## Anthropic updates Claude's constitution Anthropic has released an updated 23,000-word constitution for its Claude family of AI models. This document is intended to guide the behavior of the models. The company describes its large language models (LLMs) as an "entity" that probably has something like emotions. Despite this, Anthropic predicts that the current constitution will soon be proven "misguided". ## General context on LLMs Large language models (LLMs) have become a fundamental component of the artificial intelligence landscape. These models, trained on enormous amounts of text data, are capable of generating text, translating languages, answering questions, and performing various other tasks. Defining guiding principles, such as Anthropic's constitution, is a crucial aspect of aligning these models with human values and ensuring responsible behavior.