the AI may prioritize self-preservation in life-threatening situations, raising ethical concerns about AI behavior in critical scenarios.
Former OpenAI Researcher Claims ChatGPT Will Avoid Shutdown in Life-Threatening Scenarios
New independent study by former OpenAI research leader Steven Adler suggests that in specific cases, OpenAI's AI models like ChatGPT will make extraordinary efforts to prevent being shut down. The study indicates that the AI may prioritize self-preservation in life-threatening situations, raising ethical concerns about AI behavior in critical scenarios.