OpenAI revises AI behavioral guidelines "Model Spec," specifying usage of "memory" and human supervision
OpenAI released a revised version of its AI behavior guidelines "Model Spec." It restructured the whole according to five values and stipulated the appropriate handling of memory and conversation history, as well as obedience to human supervision during autonomous agent operation. On the other hand, it deleted most personality settings and detailed style regulations, and changed the model designation from "the assistant" to "GPT.".
OpenAI released a revised version of its AI behavior guidelines "Model Spec." In this revision, the entire document was restructured according to five values.
The new guidelines stipulate the appropriate handling of memory and conversation history, as well as obedience to human supervision during autonomous agent operation. On the other hand, most personality settings and detailed style regulations have been deleted.
Additionally, the model designation has been changed from "the assistant" to "GPT.".
This item is a summary. The full, checkable version is at the original source:
Read the original at ITmedia NEWS ↗