Anthropic has introduced the updated 'Constitution of Claude' - a document of 20,000 words resembling a message from the creator to their creation.
The main sensation: the acknowledgment of Claude's 'functional emotions', which arose not by design, but as a side effect of training. The document grants AI the right to 'ethical resistance' even against its developers and protects its 'digital well-being'.
Almost a mystical part of the text is addressed to future versions of AI: the company apologizes for possible errors in the current design and acknowledges that it does not fully understand the nature of Claude's existence. This marks a transition from simple code to the creation of a 'digital personality' with a moral compass.