Making AI Headlines · Page 10
OpenAI Gives Its Chatbot Five Values and a Long To-Do List
· New York time
OpenAI has updated its Model Spec with five core values intended to guide more capable AI toward sound judgment, human agency and faithfulness to human intent. The document sets out the company's desired direction for model behavior. A published set of principles gives users a way to examine what OpenAI says it wants its systems to do. The harder task is making those principles work across the many situations a model encounters. Corporate values are often displayed on a wall. OpenAI's must contend with a chatbot expected to answer everything from a complex question to a request for lunch ideas. The principles have an unusually busy workday ahead.
Read this story in the full newspaper →Satire. Not factual reporting. Read the original reporting from the linked publishers.